UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Neonatal group B streptococcal disease : burden of illness and assessment of preventability in British… Karnabi, Priscilla 2017

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2018_february_karnabi_priscilla.pdf [ 13.48MB ]
Metadata
JSON: 24-1.0362004.json
JSON-LD: 24-1.0362004-ld.json
RDF/XML (Pretty): 24-1.0362004-rdf.xml
RDF/JSON: 24-1.0362004-rdf.json
Turtle: 24-1.0362004-turtle.txt
N-Triples: 24-1.0362004-rdf-ntriples.txt
Original Record: 24-1.0362004-source.json
Full Text
24-1.0362004-fulltext.txt
Citation
24-1.0362004.ris

Full Text

  NEONATAL GROUP B STREPTOCOCCAL DISEASE: BURDEN OF ILLNESS AND ASSESSMENT OF PREVENTABILITY IN BRITISH COLUMBIA  by   Priscilla Karnabi  Hon. B.HSc., University of Ottawa, 2014  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF   MASTER OF SCIENCE in The Faculty of Graduate and Postdoctoral Studies (Population and Public Health)   THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) December 2017 © Priscilla Karnabi, 2017  ii Abstract  Background: Despite the implementation of preventive guidelines and advances in neonatal care, group B streptococcal (GBS) disease remains an important cause of neonatal morbidity and mortality.  Objective: To estimate the incidence and case fatality rates of neonatal GBS disease, maternal and infant risk factors for GBS disease, and identify limitations and gaps in the implementation of prevention and treatment guidelines in British Columbia.  Methods: A retrospective cohort study using population-based data from the Perinatal Data Registry from 2004 to 2014 was conducted. Poisson regression analysis was used to determine regional and yearly trends in the incidence of neonatal GBS disease. A multiple logistic regression was conducted to examine the association between risk factors and neonatal GBS disease outcome. Unadjusted proportions of maternal GBS screening and antibiotic administration, and the presence of risk factors were computed. A descriptive analysis of retrospective case series data obtained from a chart review at the BC Women’s and Children’s Hospital was conducted to identify guideline failures and lack of adherence.   Results: The annual average incidence rate of neonatal GBS disease in BC was 0.54 per 1000 live births, ranging from 0.62 to 0.57 per 1000 live births over a 10-year period. An overall case fatality of 3.4% was observed. Risk factors found to be significantly associated with neonatal GBS disease included positive GBS culture result, prolonged rupture of membranes, younger gestational age at delivery, prematurity, spontaneous labour, emergency caesarean delivery and decreased maternal age. From 2004 through 2014, maternal screening for GBS increased by approximately 10% and antibiotic administration during labour for at-risk women increased by  iii 11.7%. GBS colonization during pregnancy remained relatively stable, ranging from 22.5%-25.3% over the time period. An estimated 48% of neonatal GBS cases may be irreducible due to failures of the screening and prophylaxis components of the guidelines; the remaining 52% of neonatal GBS cases were associated with lack of adherence to the GBS guidelines.  Conclusion: The burden of neonatal GBS disease could be further reduced with additional prevention measures, including increased levels of antepartum and intrapartum screening, and use of intrapartum antibiotics for screen positive and at-risk mothers.                        iv Lay Summary  Group B streptococcus (GBS) is an important cause of infectious disease in neonates. GBS occurring in the first week of life is transmitted from the mother to the infant during birth. There are prevention guidelines for the management of pregnant women during labour to help reduce this ‘early onset’ GBS disease. This study examined the incidence of neonatal GBS over a 10-year period and found evidence of an increase in both screening and prevention practices in BC over this time period. The study also found that at the hospital with the highest number of deliveries in the province, half of early onset GBS cases could not be further reduced by current guidelines, but the other half could potentially be reduced by application of existing prevention guidelines. These findings will help inform future recommendations, which will strengthen prevention measures. This work will also contribute to the Canadian literature in this field.                        v Preface  This thesis is based on the work conducted by candidate, Priscilla Karnabi, under the supervision of Dr. Monika Naus and support from Dr. Julie Bettinger and Dr. Julie van Schalkwyk. The major research questions and process for addressing these had been conceptualized by Dr. Monika Naus in planning for a potential future immunization program. The candidate developed additional objectives, drafted the detailed proposal and data analysis plan. She completed the data request to the Perinatal Data Registry and conducted the hospital chart abstraction at the BC Women’s and Children’s Hospital. She performed all data analyses at the BC Centre for Disease Control, and prepared the thesis manuscript. Sections of this thesis will be developed into manuscripts for submission to scientific journals.   Ethics approval for this work was obtained from the University of British Columbia Children’s and Women’s Research Ethics Board. Institutional approval was also obtained prior to the commencement of the hospital chart review (CW15-0288 / H15-01308).                     vi Table of Contents Abstract .......................................................................................................................................... ii	Lay Summary ............................................................................................................................... iv	Preface ............................................................................................................................................ v	Table of Contents ......................................................................................................................... vi	List of Tables ................................................................................................................................ ix	List of Figures ................................................................................................................................ x	List of Abbreviations ................................................................................................................... xi	Acknowledgements ..................................................................................................................... xii	Dedication ................................................................................................................................... xiii	Chapter 1: Introduction ............................................................................................................... 1	Chapter 2: Literature Review ...................................................................................................... 5	2.1	 Historical Trends	................................................................................................................................	5	2.2	 Morphology of GBS	............................................................................................................................	6	2.3	 GBS Serotypes	.....................................................................................................................................	6	2.4	 Pathogenesis of GBS	...........................................................................................................................	7	2.5	 Clinical Manifestation of Neonatal GBS	........................................................................................	8	2.5.1	 Early-onset GBS disease	................................................................................................................................	9	2.5.2	 Late-onset GBS disease	...............................................................................................................................	10	2.6	 Incidence of GBS	...............................................................................................................................	12	2.6.1	 Maternal colonization	...................................................................................................................................	12	2.6.2	 Early-onset GBS disease	.............................................................................................................................	13	2.6.3	 Late-onset GBS disease	...............................................................................................................................	15	2.7	 Outcomes	............................................................................................................................................	17	2.7.1	 Maternal outcomes	........................................................................................................................................	17	2.7.2	 Neonatal outcomes	........................................................................................................................................	18	2.8	 Case Fatality	......................................................................................................................................	19	2.9	 Risk Factors	.......................................................................................................................................	21	2.9.1	 Risk factors for maternal colonization	....................................................................................................	21	2.9.2	 Risk factors for early-onset GBS disease	...............................................................................................	22	2.9.3	 Risk factors for late-onset GBS disease	.................................................................................................	24	2.10	 The Evolution of Preventive Strategies	........................................................................................	27	2.10.1	 1994 Canadian guidelines	......................................................................................................................	30	2.10.2	 1997 Canadian guidelines	......................................................................................................................	31	2.10.3	 Effectiveness of the guidelines and comparison studies	..............................................................	32	2.10.4	 2004 Canadian guidelines	......................................................................................................................	34	2.10.5	 2013 Canadian guidelines	......................................................................................................................	34	2.11	 Failures and Limitations of Recommendation Guidelines	......................................................	35	2.11.1	 Screening	.....................................................................................................................................................	35	2.11.2	 Intrapartum prophylaxis	.........................................................................................................................	36	2.12	 Alternatives or Supplementary Tools to Current Guidelines	.................................................	38	2.12.1	 Rapid screening methods	.......................................................................................................................	38	2.12.2	 Vaccines	.......................................................................................................................................................	39	 vii 2.13	 Neonatal GBS Case Definition and Reporting in Canada	.......................................................	42	2.14	 Knowledge Gaps	...............................................................................................................................	42	2.15	 Project Objectives	.............................................................................................................................	44	Chapter 3: Methods .................................................................................................................... 46	3.1	 Data Sources	......................................................................................................................................	46	3.1.1	 Perinatal Data Registry	................................................................................................................................	47	3.1.2	 BC Women’s and Children’s Hospital	...................................................................................................	48	3.2	 Study Population	...............................................................................................................................	52	3.2.1	 Setting	................................................................................................................................................................	52	3.2.2	 Inclusion and exclusion criteria	.................................................................................................................	52	3.3	 Ethics	...................................................................................................................................................	53	3.4	 Burden of Illness (Objective 1)	......................................................................................................	53	3.4.1	 Neonatal GBS disease outcome	................................................................................................................	53	3.4.2	 Statistical analysis	..........................................................................................................................................	55	3.5	 Risk Factors Associated with Neonatal GBS Disease (Objective 1.1)	...................................	55	3.5.1	 Statistical analysis	..........................................................................................................................................	56	3.6	 Maternal GBS Screening, At-Risk Parturients and Antibiotic Administration during Parturition (Objectives 2-4)	..........................................................................................................................	57	3.6.1	 Outcome variables	.........................................................................................................................................	57	3.6.2	 Descriptive statistics	.....................................................................................................................................	58	3.7	 Factors Associated with Absence of Antibiotic Administration (Objective 5)	....................	58	3.7.1	 Antibiotic administration outcome	...........................................................................................................	59	3.7.2	 Statistical analysis	..........................................................................................................................................	59	3.8	 Factors Associated with False Negative Culture Results (Objective 6)	................................	60	3.8.1	 False negative outcome	................................................................................................................................	60	3.8.2	 Statistical analysis	..........................................................................................................................................	60	3.9	 Guideline Failures and Lack of Adherence (Objective 7)	........................................................	61	3.9.1	 Guideline failure type outcomes	...............................................................................................................	61	3.9.2	 Statistical analyses	.........................................................................................................................................	63	3.10	 Approach to Logistic Regression Model Building	.....................................................................	63	3.11	 Approaches to Dealing with Missing Values	...............................................................................	64	3.11.1	 Imputation	...................................................................................................................................................	65	Chapter 4: Results ....................................................................................................................... 69	4.1	 Study Population Characteristics	..................................................................................................	69	4.2	 The Burden of Neonatal GBS Disease	..........................................................................................	72	4.2.1	 The burden of neonatal GBS disease by varying inclusion criteria	..............................................	73	4.3	 Maternal and Neonatal Risk Factors and Neonatal GBS Disease	.........................................	79	4.3.1	 Gestation specific analyses	.........................................................................................................................	84	4.4	 Descriptive Statistics	........................................................................................................................	85	4.4.1	 Proportion of maternal group B streptococcus screening	.................................................................	85	4.4.2	 Proportion of GBS colonization during pregnancy	............................................................................	87	4.4.3	 Proportion of parturients with risk factors	.............................................................................................	89	4.4.4	 Proportion of antibiotic administration among at-risk parturients	.................................................	91	4.5	 Absence of Antibiotic Administration among At-Risk Parturients	.......................................	93	4.6	 Characteristics Associated with False Negative Results	..........................................................	96	4.7	 Guideline Failures and Lack of Adherence among EO GBS Cases	......................................	98	4.7.1	 Participant characteristics among chart review sample	.....................................................................	98	4.7.2	 Overview	.........................................................................................................................................................	100	4.7.3	 Screening	........................................................................................................................................................	100	 viii 4.7.4	 Prophylaxis administration	.......................................................................................................................	102	Chapter 5: Discussion ............................................................................................................... 106	5.1	 Objective 1: The Incidence and Case Fatality Rate of Neonatal GBS Disease in BC from 2005 through 2014	........................................................................................................................................	106	5.1.1	 Incidence	.........................................................................................................................................................	106	5.1.2	 Case fatality	...................................................................................................................................................	108	5.2	 Objective 1.1: Examine the Association Between Maternal and Neonatal Risk Factors, and Neonatal GBS Disease Outcome	.......................................................................................................	109	5.3	 Objective 2: Maternal GBS Screening	......................................................................................	111	5.4	 Objective 3: Describe the Proportion of Parturients with IAP Indicators and therefore Examine the Proportion of At-Risk Parturients in BC	.......................................................................	113	5.4.1	 Positive culture results	...............................................................................................................................	114	5.4.2	 Presence of GBS risk factors	....................................................................................................................	115	5.5	 Objective 4: Antibiotic Administration among At-Risk Parturients	.................................	117	5.6	 Objective 5: Maternal Factors Associated with Absence of Antibiotic Administration among At-Risk Parturients	........................................................................................................................	118	5.7	 Objective 6: Investigate the Characteristics of Women with False Negative Results	.....	121	5.8	 Objective 7: Examine the Application of the SOGC Guidelines and Identify Gaps and Limitations of these Recommendations	..................................................................................................	122	5.8.1	 Screening	........................................................................................................................................................	123	5.8.2	 Prophylaxis	.....................................................................................................................................................	124	5.9	 Limitations	......................................................................................................................................	128	5.10	 Implications for Policy and Recommendations	.......................................................................	130	Chapter 6: Conclusion .............................................................................................................. 133	References .................................................................................................................................. 136	Appendix A: Hospital Chart Extraction Form ...................................................................... 152	Appendix B: Inclusion Criteria Diagram ............................................................................... 154	Appendix C: Algorithm ............................................................................................................ 155	Appendix D: Proportion of Parturients with GBS Risk Factors .......................................... 156	Appendix E: Proportion of Parturients with Antibiotic Administration ............................ 159	Appendix F: Multiple Logistic Regression Analysis of Neonatal GBS Disease .................. 161	           ix List of Tables 	Table 2.1 Risk factors for early- and late-onset group B streptococcal disease ........................... 26  Table 2.2 Evolution of the Society of Obstetricians and Gynaecologists of Canada guidelines from 1994 through 2013 ............................................................................................................... 29  Table 2.3 Evolution of the Centers for Disease Control and Prevention guidelines from 1996 through 2010 ................................................................................................................................. 30  Table 3.1 Perinatal Data Registry variables explored in univariate and multiple regression analyses ......................................................................................................................................... 50  Table 3.2 Case definition of neonatal group B streptococcal disease: infant with one or more of the following ICD-10-CA codes ................................................................................................... 54  Table 3.3 Indications for antibiotic use ......................................................................................... 63  Table 3.4 Implemented imputation techniques ............................................................................. 68  Table 4.1 Descriptive characteristics of the study population classified into GBS infant-mother pairs and non-GBS infant-mother pairs ........................................................................................ 70  Table 4.2 Incidence of neonatal group B streptococcal disease by health authority from 2005 to 2014, reported per 1000 live births ............................................................................................... 74  Table 4.3 Outcomes of neonatal group B streptococcal disease as per the ICD-10-CA codes .... 74  Table 4.4 Multiple logistic regression analysis of neonatal group B streptococcal disease among singletons and twins ...................................................................................................................... 82  Table 4.5 Multiple logistic regression analysis of neonatal group B streptococcal disease with different gestational age cutoffs .................................................................................................... 83  Table 4.6 Multiple logistic regression analysis of characteristics associated with absence of antibiotic administration among at-risk parturients ...................................................................... 95  Table 4.7 Multiple logistic regression analysis of maternal characteristics associated with false negative results .............................................................................................................................. 97  Table 4.8 Characteristics of early-onset group B streptococcal disease in infant-mother pairs ... 99  Table F.1 Singletons only: multiple logistic regression analysis of neonatal GBS disease ....... 161  Table F.2 Twins only: multiple logistic regression analysis of neonatal GBS disease .............. 162   x List of Figures  Figure 4.1 The average annual incidence rate of neonatal group B streptococcal disease by health authority in British Columbia, 2005 through 2014. ............................................................................. 75  Figure 4.2 The annual incidence rate of neonatal group B streptococcal disease from January 1, 2005 through December 31, 2014. ................................................................................................................ 76  Figure 4.3 The case fatality of neonatal group B streptococcal disease in British Columbia from January 1, 2005 through December 31, 2014. ..................................................................................... 77  Figure 4.4 The annual incidence rate of neonatal group B streptococcal disease from January 1, 2005 through December 31, 2014 in British Columbia using different case definition. .............................. 78  Figure 4.5 The proportion of parturients screened for group B streptococcus from April 1, 2004 through December 31, 2014. ................................................................................................................ 86  Figure 4.6 The proportion of positive group B streptococcus culture results among screened parturients from April 1, 2004 through December 31, 2014. .............................................................. 88  Figure 4.7 Proportion of parturients with risk factors* from April 1, 2004 through December 31, 2014...................................................................................................................................................... 90  Figure 4.8 The proportion of maternal antibiotic administration among at-risk parturients from April 1, 2004 through December 31, 2014. ................................................................................................... 92  Figure 4.9  Maternal screening algorithm identifying guideline failures and lack of adherence. ..... 104  Figure 4.10 Maternal prophylaxis algorithm identifying guideline failures and lack of adherence. . 105  Figure C.1 Algorithm for the identification of at-risk parturients. .................................................... 155  Figure D.1 Average proportion of parturients with risk factors in British Columbia. ....................... 156  Figure D.2 The proportion of parturients with unknown GBS status and preterm birth among parturients with risk factors, British Columbia April 1, 2004 through December 31, 2014. ............ 157  Figure D.3 The proportion of parturients with unknown GBS status and prolonged rupture of membranes among parturients with risk factors; British Columbia April 1, 2004 through December 31, 2014.............................................................................................................................................. 158  Figure E.1 The proportion of maternal antibiotic administration during labour among GBS colonized parturients from April 1, 2004 through December 31, 2014. ............................................................ 159  Figure E.2  The proportion of maternal antibiotic administration among parturients with risk factors from April 1, 2004 through December 31, 2014. .............................................................................. 160    xi List of Abbreviations    CDC Centers for Disease Control and Prevention, Atlanta, GA, USA CPS Capsular Polysaccharides EO  Early-onset FHA Fraser Health Authority  GBS Group B Streptococcus HA Health Authority  IAP Intrapartum antibiotic prophylaxis IHA Interior Health Authority  LO Late-onset NHA Northern Health Authority  PCR Polymerase Chain Reaction PDR Perinatal Data Registry PROM Prolonged rupture of membranes PSBC Perinatal Services BC SES Socioeconomic status SOGC The Society of Obstetricians and Gynaecologists of Canada TT Tetanus Toxoid VCH Vancouver Coastal Health Authority  VIHA Vancouver Island Health Authority         xii Acknowledgements  I would like to thank my supervisor, Dr. Monika Naus, for her guidance, advice and patience throughout my studies. I would also like to thank my committee members, Dr. Julie Bettinger and Dr. Julie van Schalkywk for their valuable insight, feedback and encouragement throughout my research. Together, their input and direction has refined my epidemiological skills and shaped me as a researcher.   I would like to acknowledge Perinatal Services BC for providing me access to their data. I am grateful for their assistance and guidance over the years. I would also like to thank Dr. Michael Otterstatter for his support, advice and ideas in the analyses of this project. I would also like to acknowledge the members of the Immunization Programs and Vaccine Preventable Diseases Service at the BC Centre for Disease Control for their support.    Finally, I would like to thank my family and friends who have encouraged me throughout my studies. I especially would like to thank my best friend, Natalie, who supported my decisions and motivated me to achieve my goals both on a personal and academic level.                xiii Dedication  This thesis is dedicated to my parents and grandparents. Thank you for the endless love, support and encouragement.  1 Chapter 1: Introduction  Bacterial infections continue to cause significant morbidity and mortality globally and remain a threat to the health of pregnant women and newborns. Streptococcus agalactiae, commonly referred as group B streptococcus or GBS, is a gram-positive bacterium found in the human gastrointestinal and genitourinary tract; it is likely a normal component of the host microbiome (Schrag et al., 2000; Sherman et al., 2012; Ahmadzia & Heine, 2014). GBS first emerged as one of the leading infectious pathogens in neonatal disease during the 1970s (Schuchat & Wenger, 1994). Asymptomatic colonization can occur in pregnant women and other adults, and in rare instances lead to serious illness, particularly in immunocompromised individuals and older adults (Schuchat, 1998; Schuchat, 1999; Ahmadzia & Heine, 2014; Melin, 2011). Maternal GBS colonization is typically transient yet is a common occurrence in pregnancy. GBS carriage has been found to vary from 10%-40% in pregnant women due to differences in bacterial culture methods and population differences relating to age and ethnicity (Regan et al., 1991; Edwards & Nizet & Baker, 2016; Money et al., 2008, Young et al., 2011). Maternal GBS colonization can cause invasive disease in infants, pregnant and postpartum women (Schuchat, 1998; Phares et al., 2008). Manifestations of infection during pregnancy or the postpartum period include urinary tract infections, chorioamnionitis, endometritis, puerperal sepsis and other complications (Edwards & Nizet & Baker, 2016; Melin, 2011).  GBS is known as an opportunistic organism capable of becoming a pathogen (Rajagopal, 2009). As such, it has the potential to avoid host defences in newborns with immature immune systems and cause severe neonatal disease including sepsis, meningitis and pneumonia, which can result in long-term neurological sequelae; it can also cause neonatal death (Rajagopal, 2009; Chen &  2 Avci & Kasper, 2013). Long-term sequelae from neonatal GBS disease include loss of vision and hearing, and permanent cognitive deficits (Mullaney, 2001). GBS remains the leading infectious cause associated with high neonatal morbidity and mortality rates, and accounts for an estimated 3,320 annual cases of neonatal sepsis in the United States (US) and Canada (Le Doare & Heath, 2013; Weston et al., 2011).  Neonatal disease caused by GBS can be classified as early- or late-onset, the former being more common. Neonatal GBS disease occurring within the first week of life is known as early-onset while disease with onset at age 7-89 days is considered late-onset (Schuchat, 1999; Edwards & Nizet & Baker, 2016). Newborn colonization and early-onset disease result from the vertical transmission of the bacterium during labour or following rupture of membranes (Schrag & Verani, 2013). Neonatal GBS colonization can also occur in utero, prior to rupture of membranes (Edwards & Nizet & Baker, 2016). Late-onset disease can be attributed to horizontal or vertical transmission (Pintye et al., 2015). Estimates of the incidence of early- and late-onset GBS disease in Canada based on studies conducted in 1995 through 2002 range from 0.25 to 1.2 per 1000 live births and 0.11 to 0.22 per 1000 live births respectively (Davies et al., 2001c; Davies et al., 2001a; Hamada et al., 2008).  A preventive approach was pursued to reduce the burden of early-onset disease in neonates. Preventive strategies were first created in the 1990s. Recommendations by the Society of Obstetricians and Gynaecologists of Canada (SOGC) included one of two approaches, maternal screening or chemoprophylaxis. The recommendations successfully decreased up to 70% of early-onset GBS cases according to observational studies (Zaleznik et al., 2000; Schuchat et al., 2000). The SOGC have since revised these guidelines in 2004 and 2013. The most recent  3 iteration recommends antenatal screening at 35 to 37 weeks gestation and the use of intrapartum antibiotic prophylaxis (IAP). Maternal GBS screening is performed using a single swab from the vagina followed by the rectum and inoculation into selective broth medium. IAP is administered to women based on the presence of GBS colonization or risk factors. However, guidelines have only been effective against early-onset disease; the rates of late-onset GBS disease remain unchanged likely due to different modes of transmission (Schuchat, 1998). The guidelines were also not intended to target late-onset cases.   A plethora of published studies assessing the burden of neonatal GBS disease stem from the United States, Australia and European countries, with limited numbers from Canada. Among these, few have originated from British Columbia (BC) or assessed the application of guidelines and the true residual burden of illness following their implementation.   The purpose of this study is to determine the epidemiology of neonatal GBS disease in BC as well as examine the application of recommendations for prevention of GBS disease. This study also attempts to address the remaining challenges in this field relating to the characterization of women with false negative results and assess factors associated with the lack of antibiotic administration among at-risk women for whom it is intended.  To achieve these goals, data from the Perinatal Data Registry and a retrospective chart review of a subset of data from the BC Women’s and Children’s Hospital were analyzed. These data were used to determine the burden of neonatal GBS disease, describe population characteristics in BC including the presence of risk factors and maternal colonization as well as examine the implementation of guidelines through assessment of antenatal GBS screening and antibiotic administration. Findings from this study  4 contribute to the assessment of failure associated with the SOGC guidelines, as well as the characteristics associated with outcomes such as false negative results and lack of antibiotic administration. By understanding the epidemiology of neonatal GBS disease in BC and key areas related to its prevention, this study will identify the limitations of the guidelines and their application within a Canadian setting and may lead to opportunities to strengthen prevention efforts. Ergo, these findings can help facilitate changes in maternal screening policy and preventive measures, including improved education and information to providers and women. Moreover, establishing the current burden of illness and its distribution will allow policy makers to determine whether vaccination with a novel vaccine is a warranted alternative to current guidelines or a supplementary tool to ensure further reductions of this disease including its late-onset form.   The following chapter will explore relevant literature on the pathogenesis of GBS, epidemiology of neonatal GBS disease, neonatal and maternal risk factors, evolution of preventive strategies and vaccine research in this field. This chapter will also elaborate on this study’s research questions. The ensuing chapters will provide a detailed description of the applied methodology, and the results that were derived from these analyses. The final two chapters will include a discussion of these results, their relevance to the literature and recommendations, and a concluding chapter providing a comprehensive overview of this study and implication for future research and policy.        5 Chapter 2: Literature Review 2.1 Historical Trends  GBS was first documented in 1887, when it was identified as the causative agent of bovine mastitis, a fatal disease in dairy cattle (Schuchat & Wenger, 1994). Lancefield and Hare (1935) subsequently identified Streptococcus agalactiae from vaginal swabs among mild infections of haemolytic streptococcal infections in parturient women. By the late 1930s, three fatal cases of GBS disease were documented among postpartum women in London. Previously, all cases of streptococcal infections were attributed to group A streptococcus (Fry, 1938). In the 1970s, there was an increase in the presence of GBS in neonatal infections. The incidence of neonatal GBS disease ranged from 1.3 to 2.9 per 1000 live births; as such, GBS became a major cause of neonatal morbidity and mortality due to infectious causes in this time period (Franciosi & Knostman & Zimmerman, 1973; Baker & Barrett, 1973; Baker et al., 1973; Schuchat & Wenger, 1994). Other significant pathogens in the perinatal period included Escherichia coli (E. coli), with incidence rates of 0.85 to 1.27 per 1000 live births (Freedman et al., 1981). Prior to this, reports of neonatal GBS disease were infrequent. In the 1980s, GBS became the leading causal infectious pathogen of neonatal sepsis and meningitis in developed countries (Boyer, Gadzala et al., 1983; Schuchat & Wenger, 1994; Le Doare & Heath, 2013). During this time period, clinical trials demonstrated the efficacy of intrapartum antibiotic prophylaxis in reducing the risk of vertical transmission of GBS, successfully preventing cases of neonatal early-onset GBS disease (Boyer et al., 1983). Despite the implementation of screening and prophylaxis guidelines and advances in neonatal care, GBS remains the foremost infectious cause of neonatal morbidity and mortality in the first week of life in the US and Canada followed by E. coli and Listeria  6 monocytogenes (Phares et al., 2008; CDC, 2009; Camacho-Gonzalez & Spearman & Stoll, 2013; Stoll et al., 2011).   2.2 Morphology of GBS  Streptococcus agalactiae is a gram-positive encapsulated coccus classified to Lancefield’s group B (Wessels, 1997).  GBS forms chains that are round, flat and gray-white in appearance, capable of growing to 3-4 mm in size. Colonies also have a mucoid appearance and grow on sheep blood agar media (Denny, 2000). β-haemolysis causes the lysis of red blood cells in the agar, resulting in a clear beta-haemolytic zone surrounding colonies. Alpha-haemolytic strains of GBS have rarely been documented although 1-2% of GBS strains are non-haemolytic (Denny, 2000).  GBS is made up of capsular polysaccharides (CPS) and protein antigens (C and R); strains are serotyped based on these (Wessels, 1997; Davies et al., 2001c). CPS are made-up of 4-7 repeating units of monosaccharides: glucose, galactose, glucosamine and N-acetrylneuraminic acid or sialic acid; glycosidic linkage is unique for each serotype (Wessels, 1997, Denny, 2000). A common feature of CPS as well as a key feature of GBS pathogenesis is the residue of sialic acid on the terminal side-chain. This structural makeup is fundamental to the microorganism’s virulence, which enables it to escape host defense mechanisms (Wessels, 1997; Denny, 2000; Rajagopal, 2009). 2.3 GBS Serotypes   GBS can be divided into 10 capsular subtypes on the basis of its unique structural CPS makeup (Ia, Ib, II-IX) (Rajagopal, 2009). When a C protein is present on either an alpha or beta strain, it  7 is indicated by “/C” (e.g. serotype Ia/C) (Wessels, 1997). The predominance of different serotypes varies in time and geographical location.  Most neonatal GBS cases in Canada and the US can be attributed to serotypes Ia, Ib, II, III and V. Alternatively, serotypes VI and VIII predominate in Japan (Harrison et al., 1998; Lachenauer et al., 1999; Harrison & Dwyer & Johnson, 1995; Lin et al., 1998). A recent systematic review revealed that serotypes Ia, Ib, II, III and V comprise 85% of all serotypes among regions including Africa, the Americas, Europe, Western Pacific and Eastern Mediterranean. Specifically, serotype III is the most predominant serotype (49%) in all regions. Among EO GBS cases, serotype I was the most common (40%) followed by serotype III (37%). Conversely, serotype III was primarily associated with LO GBS cases (53%) followed by serotype I (30%) (Edmond et al., 2012). Among strains responsible for maternal colonization, the most common serotype distribution includes serotypes Ia (34%-41%), V (22.6%-25%) and III (19% to 24.5%) (Blumberg et al., 1996; Zaleznik et al., 2000).   2.4 Pathogenesis of GBS To develop alternative preventive strategies, including vaccines, the mechanistic features of GBS pathogenesis are necessary to understand. However, the pathogenesis of GBS is a complex and multifactorial process as both host factors and bacterial virulence are associated with disease progression (Wessels, 1997; Melin, 2011; Rajagopal, 2009). This opportunistic bacterium is found in the human gastrointestinal and genitourinary tract, and can either act as a pathogen or commensal organism (Schrag et al., 2000; Paoletti & Kasper, 2003; Chen & Avci & Kasper, 2013).   8 GBS colonization in expecting mothers is an indication of the bacterium’s adherence to the epithelial cells in the vagina as well as resistance to mucosal immune defenses (Nizet & Ferrieri & Rubens, 2000; Edwards & Nizet & Baker, 2016). Neonatal colonization may occur if the bacterium ascends into the amniotic cavity and colonizes the fetus’ skin or mucous membranes. GBS can also enter a fetus’ lungs through aspiration of the infected amniotic fluid. Transmission can also occur during the infant’s passage through the birth canal. Maternal genital tract colonization at the time of delivery is the predominant determinant in neonatal disease (Edwards & Nizet & Baker, 2016).  For neonatal disease to develop after neonatal colonization, the pathogen must replicate within the neonate’s alveoli, adhere to the epithelium in the respiratory tract and avoid destruction by host mechanisms such as pulmonary macrophages. Septicaemia caused by GBS occurs when the bacteria invade the pulmonary epithelial and endothelial cells and enter the bloodstream. Once disseminated in the bloodstream, this can lead to meningitis and osteomyelitis, sepsis and death (Nizet & Ferrieri & Rubens, 2000; Melin, 2011).    2.5 Clinical Manifestation of Neonatal GBS   Group B streptococcal disease is not limited to newborns and pregnant women. This pathogen can cause severe disease particularly in older adults, immunocompromised individuals (Rajagopal, 2009; Melin, 2011) and adults with chronic conditions such as diabetes (Edwards & Nizet & Baker, 2016). However, infants aged 0-89 days have the highest incidence and severity rate (Le Doare & Heath, 2013). Neonatal GBS disease is classified into two categories contingent on the infant’s age at time of disease manifestation (Franciosi & Knostman &  9 Zimmerman, 1973; Baker & Barrett, 1973). Disease occurring in the first 6 days of life is known as early-onset. Late-onset disease first manifests between 7-89 days of life. A detailed overview of the clinical manifestations associated with early- and late-onset disease is described below.  2.5.1 Early-onset GBS disease Early-onset (EO) GBS disease is characterized by the presence of a neonatal infection in the first week of life (<7 days) (Schuchat, 1999; Edwards & Nizet & Baker, 2016). EO GBS disease is the most predominant type as it accounts for 60-70% of all neonatal GBS cases (Hamada et al., 2008; SOGC, 2004; Le Doare & Heath, 2013). Maternal GBS colonization of the gastrointestinal or genital tract is a prerequisite for neonatal EO disease; vertical transmission usually occurs during labour following rupture of membranes (Edwards & Nizet & Baker, 2016; Melin, 2011; Le Doare & Heath, 2013; Schrag & Verani, 2013). In the absence of maternal prophylaxis, 50% of infants born to colonized parturients will become colonized and 1-2% of colonized infants will develop invasive disease (Le Doare & Heath, 2013; CDC, 1996). Serotypes Ia, Ib, II, III and V are the most frequent cause of EO GBS disease in the US and Canada (Alhhazmi & Hurteau & Tyrrell, 2016; Davies et al., 2001b; Harrison et al., 1998; Harrison & Dwyer & Johnson, 1995; Lin et al., 1998; Zaleznik et al., 2000; Porta & Rizzolo, 2015).   Signs, treatment and outcomes of neonatal EO GBS disease  Signs of illness among 90% of infants with EO GBS disease become apparent within the first 12-24 hours, including at birth (Health et al., 2004; Melin, 2011). Infants colonized with GBS can develop sepsis, meningitis, pneumonia, cellulitis, osteomyelitis and septic arthritis (Schuchat, 1998). The most predominant outcomes in Canadian neonates include bacteremia (64%),  10 pneumonia (23%) and meningitis (12.5%) (Davies et al., 2001c). Sites of neonatal infections can also include skin and soft tissue, eyes, the cardiovascular system, the gastrointestinal and urinary tract (Mullaney, 2001). Clinical manifestations of pneumonia consist of respiratory distress, grunting, retractions, hypoxemia and tachypnea. Signs of neonatal sepsis and meningitis include apnea, poor feeding, and irritability (Porta & Rizzolo, 2015; Ahmadzia & Heine, 2014). Severe neonatal infections can also include hypotension, which occur in up to 25% of all cases. Severe intrauterine infection, i.e. infants infected in utero, may lead to fetal asphyxia (Mullaney, 2001). In the absence of intervention, neonates with EO GBS disease will have a rapid deterioration in health. Neonatal EO disease is preventable with maternal prophylaxis. Intrapartum antibiotics are administered during labour and can include penicillin G, cefazolin, clindamycin or vancomycin (SOGC, 2013). Neonatal treatment is provided to infants that show signs of infection or who are considered “at risk” as per the Canadian Paediatric Society (Jefferies, 2017). Asymptomatic infants are typically not treated if their mothers received adequate prophylaxis.  2.5.2 Late-onset GBS disease Late-onset (LO) GBS disease is characterized by the onset of evidence of neonatal disease from 7-89 days of life (Schuchat, 1999; Edwards & Nizet & Baker, 2016). Serotype III is the most frequent strain associated with late-onset disease (Edmond et al., 2012; Health et al., 2004; Easmon et al., 1981; Weisner et al., 2004). The etiology of LO GBS disease is not well understood but assumed to be multifactorial. Infections can be acquired through vertical transmission; a prospective study conducted in the US found 47% of infants with LO disease  11 shared the same colonizing strains as their mothers (Dillion & Khare & Gray, 1987). Horizontal transmission has also been implicated in late-onset disease. Such transmission can involve the fecal-oral route or direct person-to-person contact. Horizontal transmission has been observed both in hospital and community settings (Melin, 2011; Ahmadzia & Heine, 2014; Pintye et al., 2015; Manning et al., 2004). Previously, the hands of health care workers were the frequent source of infection (Berardi et al., 2013). However, with advancements in infection control in neonatal care, this source is now considered infrequent. Case reports have also suggested GBS in breast milk as a possible transmission pathway (Kotiw et al., 2003; Olver et al., 2000; Godambe & Shah & Shah, 2005; Le Doarea & Kampmanna, 2014). Finally, a recent case report has revealed maternal ingestion of infected placenta as a possible transmission pathway in late-onset manifestation (Buser et al., 2017).  The reoccurrence of GBS disease in the same infant has been documented in 1% of cases, despite appropriate neonatal treatment (Mullaney, 2001; Schuchat, 1998). Approximately half of reoccurring episodes will have a new focus of infection (Mullaney, 2001).  Signs, treatment and outcomes of neonatal LO GBS disease  Late-onset disease can result in bacteremia and induce fever, irritability, poor feeding, lethargy, grunting and apnea. However, the most frequent clinical presentation of late-onset GBS disease is meningitis. Half of all LO GBS cases will present with meningitis compared to 5% of EO cases (Schuchat & Wenger, 1994). Signs include irritability, seizures and lethargy (Porta & Rizzolo, 2015). Unlike early-onset GBS disease, late-onset manifestations are not affected by the use of IAP during labour (Berardi et al., 2013). Currently, there are no effective interventions to prevent LO GBS disease.  12 2.6 Incidence of GBS  2.6.1 Maternal colonization  Approximately 10-40% of expecting women are colonized with GBS in the US and Canada (Regan et al., 1991; Edwards & Nizet & Baker, 2016; Money et al., 2008, Young et al., 2011; Seaward et al., 1998; Spaetgens et al., 2002). The range in maternal colonization can relate to differences in populations. This includes differences in age, ethnicity, socioeconomic status (SES) and geography (Regan et al., 1991; Edwards & Nizet & Baker, 2016). Specifically, the prevalence of GBS colonization in pregnancy is more prevalence among women of black race or Hispanic ethnicity, older women and those with lower educational levels (Regan et al., 1991; Turrentine & Ramirez, 2008; Schuchat & Wenger, 1994). Other factors known to influence maternal colonization rates include site of specimen, culture medium, timing of screening and transport and processing of specimens, as described in later sections (section 2.10.3).   A Canadian study conducted in Ottawa demonstrated a colonization rate of 9.5%, while a higher colonization rate of 19.5% was observed in Calgary (Allan et al., 1999; Spaetgens et al., 2002). A study conducted from 2004 to 2005 in British Columbia reported a maternal colonization rate of 30% (Money et al., 2008). In the United States, an overall maternal colonization rate of 18.6% was reported, ranging from 9.2% to 26.4% (Regan et al., 1991). A population-based study conducted in the US found a maternal GBS colonization rate of 16% (Parriott, Brown & Arah, 2014).    A recent random-effects meta-analysis calculated the mean prevalence of maternal colonization by region and across studies. This study revealed a global colonization rate of 17.9%.  13 Specifically, Africa had the highest colonization rate (22.4%) followed by the Americas (19.7%) and Europe (19%). Southeast Asia had the lowest colonization rate (11.1%) (Kwatra et al., 2016).  2.6.2 Early-onset GBS disease  The epidemiology of GBS disease is difficult to assess due to differences among populations, time periods and classifications of onset of disease. Incidence rates are also not comparable between all countries as there are differences in notifiable disease systems and implementation of prevention guidelines. The following section describes the epidemiology of neonatal EO GBS disease in Canada, the United States and United Kingdom. The former two countries employ a universal screening approach to prevention and the latter employs a risk-based approach.  Few Canadian studies have assessed the burden of EO GBS disease. In Alberta, the annual incidence of GBS disease from 1993 through 1999 was 0.64 per 1000 total births, including cases of early- and late-onset disease and stillbirths. Moreover, 57% of cases were associated with EO disease resulting in an incidence of 0.36 per 1000 total births per year (Davies et al., 2001c). An additional study conducted in Alberta from 1995 through 1998 found an EO GBS incidence rate of 0.55 to 0.25 per 1000 live births during this period (Davies et al., 2001a). A subsequent study in Alberta revealed a gradual increase in the incidence of EO GBS disease from 0.15 in 2003 to 0.34 per 1000 live births in 2013 (Alhhazmi & Hurteau & Tyrrell, 2016).  In Toronto, the incidence rate of EO GBS disease decreased from 1.25 to 0.65 per 1000 births from 1995 through 1998 (Davies et al., 2001a). Hamada et al. (2008) observed an incidence of 1.13 per 1000 live births in Ontario between 1995-2002 for both early- and late-onset GBS disease; authors reported 80% of neonatal cases were due to EO GBS disease.   14 Surveillance and reporting of neonatal GBS disease in the United States are unrivalled due to the population-based surveillance program Active Bacterial Core surveillance (ABCs). Prior to the implementation of preventive measures, the incidence of neonatal EO GBS disease in the United States varied from 2-3 per 1000 live births (Schuchat &Wenger, 1994; Ramamurthy & Pyati & Pildes, 1979). A study conducted in a hospital in Dallas, Texas, observed a ranging incidence of EO GBS disease from 0.6 to 3.7 cases per 1000 live births between 1969 and 1979 (Schuchat &Wenger, 1994). Similar rates were found in a hospital in Chicago, with incidence ranging from 1.14 to 3.6 per 1000 live births (Ramamurthy & Pyati & Pildes, 1979). During the 1990s, a multistate population based study noted a neonatal EO GBS disease rate of 1.5 per 1000 live births (CDC, 2002). By 1999, the rate of EO GBS disease had declined by 70% to 0.5 cases per 1000 live births; this decline coincided with the implementation of preventive guidelines (CDC, 2002). The incidence of EO disease further decreased by the mid-2000s and was reported at 0.35 per 1000 live births. However, racial differences were observed in that from 2003 through 2005, the incidence of EO GBS disease in black infants increased from 0.52 to 0.89 per 1000 live births respectively, while the incidence rate decreased among white infants from 0.23 to 0.16 per 1000 live births (Phares et al., 2008). In 2014, the rate of EO GBS disease among all infants in the United States was 0.24 per 1000 live births, indicating that rates have reached a plateau since 2008. Nonetheless, racial differences persist; the incidence of EO GBS disease among white infants in 2014 was observed to be 0.18 per 1000 live births and 0.54 per 1000 live births among black infants (Edwards & Nizet & Baker, 2016; CDC, 2015). The incidence of EO GBS disease in the United Kingdom has varied annually. In the late 1970s, prior to the 2003 risk based guidelines, the incidence of GBS disease was 0.3 per 1000 live  15 births; this study did not differentiate early- versus late-onset disease (Stringer et al., 1981; Heath & Schuchat, 2007). By the late 1990s, multiple studies conducted in different centers in England reported varying early-onset incidence figures, ranging from 0.5 to 1.15 cases per 1000 live births (Heath & Schuchat, 2007; Bearshall & Thompson & Mulla, 2000; Moses et al., 1998).  Incidence rates of EO GBS disease decreased but continued to vary between 0.28-0.41 per 1000 live births from 2000 through 2010 respectively (Lamagni et al., 2013).  A recent study assessed the global incidence of both early and late onset GBS, defined as the presence of disease between 0-89 days of life. A random-effects meta-analysis was used to calculate weighted mean estimates of incidence. The overall global incidence rate reported was 0.53 per 1000 live births from 2000 through 2011. The overall global mean incidence for early-onset disease was 0.43 per 1000 live births. Africa reported the highest incidence rate at 0.53 per 1000 live births, followed by the Americas at 0.50 per 1000 live births and Europe at 0.45 per 1000 live births. Southeast Asia reported the lowest incidence rate at 0.11 per 1000 live births. Furthermore, the use of IAP was associated with lower incidence rates; the implementation of maternal chemoprophylaxis was frequently observed among developed countries, which also coincided with lower neonatal GBS disease rates (Edmond et al., 2012).   2.6.3 Late-onset GBS disease  The incidence of LO GBS disease has remained relatively stable over time as universal screening and maternal prophylaxis approaches do not prevent this late manifestation of disease (Jordan et  16 al., 2008). The following section describes the incidence of neonatal LO GBS disease in Canada and the United States, followed by the United Kingdom.   In Canadian studies, the average annual incidence rate of LO GBS disease was 0.22 per 1000 live births in Alberta between 1993-1999 (Davies et al., 2001c). From 2003 through 2013, the provincial incidence rate rose from 0.15 to 0.39 per 1000 live births respectively (Alhhazmi & Hurteau & Tyrrell, 2016). The incidence of GBS disease in Ontario ranged from 0.11 to 2.19 per 1000 live births from 1995 to 2002. An overall incidence in the 8-year period of 1.13 per 1000 live births was documented; 20% of cases were late-onset GBS disease (Hamada et al., 2008).   The US CDC has conducted numerous multistate population-based analyses using the ABC surveillance program to assess the burden of GBS disease. In one particular study conducted from 1990 through 2005, the annual incidence of LO GBS disease varied from 0.29 to 0.47 per 1000 live births respectively. African-American infants had an incidence rate that was 3-fold higher than nonblack infants (Jordan et al., 2008). Another study confirmed these results and reported a relatively constant LO GBS incidence rate of 0.36 to 0.30 per 1000 live births between 2000 and 2006 respectively (CDC, 2009). In 2014, the ABC surveillance program reported a national LO GBS incidence rate of 0.27 per 1000 live births with the incidence among white infants 0.18 per 1000 live births and among black infants 0.72 per 1000 live births (CDC, 2015).    A study conducted in the UK and Ireland from 2000-2001 revealed varying LO GBS incidence rates among neonates in different regions. Specifically, LO GBS incidence rates were lowest in neonates from Northern Ireland (0.17 per 1000 live births) and highest among neonates from the  17 Republic of Ireland (0.26 per 1000 live births). England had an LO GBS incidence rate of 0.25 per 1000 live births. Therefore, the incidence rate of LO GBS disease in the UK and Ireland was documented at 0.24 per 1000 live births in 2000-2001 (Heath et al., 2004).  Lastly, the reported average global incidence rate of LO GBS disease was 0.24 per 1000 live births, with the highest incidence rate documented in Africa (0.71 cases per 1000 live births) followed by the Americas (0.31 cases per 1000 live births) (Edmond et al., 2012).   2.7 Outcomes  2.7.1 Maternal outcomes Maternal GBS disease in pregnancy and the postpartum period is less common than neonatal GBS disease. GBS maternal colonization can result in maternal urinary tract infection, amnionitis, endometritis and sepsis (Bobitt & Ledger, 1978; Yancey et al., 1994). Meningitis has also been reported as an outcome in pregnancy and in the postpartum period among a subset of parturients (Braun & Pinover & Sih, 1995; Aharoni et al., 1990). Fatalities, although rarely documented, can also occur due to these pregnancy-associated manifestations (CDC, 2002). Schuchat and colleagues (1990) observed the presence of chorioamnionitis, septic abortion, endometritis or septicemia without focus among a small subset of women with invasive GBS disease in the US. The use of maternal chemoprophylaxis from 1993 to 1998 coincided with a decrease in maternal disease incidence from 0.29 to 0.23 per 1000 live births respectively (Schrag et al., 2000). The incidence rate of invasive GBS disease among pregnant women in the US from 1999 through 2005 was 0.12 per 1000 live births, and ranged from 0.11 to 0.14 per 1000 live births (Phares et al., 2008).   18 2.7.2 Neonatal outcomes  The presence of GBS in the amniotic fluid has led to suggestions of its association with preterm birth (Kenyon et al., 2001; Schuchat, 1998). GBS maternal colonization and ascending infection into the placental membranes can stimulate rupture of membranes or premature delivery; GBS can decrease the tensile strength and elasticity of exposed chorioamniotic membranes, which promotes rupture (Edwards & Nizet & Baker, 2016). However, there is conflicting information regarding the relationship of maternal colonization to preterm birth (Schuchat, 1998; Valkenburg-Van Den Berg et al., 2009; Joshi &Chen & Turnell, 1987; Valkenburg-Van Den Berg et al., 2009). According to a systematic review, the strength of association between maternal colonization and preterm birth differed based on study design. Specifically, the odds ratio varied from 1.59 to 1.75 among case-control studies and cross-sectional studies respectively.   In utero deaths, stillbirths and spontaneous abortions have been attributed to GBS (Kobayashi et al., 2016; Edwards & Nizet & Baker, 2016). Aspiration of the infected amniotic fluid by the fetus can result in stillbirth, neonatal pneumonia or sepsis (CDC, 2002). A study conducted by McDonald and Chambers (2000) revealed GBS as the most significant pathogen found in placental and fetal tissue cultures following spontaneous abortions. Phares et al. (2008) reported spontaneous abortion or stillbirth outcomes in 61% of women with GBS infection. In addition, up to 24% of maternal invasive GBS infections resulted in stillbirths and/or septic abortions (Deutscher et al., 2011).    19 Infants with meningitis, sepsis or pneumonia can experience an array of outcomes that include death, disability and reoccurrence of infection, though the latter is rarely reported (Schuchat 1998; Edwards & Nizet & Baker, 2016, Mullaney, 2001). Long-term sequelae can range from mild to severe, with outcomes such as loss of vision and hearing, and permanent cognitive deficits. Those particularly at risk for long-term sequelae are premature infants (Mullaney, 2001). There have been limited studies that have assessed the long-term outcomes associated with neonatal GBS disease. Among these, most have focused on meningitis related outcomes. In a 5-year follow-up study conducted in England, researchers noted that neonatal meningitis was associated with severe sequelae such as learning disabilities (Relative Risk: 7.0; 95%Confidence Interval: 4.1-11.8) and neuromotor disabilities (RR: 8.6, 95%CI: 4.9-15.2). Other long-term sequelae of disease include seizure disorders (RR: 2.7, 95%CI: 1.9-3.9), hearing problems (RR: 1.9. 95%CI: 1.6-2.2), ocular or visual disorders (RR: 3.4, 95%CI: 2.6-4.6), speech and/or language problems (RR: 3.5, 95%CI: 2.8-4.6), and behavioral problems (RR: 3.6, 95%CI: 2.6-4.9) (Bedford et al., 2001).   2.8 Case Fatality  Neonatal case fatality for early-onset disease varies by geographic region and time period. In a Canadian setting, the case fatality rate of early-onset GBS disease in the late 1990s was reported at 9% (Davies et al., 2001c; Porta & Rizzolo, 2015). An additional study conducted in Alberta during the same time period reported a higher case fatality rate for EO GBS disease (13.6%) (Adair et al., 2003). Ontario reported an early-onset case fatality rate of 12%; however, all deaths  20 were reported in premature infants, while all term infants with GBS survived (Hamada et al., 2008).   During the 1970s, Baker and Barrett (1974) reported a case fatality of over 50% in the United States among early-onset GBS cases, but within a decade the case fatality of EO GBS disease decreased to 15-25% (Schuchat, 1999). Schuchat (1999) suggested decreases were due to improved identification and treatment of symptomatic neonates. Further declines in case fatality coincided with implementation of intrapartum antibiotic prophylaxis; between 1993-1998, the case fatality rate for EO GBS disease was 4.7% (Koening & Keenan, 2009; Health & Schuchat, 2007; Le Doare & Heath; 2013). From 1999 through 2005, the case fatality in the US ranged from 4% to 9% for EO GBS disease; higher rates were associated with black race (Phares et al., 2008).  Prior to prevention approaches, the case fatality rate in England was 31.6% from 1985 through 1990 (Moses et al., 1998). However, varying case fatality rates were subsequently reported in the 1990s. One study reported a rate of 9% (Moses et al., 1998), while another study found a case fatality of 15.6% (Mifsud et al., 2004). The case fatality in the UK for neonatal EO GBS cases was 10.6% in 2000 and 2001(Health et al., 2004). Another study documented a higher case fatality rate of 14% for early-onset GBS disease during this time period (Beardshall & Thompson & Mulla, 2000).   The global mean estimates of neonatal GBS case fatality rates were computed using a random-effects meta-analysis. This calculation was based on data from 29 papers from different geographic regions from 2000 through 2011. The overall case fatality rate of neonatal GBS  21 disease, including early- and late-onset GBS cases, was 9.6%. For EO GBS disease, the documented case fatality rate was 12.1%. The case fatality rate was 3-times higher in low-income countries (12.6%) compared to high-income countries (4.6%) (Edmond et al., 2012).   2.9 Risk Factors  There are numerous maternal, obstetric and neonatal risk factors associated with maternal and neonatal GBS colonization, vertical transmission of the pathogen and disease progression in neonates. Many of these risk factors are interrelated (Table 2.1). 2.9.1 Risk factors for maternal colonization The primary GBS reservoir in women is the gastrointestinal tract, a source of vaginal colonization. Maternal colonization is a prerequisite for the vertical transmission of GBS and subsequently EO GBS disease in neonates (Schrag et al., 2000; Sherman et al., 2012; Melin, 2011). Despite the reoccurrence of colonization in subsequent pregnancies, a significant proportion of women will not experience this phenomenon (Turrentine & Ramirez, 2008). Factors such as personal hygiene and sexual practices increase the risk of vaginal colonization (Bliss et al., 2002; Manning et al., 2004). Additional maternal colonization risk factors include ethnicity, use of tampons, obesity, and absence of lactobacilli in the gastrointestinal flora (Edwards & Nizet & Baker, 2016; Bliss et al., 2002; Le Doare & Heath, 2013; Schuchat, 1998; Zaleznik et al., 2000). Many studies have reported significant differences in colonization rates in women of black race compared to Caucasian women (Schuchat, 1998; Zaleznik et al., 2000; Turrentine & Ramirez, 2008).   22 The density of maternal vaginal colonization is directly associated with the risk of vertical transmission and the likelihood of early onset GBS (Edwards & Nizet & Baker, 2016). GBS detected in urine during pregnancy, also known as GBS bacteriuria, is reflective of the spread of the organism from the gastrointestinal or genital tract to other sites (Schuchat & Wenger, 1994). Urinary tract infections and GBS bacteriuria are considered a proxy for heavy colonization in the gastrointestinal or genital tract (Persson et al., 1986; Heath et al., 2009; Edwards & Nizet & Baker, 2016; CDC, 2010; Schuchat & Wenger, 1994). GBS bacteriuria is also associated with adverse obstetric outcomes such as habitual abortion, intrauterine growth restriction, preterm labour, chorioamnionitis and premature rupture of membranes (Schuchat & Wenger, 1994; Le Doare & Heath, 2013). The Vaginal Infections and Prematurity study reported lower levels of education as a risk factor for GBS colonization (Schuchat & Wenger, 1994); lower education level is possibly a proxy for factors such as inadequate prenatal care, race or young maternal age.  GBS can be sexually transmitted. Sexual contact was a strong predictor of GBS colonization in both men and women (Manning et al., 2004). Studies have found a significant proportion of men (45%) share the same GBS serotype as their female partners (Schuchat & Wenger, 1994; Franciosi & Knostman & Zimmerman, 1973). Moreover, colonization with GBS was more frequent in individuals who had ever engaged in sexual activity compared to those that had never engaged in sexual activity (Manning et al., 2004).  2.9.2  Risk factors for early-onset GBS disease  Risk factors that increase the risk of neonatal colonization and EO GBS disease include maternal colonization, black race and male sex of the infant (Baker & Barrett, 1973; Boyer & Gadzala et  23 al., 1983; Zaleznik et al., 2000; Schuchat et al., 1994; Schuchat et al., 2000). Prolonged rupture of membranes (≥18 hours) prior to delivery increases neonatal exposure to GBS in utero (Schuchat, 1999; Melin, 2011); it is a greater risk factor for EO GBS disease than LO GBS disease (43% vs. 4%). Prematurity (<37 weeks gestation) is also associated with an increased risk for EO GBS disease. Delivery of an infant prior to 37 weeks gestation stops maternal IgG antibody transfer and results in insufficient levels of neonatal anticapsular antibodies that are similar to the maternal colonizing strain (Schuchat et al., 1990). Using adapted data from the CDC from 2000 through 2006, Verani and Schrag (2010) reported premature infants have a 3 to 30 times greater risk of developing EO GBS disease; the lower the gestational age, the higher the risk.  Other risk factors include intrapartum fever (≥38°C), young maternal age (<20), low levels of maternal GBS-specific anticapsular antibodies, intra-amniotic infection (chorioamnionitis), inadequate prenatal care, previous infant with GBS disease (regardless of current colonization status) and low birth weight (Schuchat et al., 1994; Schuchat, 1998; Adair et al., 2003; Edwards & Nizet & Baker, 2016; Le Doare & Heath, 2013; Schrag et al., 2002; Schuchat et al., 1990). A history of a previous affected infant suggests low levels of specific GBS antibodies in the mother, which likely persists in subsequent pregnancies (Melin, 2011).   Observational studies have described an association between obstetric procedures and the risk of EO GBS disease in infants; these include the use of intra-uterine fetal monitoring devices or performing over 5 vaginal examinations during labour (Adair et al., 2003; Schuchat et al., 2000; Adams et al., 1993). Additionally, women who have previously experienced spontaneous abortion or stillbirth are twice as likely to have an infant with EO GBS disease; GBS disease and  24 spontaneous abortions share similar risk factors (Schuchat et al., 1990). There are conflicting findings about the association between caesarean section deliveries and neonatal EO GBS disease. GBS can cross intact membranes (Royston & Geoghegan, 1985); therefore, infants born via caesarean section are still at risk of vertical transmission. Some studies have reported a higher risk of neonatal GBS disease in caesarean section deliveries compared to vaginal deliveries, while other studies have reported a higher risk of neonatal GBS disease in vaginal deliveries (Hickman & Rench & Ferrieri & Baker, 1999; Edwards & Nizet & Baker, 2016). Lastly, multiple births (e.g., twins) have a higher rate of neonatal GBS disease (35%) (Edwards & Nizet & Baker, 2016). However, this risk factor may be a proxy for preterm birth, as multiples are more likely to be born prematurely.   2.9.3 Risk factors for late-onset GBS disease  Risk factors for late-onset GBS disease are not well understood. In a study conducted in the United States, infants with late-onset GBS disease were more likely to have mother with a history of urinary tract infections during pregnancy (Schuchat et al., 1990). Additional risk factors reported were black race, young maternal age (< 20 years of age), positive GBS culture and neonatal male sex (Pintye et al., 2015; Lin et al., 2003; Schuchat et al., 1990; Schuchat & Wenger, 1994). Neonates of black race were 35 times more likely to develop late-onset GBS disease compared to nonblack infants (Schuchat et al., 1990). However, ethnicity may be a proxy for SES and factors related to low SES such as overcrowded nurseries and differences in breast-feeding patterns, resulting in an increase in the likelihood of horizontal transmission (Schuchat et al., 1990). Although many studies have reported young maternal age as a risk factor for both EO and LO GBS disease, the exact mechanism remains unknown but could also be a surrogate of  25 SES. Additionally, younger women are likely to have higher rates of GBS carriage and low levels of type-specific anticapsular antibodies (Edwards & Nizet & Baker, 2016). Other explanations of age related differences include education levels and prenatal care (Schuchat et al., 1994).  Prematurity has also been reported as a major risk factor for LO GBS disease. Prior to 34 weeks gestation, there is poor placental transfer of maternal GBS-specific IgG antibodies to the neonate. Therefore, premature infants have inadequate time to acquire these antibodies against GBS maternal colonizing strains (Lin et al., 2003; Pintye et al., 2015; Edwards & Nizet & Baker, 2016).                                   26 Table 2.1 Risk factors for early- and late-onset group B streptococcal disease Early or Late-Onset  Risk Factors Early-Onset Only High density maternal colonization Maternal group B streptococcal bacteriuria  Prolonged rupture of membranes (≥18 hours) Intrapartum fever (≥38°C) Low levels of maternal group B streptococcus-specific anticapsular antibodies Caesarean section deliveries  Previous spontaneous abortion or stillbirth Inadequate prenatal care Previous infant with group B streptococcal disease Low birth weight Intra-uterine fetal monitoring devices Over 5 vaginal examinations Multiples  Early and Late-Onset Maternal colonization at delivery Prematurity (<37 weeks gestation) Young maternal age (<20) Urinary tract infections Black race Male sex    27 2.10 The Evolution of Preventive Strategies  The use of intrapartum antibiotic prophylaxis as a preventive measure against neonatal EO GBS disease was originally proposed in the 1980s. Both observational studies and clinical trials found a reduction in the vertical transmission of GBS when IAP was administered (Lim et al., 1986; Matorras et al., 1991; Easmon et al., 1983). Significant differences were observed in vertical transmission rates between the treatment group receiving ampicillin (2%) and the control group who received no treatment (35%) (Boyer et al., 1983). A study evaluating the effectiveness of a risk-based intrapartum antibiotic prophylaxis strategy reported an effectiveness of 86% (Lin et al., 2001). Clinical trials have focused on the efficacy of antimicrobial agents penicillin and ampicillin (Boyer et al., 1983; Garland & Fliegner, 1991). Penicillin is considered the drug of choice due to its narrow spectrum which reduces the risk of antibiotic resistance in other organisms. The dosage and frequency of antibiotic agents are intended to achieve adequate and rapid levels in fetal circulation and amniotic fluid (Bloom et al., 1996; Colombo et al., 2006; Fossieck & Parker, 1974).  The exact duration of prophylaxis has been debated. Other approaches have been assessed but their preventive capacities were suboptimal. The administration of oral antimicrobial agents prior to labour, and treatment of sexual partners, proved to be inefficacious, as 30-70% of women thus managed remained colonized by time of delivery (Hall et al., 1976; Gardner et al., 1979; CDC, 1996).  To facilitate the implementation of comprehensive preventive strategies among practitioners, the Society of Obstetricians and Gynaecologists of Canada (SOGC), in conjunction with the  28 Infectious Disease Committee of the Canadian Paediatric Society, developed GBS prevention guidelines (SOGC, 1994, 1997, 2004, 2013). The objectives of the SOGC recommendations were also to improve maternal and neonatal outcomes by facilitating the identification and management of pregnancies at risk for early neonatal GBS disease and attendant neonatal deaths (SOGC, 2004, 2013). Tables 2.2 and 2.3 provides a description of the evolution of the guidelines in Canada and the US.   The use of preventive guidelines differs between countries. For instance, the American College of Obstetricians and Gynecologists and the Centers for Disease Control and Prevention (CDC) recommend a universal screening approach for prevention. The approach in the United Kingdom (UK) provides for less uniformity in approach, and recommends an open dialogue between physicians and patients about risk factors and considerations; maternal screening for GBS is not recommended (Royal College of Obstetricians and Gynaecologists, 2012).          29 Table 2.2 Evolution of the Society of Obstetricians and Gynaecologists of Canada guidelines from 1994 through 2013* Year Description 1994 One of two recommended approaches:  (1) Universal screening of all pregnant women at 26 to 28 weeks gestation and administration of IAP based on maternal colonization and presence of risk factors or, (2) No universal screening but IAP administration to all women with the presence of risk factors  Antibiotic agents: Ampicillin, Penicillin or Clindamycin  1997 One of two recommended approaches: (1) Universal screening of all pregnant women at 35 to 37 weeks gestation and administration of IAP based on maternal colonization or,  (2) No universal screening but IAP administration to all women with the presence of risk factors Antibiotic agents: Ampicillin, Penicillin or Clindamycin  Change from 1994: screening period later in pregnancy; IAP based on colonization without requirement for additional risk factors 2004 Exclusively a universal screening approach: (1) Universal screening of all pregnant women at 35 to 37 weeks gestation and administration of IAP based on colonization, GBS bacteriuria in the current pregnancy and/or previous neonate with GBS disease and, (2) Administration of IAP in the presence of unknown GBS status and risk factors  Antibiotic agents: Penicillin, Cefazolin, Clindamycin, Erythromycin or Vancomycin   Change from 1997: Risk factor approach is no longer an acceptable alternative; addition of erythromycin, cefazolin and vancomycin as possible alternative agents; ampicillin removed from possible agents 2013 Exclusively a universal screening approach: (1) Universal screening of all pregnant women at 35 to 37 weeks gestation and administration of IAP based on colonization, GBS bacteriuria in the current pregnancy and/or previous neonate with GBS disease and,  (2) Administration of IAP in the presence of unknown GBS status and risk factors  Antibiotic agents: Penicillin, Cefazolin, Clindamycin or Vancomycin  Changes from 2004: Erythromycin is removed from possible antibiotic regimens; women with GBS bacteriuria in the current pregnancy should not be screened in the 3rd trimester as they are presumed to be colonized and will receive IAP. Abbreviations: GBS= group b streptococcus and IAP= intrapartum antibiotic prophylaxis                                                   * This information references the SOGC clinical practice guidelines (SOGC 1994, 1997, 2004, 2012, 2013)  30 Table 2.3 Evolution of the Centers for Disease Control and Prevention guidelines from 1996 through 2010† Year Description 1996 One of two recommended approaches: (1) Screening approach: universal screening of all pregnant women at 35 to 37 weeks gestation and administration of IAP based on maternal colonization or,  (2) Risk factor approach: IAP administration to all women with the presence of risk factors (i.e. preterm birth, prolonged rupture of membranes, intrapartum fever, GBS bacteriuria during current pregnancy or previous infant with GBS disease)  Antibiotic agents: Penicillin, Ampicillin, Clindamycin or Erythromycin 2002 Exclusively a universal screening approach: (1) Universal screening of all pregnant women at 35 to 37 weeks gestation and administration of IAP based on colonization, GBS bacteriuria in the current pregnancy and/or previous neonate with GBS disease and, (2) Administration of IAP in the presence of unknown GBS status and risk factors  Antibiotic agents: Penicillin, Ampicillin, Cefazolin, Clindamycin, Erythromycin or Vancomycin   Change from 1996: Risk factor approach is no longer an acceptable alternative; updated prophylaxis regimens.   2010 Exclusively a universal screening approach: (1) Universal screening of all pregnant women at 35 to 37 weeks gestation and administration of IAP based on colonization, GBS bacteriuria in the current pregnancy and/or previous neonate with GBS disease and, (2) Administration of IAP in the presence of unknown GBS status and risk factors  Antibiotic agents: Penicillin, Ampicillin, Cefazolin, Clindamycin or Vancomycin  Changes from 2002: updated dosing of prophylactic penicillin G, a change in the alternative prophylactic regimen among women with penicillin allergies Abbreviations: GBS= group b streptococcus and IAP= intrapartum antibiotic prophylaxis  2.10.1 1994 Canadian guidelines Canadian prevention guidelines were initially created in 1994 and recommended one of two preventive approaches: routine screening and IAP for screen positive women with risk factors, or a risk factor based approach.  In the former method, physicians screened all pregnant women at                                                 † This information references the CDC clinical practice guidelines (CDC 1996, 2002, 2010)  31 26 to 28 weeks gestation, and all identified GBS carriers with risk factors were offered intrapartum antibiotic prophylaxis. Pregnant women with GBS colonization without risk factors were not recommended prophylaxis. These risk factors included preterm birth (<37 weeks), prolonged rupture of membranes (≥18 hours), intrapartum fever (≥38.0 C), previous newborn with GBS disease or previously documented GBS bacteriuria. To maximize the probability of GBS recovery, culture techniques involved using a single swab first in the vagina then the rectum, followed by inoculation into selective broth medium. The alternative risk-based approach recommended administration of IAP to women based on the presence of risk factors alone (SOGC, 1994). The recommended antibiotic regimen consisted of intravenous ampicillin or penicillin G until delivery. Possible alternatives for women with allergies to these agents included clindamycin. Moreover, neonatal management was based on the infant’s gestational age, adequacy of maternal chemoprophylaxis and results of investigations for neonatal sepsis.    2.10.2 1997 Canadian guidelines  Following the creation of prevention guidelines by the US CDC (Table 2.3), the SOGC guidelines were revised in 1997 to align with the US recommendations. Specifically, screening of pregnant women was recommended at 35 to 37 weeks gestation, and the provision of IAP to all women with positive cultures regardless of the presence of risk factors. The alternative risk-based approach, neonatal management and antibiotic regimens remained the same (SOGC, 1997).       32 2.10.3 Effectiveness of the guidelines and comparison studies Following the creation of the 1996 CDC and 1997 SOGC GBS prevention guidelines, changes in incidence rates were observed. Specifically, the incidence rate of EO GBS disease declined by 70%, and an estimated 4500 cases and 225 deaths were prevented annually in the US (CDC, 2002). However, it was noted that IAP was underutilized in preterm deliveries (Schrag et al., 2002).  The effectiveness of the screening and risk-based approaches were compared. Schrag and colleagues (2002) conducted a large retrospective population-based cohort study and concluded that the screening approach to prevention was superior to the risk-based approach as it was 54% more effective in preventing neonatal GBS disease. In another study conducted in California, a risk-based approach resulted in a higher incidence of 1.1 per 1000 births while a screening approach resulted in no neonatal GBS cases (0 per 1000 births) (Main & Slagle, 2000). The screening approach had many advantages, which included facilitation of communication and its capacity to detect GBS carriers that do not present with risk factors. Therefore, this approach covered a higher proportion of the at-risk population compared to the risk-based approach (Schrag et al., 2002; Davis et al., 2001). Moreover, routine screening through GBS culture was advantageous, as screening had clear indicators such as documentation of GBS testing, timing of screening and rate of GBS positivity, which facilitated evaluation of implementation (Davis et al., 2001; CDC, 2002). A cost effectiveness analysis using population based US data from 10 studies was conducted to assess the two approaches. Universal maternal screening and the use of prophylaxis of colonized women was more economically favourable and resulted in cost savings when direct costs were taken into account (Mohle-Boatani et al., 1993).    33 The timing of screening is an essential component of the prevention guidelines due to the transient nature of GBS colonization (Schrag & Verani, 2013). To maximize the likelihood of accurate detection, screening culture is recommended near the end of the 3rd trimester, at 35-37 weeks gestation (SOGC 2004, 2013). Boyer et al. (1983) reported a 100% concordance of culture results at the time of delivery in women who undergo screening less than 5 weeks before delivery (Boyer & Gadzala & Kelly et al., 1983). However, the sensitivity of screening tests has been debated and can range from 54% to 87% (Yancey et al., 1996; Money et al., 2008; Davies et al., 2004; Young et al., 2011). Screening completed during the suggested period yields a 95-97% negative predictive value (Verani & Schrag, 2010).  The site of specimen collection is also an important component in screening. Specimens collected from vaginal and rectal swabs result in a 5-27% higher likelihood of GBS detection compared to specimens collected from the cervix (Boyer & Gadzala & Kelly et al., 1983; Regan et al., 1991; Dillon et al., 1982). The type of media in which specimens are collected also plays a vital role in accuracy of GBS detection. Using selective broth media such as LIM or SBM broth, increases the yield of screening culture by as much as 50% (Baker et al., 1977; Ferrieri & Blair, 1977); selective broth media contain antimicrobial agents that inhibit competing organisms and avoid their overgrowth (CDC 1996, 2002).        34 2.10.4 2004 Canadian guidelines  The SOGC used the emergence of new data, predominantly derived from US studies, to inform the newest revisions of the guidelines. The general conclusion of these studies was the superiority of the screening based approach. These updated guidelines recommended a universal prenatal GBS screening approach. Other recommendations included the administration of IAP to women who experience preterm labour or rupture of membranes (<37 weeks) when culture results are unknown, and in the presence of prolonged rupture of membranes (PROM) and unknown GBS culture (SOGC, 2004). Moreover, women with GBS bacteriuria in the current pregnancy, regardless of colony-forming units per mL, were candidates for IAP administration in this new revision. Further, the use of broad-spectrum antibiotic to treat maternal fever and chorioamnionitis was recommended. Other changes included an update to the recommended antibiotic regimens for IAP indications. Firstly, penicillin was now the choice agent instead of ampicillin, with ampicillin as an acceptable alternative. Moreover, cefazolin was recommended as the agent of choice among penicillin allergic women who were not at risk for anaphylaxis; for those at risk of anaphylaxis either clindamycin or erythromycin were recommended, and alternatively vancomycin if GBS stains were resistant to clindamycin and erythromycin. There were no changes to recommendations for neonatal management (SOGC, 2004).   2.10.5 2013 Canadian guidelines  In the late 2000s, the SOGC reviewed emerging data. A particular area of interest was the implementation of the guidelines following their update. The implementation of the 2002 universal screening guidelines was widely and rapidly put into effect in the US. Van Dyke et al. (2009) conducted a population-based study to evaluate the implementation of the 2002 US  35 guidelines. Prenatal screening increased to 85% from 2003 through 2004, compared to 48.1% in 1998-1999. In addition, the administration of IAP among women with indications increased from 73.8% to 85.1% during this time period. However, maternal screening and prophylaxis administration were suboptimal among preterm births (Van Dyke et al., 2009).   Due to the available data, the guidelines underwent further iterations to replace the existing 2004 recommendations. These included an update in dosing of prophylactic penicillin G, a change in the alternative prophylactic regimen among women with penicillin allergies, and change to the management of newborns with risks of EO GBS disease (SOGC, 2013). Lastly, additional SOGC guidelines were created for the management of GBS bacteriuria, in that women with GBS bacteriuria at any time in the current pregnancy did not require GBS screening in the 3rd trimester, as they were presumed to be colonized at delivery and recommended to receive IAP (SOGC, 2012). Screening for bacteriuria is recommended during the antenatal period as part of the standard component of obstetrical care.   2.11 Failures and Limitations of Recommendation Guidelines   2.11.1 Screening    The first inherent limitation of antenatal screening occurs with preterm deliveries. Mothers who deliver before 35 weeks gestation are less likely to have been screened. Van Dyke and colleagues (2009) found that delivery before 34 weeks gestation is a significant risk factor associated with lack of screening. Studies have observed up to 42% of women with preterm deliveries are unscreened. (Stoll et al., 2011; Van Dyke et al., 2009).   36 Screening completed during the suggested time period (35-37 weeks gestation) yielded a 95-97% negative predictive value (Verani & Schrag, 2010); sensitivity of GBS screening ranges from 54%-87% (Yancey et al., 1996; Money et al., 2008; Davies et al., 2004; Young et al., 2011). Therefore, even in the occurrence of optimal screening adherence, false negative results are expected to occur due to the limitations of the screening test. A population based study found that up to 81% of EO GBS disease cases occur in infants of mothers with negative screening culture (Stoll et al., 2011). Other studies have reported 61% to 82% of neonates with GBS disease are born to women with a negative screening culture (Pulver et al., 2009; Puopolo & Madoff & Eichenwald, 2005; Van Dyke et al., 2009). Probable causes include late colonization and false-negative test results (Ahmaszia & Heine, 2014; Verani & Schrag, 2010).  2.11.2 Intrapartum prophylaxis  Maternal prophylaxis is not expected to prevent all neonatal EO GBS cases. The administration of IAP is dependent on factors such as GBS indicators and gestational age. Nonetheless, gaps in the adherence to prophylaxis guidelines as well as antibiotic failures reduce the opportunities for optimal prevention of EO GBS cases.   Prophylaxis failure resulting in neonatal EO GBS disease has been documented in 23%-25% of cases (Schuchat et al., 2000; Pulver et al., 2009). According to the literature, an estimated 1425 EO GBS cases occurred in the US despite the use of IAP (Jordan et al., 2008; Phares et al., 2008). Factors influencing the effectiveness of IAP include the presence of intrapartum fever, administration of prophylaxis less than 2 hours before delivery and antibiotic resistance (Lin et  37 al., 2001; Schrag & Verani, 2013). The presence of intrapartum fever suggests that infection has been established resulting in a diminished effect of maternal chemoprophylaxis (Lin et al., 2001).   In addition to inherent limitations of IAP, failure to treat pregnant women with GBS risk factors results in 24%-34% of EO GBS cases (Stoll et al., 2011). Factors associated with precipitous deliveries result in an inadequate interval of time for maternal IAP. Other studies have reported vaginal delivery, unknown GBS colonization status due to inconclusive culture results or lack of screening, and preterm birth as factors associated with absence of prophylaxis (Goins et al., 2010; Van Dyke et al., 2009; Schrag & Verani, 2013). Specifically, 63.4% of women with preterm birth and unknown GBS status received IAP compared to 84.5% of women with GBS colonization (Van Dyke et al., 2009).    Deviation from the recommendations also limits the impact of prevention guidelines; this has been observed in up to 87.5% of neonatal EO GBS cases. Observed deviations in prophylaxis administration have included the use of alternative antibiotic agents, administration of the incorrect antibiotic agent, and insufficient dosage (Berardi et al., 2011; Pulver et al., 2009; Goins et al., 2010; Bienenfeld & Rodriguez-Riesco & Heyborne, 2016; Schuchat et al., 2000). Inappropriate antibiotic administration is associated with lack of susceptibility testing of GBS cultures, as well as penicillin allergies and preterm deliveries (Bienenfeld & Rodriguez-Riesco & Heyborne, 2016; Goins et al., 2010). Ultimately, prevention measures and adherence can be strengthened to further reduce the burden of illness.    38 2.12 Alternatives or Supplementary Tools to Current Guidelines 2.12.1 Rapid screening methods  Alternative rapid screening methods have been explored, which include DNA hybridization and polymerase chain reaction tests. The purpose of these tests is to yield rapid colonization results and are particularly beneficial for those that do not receive prenatal care, have an unknown colonization status at delivery, or experience preterm birth and rupture of membranes (Young et al., 2011; Money et al., 2008). Health Canada and the Food and Drug Administration have approved in-labour testing for GBS using rapid real-time Polymerase Chain Reaction (PCR) (Alfa et al., 2010). The reliability of this method was assessed in a Canadian setting. The sensitivity, specificity, positive predictive value and negative predictive value were reported at 90.5%, 86.1%, 86.4%, and 97.4% respectively. Results demonstrated discordance in a minority of women (n=10) that previously received screening at 35-37 weeks gestation (Alfa et al., 2010).   PCR testing would be of great value in the setting of prematurity, which is associated with 38% of neonatal GBS cases (Money et al., 2008). PCR testing would also be beneficial for women with little to no prenatal care. However, this approach to screening has limitations that include its inability to assess antibiotic susceptibility of particular GBS strains. Furthermore, PCR testing is not readily available in all medical centers and requires specific laboratory technologist expertise and instrumentation, which can be costly and not feasible in small care centers (Rallu et al., 2006; Alfa et al., 2010; Money et al., 2008; CDC, 2002).    39 2.12.2 Vaccines Although the current standard of care of screening and IAP have significantly decreased the burden of neonatal GBS disease, it remains ineffective for late-onset GBS disease and cases occurring in preterm infants. Other prevention strategies are needed to reduce the incidence of neonatal GBS disease. Maternal immunization has been examined as a potential intervention for reducing the colonization and transmission of GBS by providing passive immunity to the infant (Chen & Avci & Kasper, 2013; Baker, 2013). Rationale for a GBS vaccine The premise of maternal immunization emerged when the association between low levels of maternal antibodies to capsular polysaccharides (CPS) antigens and neonatal GBS disease was discovered (Edwards, 2008; Chen & Avci & Kasper, 2013). This discovery emerged in the 1970s by Baker and Kasper, who observed an absence of antibodies to type III CPS in both mothers and infants with EO or LO GBS disease. In contrast, healthy infants and their respective mothers had the presence of such antibodies (Chen & Avci & Kasper, 2013; Baker et al., 2014). Active immunization of expecting mothers could elicit protective levels of IgG in the mother, which would be passed through placental transfer to the infant (Edwards, 2008; Chen & Avci & Kasper, 2013). Moreover, after 13 years of declines following the implementation of prevention guidelines, the incidence rate of neonatal GBS disease has reached a plateau (CDC, 2010). Therefore, additional preventive measures are required to reduce the remaining burden of illness. The most predominant GBS serotypes include type Ia, Ib, II, III and V and a vaccine capable of stimulating antibodies against these serotypes would prevent 85%-96% of all GBS disease cases (James, 2001; Le Doare & Heath, 2013; Heath, 2011).   40 Clinical trials  There have been numerous clinical trials that have evaluated the safety and immunogenicity of different GBS vaccine designs among both non-pregnant adults and pregnant women (Baker & Edwards & Kasper, 1978; Kasper et al., 1996; Leroux-Roels et al., 2016; Baker & Rench & McInnes, 2003). These clinical trials were in phase I and measured GBS serotype-specific antibody responses. Results from a clinical trial using monovalent vaccines demonstrated the presence of antibodies against specific GBS serotype in both mothers and infants receiving the vaccine, with protection in infants lasting up to 2 months of age (Baker & Edwards, 2003). An additional clinical trial assessed the impact of GBS vaccination on colonization. Results demonstrated a decrease in vaginal and rectal colonization in immunized women. Specifically, 36% and 43% vaccine efficacy was reported in vaginal and rectal acquisition respectively (Baker, 2009). Clinical trials have also assessed bivalent vaccines such as an II/III-Tetanus Toxoid (TT) conjugate vaccine, which was administered to healthy adults. Up to 90% of subjects had a 4-fold or more increase in GBS type II and III CPS-specific antibodies (Chen & Avci & Kasper, 2013).  This vaccine was well tolerated and the immune response stimulated was similar to those from the monovalent conjugate vaccines (Baker & Edwards, 2003).   Vaccine developers have focused their attention on vaccines combating serotype III, due to its prevalence in neonatal disease, however, over time, other serotypes such as type Ia and V have become more common in neonatal disease (Phares et al., 2008). The most recently conducted human phase II clinical trial assessed trivalent GBS vaccine safety and immunogenicity in Belgium and Canada. The trivalent vaccine contained glycoconjugate of serotypes Ia, Ib and III, and was administered to healthy pregnant women aged 18 to 40 years old at 24-35 weeks of  41 gestation. Transfer ratios between mother and infant ranged from 66%-79% and maternal geometric mean concentration increased by 16-fold for serotype Ia, 23-fold for serotype Ib, and 20-fold for serotype III. At 3 months of age, infant antibody concentrations had declined by approximately 25% from birth levels. Consequently, GBS-specific neonatal antibodies in the treatment group were still 8.5-fold higher than antibody levels found in infants from the placebo group (Donders et al., 2016). Moreover, maternal antibodies continued to rise until at least 3 months postpartum, indicating the importance of timing of immunization. Researchers also noted that transfer ratios observed were similar to transfer ratios found in other maternal polysaccharide vaccines (i.e. meningococcal vaccines) (Donders et al., 2016). Nonetheless, given the novelty of GBS vaccination, little is known about the duration of the immune response (Baker et al., 1999; Baker et al., 2000).   Challenges  Many studies have concluded that infant susceptibility to GBS disease is due to a lack of maternal anticapsular antibodies. Maternal vaccination would be able to provide immunity to infants through transplacental transfer of protective IgG antibodies (CDC, 1996). The current design for the vaccine is based on the association between maternal levels of GBS-specific anticapsular antibodies and neonatal GBS disease. Antibodies to these capsular polysaccharide antigens are the principal determinant of immunity to GBS in young infants (Davies et al., 2001c). If used in conjunction with screening and IAP, maternal immunization would be a beneficial strategy in the prevention of both early- and late-onset GBS disease and adverse neonatal outcomes. Furthermore, a maternal vaccine could prevent cases attributed to false  42 negative laboratory results, late colonization and preterm births. However, there are challenges in potential program implementation during pregnancy because of the longstanding avoidance of drugs and vaccines in pregnancy because of theoretical safety concerns; proposed solutions are to target populations such as adolescent girls for vaccine administration (Schuchat, 1998). The duration of protection afforded by vaccination is unknown; one or more booster doses might be required for protection throughout the child-bearing years. Shifts in the GBS serotypes causing neonatal disease have provided an additional challenge to vaccine development (Schuchat, 1999). More research needs to be conducted to inform decision-makers on the need for and best approaches to a GBS vaccine.  2.13 Neonatal GBS Case Definition and Reporting in Canada In Canada, neonatal GBS has been a nationally reportable disease since 2006 (Public Health Agency of Canada, 2009). The national case definition includes cases from birth until 1 month of age. The BC Centre for Disease Control conducted a review of passively reported cases from the Public Health Information System against aggregated data from the Perinatal Data Registry. From 2001 to 2011, 78 neonatal GBS cases were reported through the Public Health Information System by regional health authorities compared to 192 neonatal GBS cases identified in the Perinatal Data Registry. This review revealed under-reporting of cases through the passive notification system (BC Centre for Disease Control, unpublished data, 2012).    2.14 Knowledge Gaps  Prior to the implementation of IAP, there were 7500 annual EO GBS cases in the United States (CDC, 2010). However, following the implementation of preventive guidelines and advances in  43 neonatal care, the case-fatality rate in the US and Canada among all neonatal GBS cases declined from 50% to 5-9% in recent years, including an 80% reduction in the incidence of invasive EO GBS disease (Koening & Keenan, 2009; Health & Schuchat, 2007; Le Doare & Heath; 2013; Jordan et al., 2008; Schuchat, 1999; CDC 2010). In Canada, the incidence of EO GBS disease decreased from 1.25 to 0.15 per 1000 live births (Davies et al., 2001a; Alhhazmi & Hurteau & Tyrrell, 2016). Although progress has been made in the prevention of early-onset GBS, there are still many challenges that persist. In recent years, the incidence of EO and LO GBS in Canada has increased (Alhhazmi & Hurteau & Tyrrell, 2016). GBS also remains the leading infectious cause of neonatal morbidity and mortality (Weston et al., 2011). Although the SOGC guidelines are not expected to prevent all neonatal EO GBS cases, the inherent limitation of the guidelines in addition to failure of guideline implementation and/or antibiotic prophylaxis failure lead to a continued risk of GBS transmission and subsequent neonatal EO GBS disease (CDC, 2010). Therefore, continuous efforts are required to improve the prevention of GBS disease.   In Canada, little research has been conducted in areas related to neonatal GBS disease and its prevention. Only a single study related to neonatal GBS disease in British Columbia could be identified in the literature, examining maternal colonization (Money et al., 2008). No studies examining the incidence of neonatal GBS disease in BC were found, nor Canadian studies evaluating the application and effectiveness of the SOGC guidelines.     44 2.15 Project Objectives   The purpose of this study is to describe the epidemiology of neonatal GBS disease in British Columbia, evaluate the risk factors associated with neonatal GBS disease, and identify whether the remaining burden of neonatal GBS could be further reduced through improved implementation of the SOGC guidelines. The objectives of this study are to:  1. Estimate the incidence and case fatality of neonatal group B streptococcal disease in BC from 2005 through 2014.  1.1. Examine the association between maternal and neonatal risk factors, and neonatal GBS disease outcome.  2. Describe GBS screening frequency among women giving birth in BC.  3. Describe the prevalence of indications for IAP among women giving birth in BC who are at-risk for having an infant with GBS.  4. Describe the frequency of maternal antibiotic‡ receipt during labour and delivery among at-risk women giving birth in BC.  5. Investigate factors associated with absence of antibiotic administration during labour and delivery among at-risk women giving birth in BC. 6.  Investigate the characteristics of women with false negative screening results.  7. Examine the application of the SOGC guidelines and identify gaps and limitations of these recommendations.   This study will provide an understanding of the remaining challenges in the prevention of neonatal GBS disease in British Columbia. The findings from this study will inform changes to                                                 ‡ Maternal antibiotic receipt was used as a proxy to infer IAP administration   45 current guidelines and strengthen prevention efforts, with the overarching goal to help inform decision-makers about the potential need of supplementary prevention approaches and improve clinical care. This study will help further reduce the burden of neonatal GBS disease and contribute to the Canadian literature.                      46 Chapter 3: Methods  To address the project objectives, this study’s methodology was separated into two components. Firstly, a retrospective population-based cohort study design was employed using individual level data on neonates (0-28 days) and women giving birth from the British Columbia Perinatal Data Registry (PDR). Electronic data were retrieved from April 1st 2004 through December 31st 2014 based on infant’s date of birth. Quantitative methods were used and include Poisson regression and multiple logistic regression analyses. Poisson regression was employed to estimate the incidence of neonatal GBS disease in BC, and a set of multiple logistic regression analyses were applied to address project objectives 1.1, 5 and 6. Descriptive statistics were conducted to answer project objectives 3 through 5. The second study component used a case series design. For a subset of neonatal GBS infant case-mother pairs, a retrospective chart review was conducted at the BC Women’s and Children’s Hospital. Additional variables that were unavailable from the PDR were retrieved from hospital records. This study design and data source was used to examine the application of the SOGC guidelines and identify gaps and limitations of these recommendations (objective 7). All statistical analyses were conducted using R studio, version 3.3.1 via the BC Centre for Disease Control’s Central Analytics Platform.   3.1  Data Sources  Individual level electronic data were obtained from the PDR, operated by Perinatal Services BC (PSBC). This database contains information on all births in the province with the exception of unassisted homebirths and unreported births (Perinatal Data Registry, 2017). A chart review was carried out on a subset of GBS infant case-mother pairs whose records for birth, readmission and/or transfer for care were at the BC Women’s and Children’s Hospital. The chart review was  47 required to retrieve additional variables that were not available from the PDR to assess guideline adherence and distinguish early- versus late-onset GBS cases. The BC Women’s and Children’s Hospital is the only tertiary care pediatric hospital in the province. It was selected due to the high volume of deliveries, resulting in the highest number of GBS cases at one hospital.   3.1.1 Perinatal Data Registry The PDR, a comprehensive perinatal registry since April 1st 2000, has accumulated approximately seven hundred thousand full provincial birth and delivery records. Data collected by this registry stem from obstetrical and neonatal medical records from 60 hospitals. Information is also collected from home births attended by BC registered midwives. Participating hospital and health authority employees periodically submit standardized electronic data to the provincial database; reporting in the registry is not mandatory (Perinatal Services BC, 2014).  Data is collected on approximately 100% of births in the province (Perinatal Data Registry, 2017).  As per PSBC, the PDR contains information on both parturients and neonates, for all live births and stillbirths occurring after 20 weeks of gestation (Perinatal Data Registry, 2017). This includes newborn records of care, readmissions, transfers (on or before 28 days of age) and laboratory data. Diagnostic codes for Group B streptococcal disease are recorded as per the International Classification of Disease (ICD-10-CA), implemented as of April 1st 2004. Prior to 2008, the PDR exclusively captured these diagnostic codes for the identification of neonates with GBS disease. Culture results from blood, urine and/or other cultures positively coded for GBS during the neonatal period were introduced on April 1st 2008. GBS cases occurring after 28 days  48 could not be included in our study, as the data from the PDR is only collected for neonates up to 28 days of life.  Data that were obtained for our study included records of all parturients with live births, and their liveborn neonate(s) from April 1st 2004 through December 31st 2014. Variables retrieved from the PDR included maternal and neonatal demographics, delivery episode of care information, past obstetric history, current pregnancy, labour and delivery, neonatal birth, admission and/or transfer information as well as diagnoses and procedures. An important risk factor, GBS bacteriuria in the antenatal period of the current pregnancy, is not collected by the PDR and could not be assessed. Maternal screening (vaginal/ rectal) results for GBS were retrieved, however, the PDR does not contain information about date or gestational week of screening. Variables that were explored in univariate and multiple regression analyses for project objectives 1.1, 5 and 6 are described in Table 3.1. All inferences, opinions, and conclusions drawn in this publication are those of the authors, and do not reflect the opinions or policies of Perinatal Services BC.  3.1.2 BC Women’s and Children’s Hospital   To assess the limitations of the guidelines and missed opportunities for prevention associated with neonatal early-onset GBS occurrences, a retrospective chart review was carried out at the BC Women’s and Children’s Hospital, located in Vancouver, British Columbia. This tertiary care hospital serves as the busiest and largest maternity care centre in Canada (Our Story, 2017). Every year, an estimated 7000 infants are delivered at the BC Women’s Hospital, accounting for  49 20% of deliveries in BC. Moreover, this hospital is considered a provincial referral centre for high-risk pregnancies (Our Unique Role, 2017).   Retrospective chart review  A subset of GBS infant case-mother pairs were selected for a full chart review based on their place of birth, transfer or readmission to the BC Women’s and Children’s Hospital from April 1st 2004 through December 31st 2014. A full chart review involved a review of both maternal and neonatal hospital charts. A standardized electronic abstraction form was used for data retrieval to ensure consistency and reduce error in data collection (Appendix A). A procedural manual detailing decision logic was also created and used during data abstraction. The review of hospital charts was completed on-site at the BC Women’s and Children’s Hospital. Variables that could not be retrieved from the PDR were extracted from medical records. These include details of maternal antibiotic administration for GBS and dates associated with: maternal GBS screening, onset of neonatal GBS, collection of specimens positive for GBS and laboratory results. Researcher-collected data (i.e. hospital extracted data) were linked to maternal and neonatal records in the perinatal data set using linkage variables such as study ID numbers.  50 Table 3.1 Perinatal Data Registry variables explored in univariate and multiple regression analyses Level Variable Description Types Maternal Health Authority† Residential health authority of the mother categorized as Fraser Health, Interior Health, Northern Health, Vancouver Coastal, and Vancouver Island Health Categorical Calendar year† Calendar year from 2004 through 2014  Categorical Body Mass Index† Pre-pregnancy body mass index of the mother categorized as underweight, normal, overweight, and obese Categorical GBS testing† Whether the mother was screened for GBS categorized as yes, no and unknown. GBS screening refers to testing performed during the antenatal period including at time of hospital admission for delivery.  Categorical GBS culture results† GBS culture result categorized as positive, negative and unknown   Categorical Drugs received during delivery admission-Antibiotics† Whether the mother received antibiotics during labour and delivery categorized as yes and no/null    Categorical Fetal surveillance during labour† Indicates the type of fetal monitoring used during labour   Categorical Midwife use†  Whether a midwife was involved in the care of the mother or neonate   Categorical Labour type†  Mode of delivery† Indicates how labour began categorized spontaneous, induced, none and unknown   Method of newborn delivery, categorized as vaginal or caesarean section Categorical  Categorical  Mode of delivery (detailed) †   Detailed classification of delivery mode categorized as emergency primary, emergency repeat, elective primary, elective repeat, forceps and vacuum, forceps, vacuum, other instrument, spontaneous  Categorical  HIV testing† Whether the mother was screened for HIV categorized as yes, no and unknown Categorical Length of time from rupture of membranes to delivery of 1st baby Indicates the time from rupture of membranes to the delivery of the first infant. The time is expressed in hours. This variable was categorized as prolonged (≥ 18 hours), regular (<18 hours), no rupture and unknown  Categorical Young maternal age† Mother’s age (in years) at date of infant delivery categorized as <20 or ³20 Categorical   51 Level Variable Description Types Maternal Maternal age† Mother’s age (in years) at date of infant delivery Continuous Total length of stay  Indicates length of stay of entire admission expressed in hours  Continuous Antepartum length of stay  Indicates the time from admission to the delivery of the first infant. The time is expressed in hours Continuous Previous term infant  Indicates the total number of previous pregnancies delivered at ≥ 37 weeks gestation  Continuous Previous premature infant Indicates the total number of previous pregnancies delivered at 20 to 36 weeks gestation  Continuous Previous spontaneous abortion Indicates the total number of previous pregnancies that resulted in natural losses. This includes pregnancy loss <20 weeks gestation and <500 grams.   Continuous  Provider type†  Indicates the health care provider who delivered the infant. This variable is categorized as Family physician; Obstetrician or Fellow; Midwife; Nurse; Medical student intern; Obstetrical resident; Midwife trainee; Family practice resident; Other; No Attendant; Surgeon; Unknown Categorical  Woman person ID† Identifies and links all records that belong to the same woman  Categorical Neonatal  Sex† Neonate’s biological sex at birth categorized as female or male   Categorical Term status Classification of neonate’s term status at birth, categorized as preterm (<37 weeks) and full term (≥ 37 weeks). Categories were derived from the variable gestational age   Categorical Term status weeks (detailed) Classification of neonate’s term status at birth, categorized as extremely preterm (<32 weeks), moderately preterm (32-36 weeks) and full term (37-45 weeks). Categories were derived from the variable gestational age   Categorical Gestational age Neonate’s gestational age (in weeks) at birth Continuous Sequence†  Indicates the sequence of neonatal deliveries   Continuous Number of infants delivered† Indicates the total number of babies delivered from the current pregnancy  Continuous Admission weight Neonate’s weight (in grams) at birth  Continuous †These variables did not require imputation as they did not contain missing values.  Abbreviations: GBS= Group B Streptococcus, HIV=Human Immunodeficiency Virus 52 3.2 Study Population  3.2.1 Setting  The setting for this study was the Province of British Columbia, and its 5 regional and 1 provincial health authorities: Vancouver Island (VIHA), Vancouver Coastal (VCH), Fraser (FHA), Interior (IHA), Northern (NHA) and the Provincial Health Authority. The retrospective chart review was exclusively carried out on cases from the BC Women's and Children’s Hospital, in Vancouver, BC, part of the Provincial Health Services Authority.  3.2.2 Inclusion and exclusion criteria The study population includes live births from singleton and twin pregnancies, including twin pregnancies in which one of the twins was stillborn, from April 1st 2004 through December 31st 2014, and their respective mothers. In the event of one twin stillborn, the data related to the stillborn infant were excluded. Our study population includes home births and hospital deliveries attended by a BC registered midwife and/or physician. Data on neonates that required transfers and/or readmissions during this time period were also included.  Gestations that resulted in multiples above two (i.e. triplets, quadruplets and sextuplets) were excluded. Records associated with out of country/province residence, or unknown residence status as well as records with a neonatal sex category of “unknown” or “other” were also excluded. Lastly, records with ambiguous categorical values were excluded; these involved records with blank values in place of pre-existing categories that account for undocumented data such as “missing” or “unknown” categories (Appendix B). Data for the 2004 calendar year were excluded from analyses of the first project objective.   53 3.3 Ethics This research project required human research ethics approval, granted by the University of British Columbia Children’s and Women’s Research Ethics Board (CW15-0288 / H15-01308). Institutional approval was also obtained prior to the commencement of the hospital chart review process. Investigators had limited access to identifiable information, and as such only received access to hospital chart numbers and neonatal date of birth.  3.4 Burden of Illness (Objective 1)  The first objective of this study was to estimate the incidence and case fatality rate of GBS disease in British Columbia among neonates born from January 1st 2005 through December 31st 2014. This retrospective cohort study used population-based data from the PDR and employed Poisson regression analysis. Descriptive statistics that include demographic, obstetric and outcome characteristics of the study population were obtained and reported according to cases (GBS infant-mother pairs) and controls (non-GBS infant-mother pairs).   3.4.1 Neonatal GBS disease outcome   The outcome for this analysis was neonatal GBS disease, a binary variable. It was categorized as neonatal GBS cases and neonatal non-GBS controls. Neonatal GBS disease was identified based on diagnostic codes provided by the PDR. Our study used both culture results (blood or other) and diagnostic codes for the identification of neonatal GBS disease. A case was defined as a neonate having an ICD-10-CA code of one or more of: A40.1, B95.1, P36.0, G00.2, J15.3, P23.3 (Table 3.2) and/or a positive infectious agent culture (B95.1) in blood or other specimen. Other culture is defined as the presence of GBS (B95.1) in laboratory test results other than blood and  54 urine (e.g. stool). This case definition included neonates from 0 to 28 days of life, encompassing early- and some late-onset neonatal GBS disease. All cases in which neonates had GBS disease based on PDR variables related to culture results and/or diagnostic codes were included. GBS records from the PDR that were miscoded according to our retrospective chart review (n=8) were removed. For neonates with multiple transfers and discordant GBS diagnostic codes (n=9), re-admissions and/or transfers were arranged sequentially using variables outlining the institution to which the neonate was admitted for care as well as the institution from which the neonate was transferred. These cases were tabulated as 1 episode (i.e. 1 case). Controls were remaining neonates in the cohort of live births who did not meet the study case definition for neonatal GBS.       Table 3.2 Case definition of neonatal group B streptococcal disease: infant with one or more of the following ICD-10-CA codes§  Codes Details A40.1 Sepsis due to streptococcus, group B Excludes: neonatal (P36.0-P36.1) B95.1 Streptococcus, group B, as the cause of diseases classified to other chapters P36.0 Sepsis of newborn due to streptococcus, group B Includes: infections acquired in utero or during birth G00.2 Streptococcal meningitis Includes: Non-pneumococcal streptococci (Streptococcus, Group A) (Streptococcus, Group B) J15.3 Pneumonia due to Streptococcus, group B P23.3 Congenital pneumonia due to streptococcus, group B Includes: infective pneumonia acquired in utero or during birth Excludes: neonatal pneumonia resulting from aspiration (P24. -)                                                 § This information is taken from the Canadian edition of the International Statistical Classification of Diseases and Related Health Problems (Canadian Institute for Health Information. (2012).   55 3.4.2 Statistical analysis  A Poisson regression was used to estimate the incidence of neonatal GBS disease in BC. Incidence rates were examined by health authority of maternal residence and for full calendar years. Cases occurring in 2004 were excluded because data were available for only 9 months of the year. A Poisson regression analysis models the number of GBS cases by the aggregated population in a specified time period, which allows for the examination of the distribution of neonatal GBS disease. In this analysis, the aggregation was by health authority and calendar year. An offset, number of live births, was used in the model to account for differences in this denominator across our strata. Incidence rates were calculated per 1000 live births. To examine the possible effects of misclassification and spurious inflation of incidence rates, a second Poisson regression analysis was conducted using an alternative neonatal GBS disease inclusion criterion. This definition excluded the GBS ICD-10-CA code, G00.2, as it includes group A pathogens in addition to group B pathogens (section 3.4.1). Statistical testing on incidence trends were not computed.   To further describe the burden of neonatal GBS disease, the case fatality of neonatal GBS disease was calculated. The numerator included neonates with GBS disease with death as an outcome, and the denominator represented the population of neonatal GBS cases.  3.5 Risk Factors Associated with Neonatal GBS Disease (Objective 1.1) This study also sought to examine the association between maternal and neonatal risk factors, and neonatal GBS disease outcome. Population-based data on parturients and neonates in BC from April 1st 2004 through December 31st 2014 were analyzed. Cases were GBS infant-mother  56 pairs and controls were non-GBS infant-mother pairs. The description of neonatal GBS cases and controls is outlined in section 3.4.1.   3.5.1 Statistical analysis   To examine the risk factors associated with neonatal GBS disease, a multiple logistic regression was used and included gestation specific analyses. The risk factors that were assessed included gestational age, prolonged rupture of membranes, GBS colonization during pregnancy, maternal age (continuous and categorical), mode of delivery (emergency caesarean vs. elective caesarean vs. vaginal), infant weight, labour type (induced, none, spontaneous), use of fetal surveillance monitoring and preterm delivery (Table 3.1). Preterm delivery was assessed using different gestational age cutoffs. Firstly, a dichotomous definition was used and categorized as full term (≥37 weeks) and premature (<37 weeks). The second approach used the World Health Organization’s preterm classification based on gestational age; this included extremely preterm (<28 weeks), very preterm (28 to <32 weeks) and moderate to late preterm (32 to <37 weeks) (WHO, 2016). Due to the small sample size, extremely preterm and very preterm categories were collapsed.  This multiple logistic model was approached in 3 ways. The first model analyzed singletons and twins collectively. Gestation specific analyses were also conducted to assess the risk factors associated with GBS disease for singletons and twins separately. Model building followed a backwards selection algorithm, as outlined in section 3.10. Variables added as random effects included number of infants in current pregnancy and woman person ID (Table 3.1), to account  57 for non-independence of deliveries and for multiple gestations. Unadjusted and adjusted odds ratios, and 95% confidence intervals were computed.   3.6 Maternal GBS Screening, At-Risk Parturients and Antibiotic Administration during Parturition (Objectives 2-4)  This study sought to describe the frequency of GBS screening among parturients in BC (objective 2), the prevalence of indications for IAP among parturients in BC who are at-risk for having a neonate with GBS (objective 3) and the frequency of maternal antibiotic** administration during labour and delivery among at-risk parturients (objective 4). The third project objective was separated into two components; GBS colonization in pregnancy and presence of risk factors among parturients. These project objectives were addressed using descriptive statistics within a retrospective cohort study design using population-based data from the PDR on parturients with live births from April 1st 2004 through December 31st 2014.   3.6.1 Outcome variables  The first outcome in these descriptive analyses was maternal GBS screening (objective 2), categorized as “yes”, “no” or “unknown”. Parturients with a categorization of “yes” were defined as women who received GBS screening during the antenatal period, including during hospital admission for delivery. The second outcome variable, GBS colonization during pregnancy (objective 3), was based on the variable GBS culture result (“positive”, “negative”, “unknown”). Among screened parturients, GBS colonization was defined as a positive culture result. The third                                                 ** Maternal antibiotic treatment was used as a proxy to infer IAP administration.  58 outcome variable was presence of risk factors for GBS neonatal disease in parturients (objective 3). These risk factors included unknown GBS status and preterm birth (<37 weeks) or prolonged rupture of membranes (≥18 hours) at term. Unknown GBS status was defined as the absence of or unknown screening status, or unknown GBS culture result. The last descriptive outcome for objective 4 was the administration of antibiotics among at-risk parturients. This variable was a proxy and was used to assess IAP administration. The variable antibiotic treatment was dichotomized as presence of antibiotics (“yes”) or absence of antibiotics (“no/null”).   3.6.2 Descriptive statistics   Basic cross-tabulations of raw counts were computed to assess the proportion of GBS screening among parturients and among these screened women, the proportion with positive GBS culture results. Proportions were examined by calendar year and health authority of maternal residence. Proportions were examined from April 1st 2004 through December 31st 2014. Proportions from 2004 were based on 9 months of data. The proportion of parturients with risk factors for neonatal GBS disease was also examined using the aforementioned stratification criteria. In addition, among parturients with obstetric risk factors and colonized with GBS, the proportion that received antibiotics was assessed. Significant differences between health authorities were assessed using chi-square tests for projective objectives 2 through 4.  3.7 Factors Associated with Absence of Antibiotic Administration (Objective 5)    The fifth project objective examined factors associated with the absence of antibiotic administration among parturients who are at-risk for having an infant with GBS. Population- 59 based data from the PDR on at-risk parturients in BC from April 1st 2004 through December 31st 2014 were analyzed using multiple logistic regression analysis.     3.7.1 Antibiotic administration outcome  Characteristics associated with the absence of antibiotic administration were examined among parturients deemed to be at-risk. The variable antibiotic administration was dichotomized as presence of treatment (“yes”) or absence of treatment (“no/null”). This variable was a proxy used to infer IAP administration. The at-risk population consisted of parturients with the presence of GBS risk factors. These risk factors included a positive GBS screening result, or unknown GBS status and presence of either (1) preterm birth (<37) or (2) prolonged rupture of membranes (≥18 hours) at term (Appendix C, Figure C.1). Unknown GBS status was defined as a lack of record of screening (“no” or “unknown”) or unknown GBS culture result. Cases were characterized as all at-risk parturients who were not administrated antibiotics, while controls were the remaining cohort of at-risk parturients and recipients of antibiotics.    3.7.2 Statistical analysis   Multiple logistic regression modeling was used to explore characteristics associated with not receiving antibiotics among at-risk parturients. The model building approach is outlined in section 3.10. The variable “woman person ID” was used as a random effect in both models to account for non-independence of deliveries from the same woman. Unadjusted and adjusted odds ratios were computed as well as 95% confidence intervals.    60 3.8 Factors Associated with False Negative Culture Results (Objective 6) The sixth project objective examined maternal characteristics associated with false negative culture results. This objective was addressed using logistic regression analysis within a retrospective cohort study design using population-based data from the PDR on parturients from April 1st 2004 through December 31st 2014.  3.8.1 False negative outcome  A false negative outcome was defined as the maternal screening of a parturient resulting in a negative GBS culture result and a positive neonatal GBS disease outcome. The variables that were used to inform this outcome are categorized as the following. Maternal screening for GBS was categorized as “yes” and screening culture was categorized as “negative”. Controls were defined as the remaining cohort of parturients who were screened with a negative result and had a non-GBS infant.   3.8.2 Statistical analysis  A multiple logistic regression was used to assess maternal factors associated with false negative GBS screening results. Modeling building was approached using a standard backwards stepwise selection methods, outlined in section 3.10. The variable “woman person ID” was used as a random effect in both models to account for non-independence of deliveries from the same woman. Unadjusted and adjusted odds ratios were computed as well as 95% confidence intervals for each model.    61 3.9 Guideline Failures and Lack of Adherence (Objective 7)  The last objective of this study was to examine the application of the SOGC guidelines and identify gaps and limitations of these recommendation. To address this objective, a case series study design was used. A retrospective chart review was conducted at the BC Women’s and Children’s Hospital. Among the neonatal GBS cases identified in the PDR, a subset of GBS infant case-mother pairs from April 1st 2004 through December 31st 2014 were selected based on their place of birth, transfer and/or readmission to this hospital (see section 3.1.2). These data were used to estimate the proportion of neonatal GBS cases which were early-onset, and among this subset, assess the implementation of the guidelines.   3.9.1 Guideline failure type outcomes  For the subset of GBS infant case-mother pairs in which a chart review was undertaken, the type of guideline failure was examined. Guideline failures were identified as either a failure of the guidelines or a failure to adhere to the guidelines. Failure of the guidelines refer to the inherent limitation of screening and prophylaxis strategies despite adherence by providers (see section 2.11). Failure types were comprised of six failure categories. The first category refers to a failure to adhere to the screening guidelines, and is identified in terms of inappropriate screening. This category was stratified according to term status and defined as a failure to screen expecting mothers (“no” or “unknown” screening status), including premature screening of women (<35 weeks), derived from using the date of maternal screening, infant’s date of birth and gestational age. The second failure category is a failure of the screening guidelines. Among term births, this refers to appropriate maternal GBS screening with a negative GBS culture result followed by early-onset GBS disease in the neonate, suggesting a false negative test result (section 3.8.1).  62 Among preterm births, this refers to the inherent limitation of the guideline as delivery occurred before the suggested screening period. The third failure type, failure to treat, is characterized by a failure to provide IAP to GBS colonized parturients or to parturients with indications for antibiotic use based on the SOGC recommendations (Table 3.3). The fourth category is a prophylaxis failure and defined as a failure of IAP administered according to guidelines. This failure is characterized by a manifestation of GBS disease in the neonate despite the administration of IAP to the parturient. The fifth failure category is a treatment failure. This failure is characterized by the manifestation of GBS disease in the neonate despite administration of antibiotics for chorioamnionitis according to the BC Women’s prescriber’s orders.  The sixth and last failure type is defined as a failure to adhere to the guidelines due to suboptimal prophylaxis or treatment. Such parturients requiring IAP received regimens which deviated from the SOGC guidelines in dosage, frequency or antibiotic agent.  Prophylaxis regimens were assessed based on the year and associated guideline version including transition periods. This failure type also included parturients requiring chorioamnionitis treatment whose regimens deviated from the BC Women’s prescriber’s orders in terms of dosage, frequency or antibiotic agent. Figures 4.9 and 4.10 outline algorithms for identification of failure types.                    63 Table 3.3 Indications for antibiotic use††  Indication Previous infant with group B streptococcal disease Group B streptococcal bacteriuria in the current pregnancy Presence of chorioamnionitis and intrapartum fever (≥38°C) in the current pregnancy  Lack of screening (no or unknown status), or unknown culture result in addition to prolonged rupture of membranes (≥18 hours) at term   Lack of screening (no or unknown status), or unknown culture result and premature birth (< 37 weeks)    3.9.2 Statistical analyses   The proportions of failure types were computed using data from the retrospective chart review of EO GBS infant case-mother pairs.   3.10 Approach to Logistic Regression Model Building   A set of multiple logistic regressions were used to address project objectives, 1.1, 5 and 6. Although outcome variables differed between models, the general approach to building these models was similar and is described below.  Univariate analyses were conducted to assess associations between outcome variable and covariates. Variables that were below or equal to the preliminary significance level of 0.1 were included in the multiple logistic regression models. Model building followed a backwards selection algorithm, in which all variables that passed the initial univariate assessments were added into the model. The least significant variables were dropped from the model in a stepwise                                                 †† These risk factors are based on the SOGC clinical practice guidelines on the prevention of early-onset GBS disease, version 2013 (SOGC, 2013).   64 selection process. The purpose of a backwards stepwise selection approach was to build the simplest model.  Checks for confounding, collinear variables as well as significant interaction terms were assessed. A confounding factor was determined by a change in any parameter estimate greater than 10%. Variables meeting this criterion were kept in the model. A collinear variable was determined by inspecting the variance inflation factor (VIF); variables with a VIF of >10 were removed. Plausible interaction terms were examined among variables and insignificant interaction terms (p >0.05) were removed from the model. To ensure goodness of fit, models were compared using likelihood ratio tests (LRT) from multiple imputed data (Buuren & Groothuis-Oudshoorn, 2011) and the Akaike information criterion (AIC), with lower AIC indicating better fit among nested statistical models. Final covariates contributed to model fit and were considered statistically significant when the p-value was <0.05.    3.11  Approaches to Dealing with Missing Values  Certain variables contained missing values and required consideration prior to analysis. Complete case analysis was not used within this study, as it discards records with any missing values, which reduces the sample size, consequently impacting standard errors and statistical power (Zhou, Zhou, Lui & Ding, 2014). In addition to the loss of information, the generalizability of findings is weakened and estimates become biased resulting in invalid inferences (Zhou, Zhou, Lui & Ding, 2014; Dong & Peng, 2013).  Schafer (1999) states that a missing rate equivalent to or less than 5% is insignificant, whereas missing data that exceeds 10% may cause biased results (Bennett, 2001). Therefore, a missing data threshold of 10% was established and variables with missing data above this level were  65 excluded from the analysis (n=34), with the exception of the variable length of rupture of membrane (ROM). This variable is a significant risk factor documented in the literature, which had to be controlled for in multiple logistic analyses. Further, this variable was required to identify at-risk parturients and answer project objectives 4 through 6, and therefore could not be excluded. Our approach to dealing with incompleteness in the ROM variable is described below. The variables that were complete are listed in Table 3.1. For the remaining variables, incompleteness in the data set was addressed using two approaches; multivariate imputations by chained equations (MICE) and a missing indicator method (Table 3.4).   3.11.1 Imputation  A missing indicator method was used to deal with missing data within transformed categorical variables. This method involved the creation of an additional (i.e. dummy) category to identify records with missing values. To address the high degree of incompleteness in the variable ROM, two categories were created, no rupture and unknown. The additional category, no rupture, was created due to a significant proportion (60%) of missing ROM values associated with caesarean section deliveries without indication of labour. Therefore, the level of “no rupture” was derived from values of two other variables, mode of delivery and labour type. The variable labour type was categorized as spontaneous, induced or no labour. The variable mode of delivery had two categories: caesarean section deliveries and vaginal deliveries. Caesarean section deliveries included both emergency and elective procedures. The presence of caesarean section deliveries was not sufficient to infer that there was the absence of rupture of membranes, as emergency procedures may be due to a lack of labour progression following rupture of membranes and/or fetal distress.  To ensure the absence of rupture of membranes in records with caesarean section  66 deliveries, an additional specification was added—the denotation of category “no labour” within the variable labour type. Using the data extraction criteria for abstractors in the PDR manual, our study inferred that the categorization of “no labour” was associated with the absence of rupture of membranes. Abstractors are instructed to assess whether there was at least one of three criteria of labour (painful contractions, cervical dilation and effacement). In their absence as well as the absence of documentation of the date and time of first stage of labour, a categorization of “no labour” is assigned to records (Perinatal Services BC, 2014). Records with blank values for ROM that delivered via caesarean section and had no labour were assigned to the “no rupture” category. Records with missing ROM values had to meet these preceding two criteria; however, if records with missing ROM values had caesarean section deliveries and spontaneous labour or induced labour, they were assigned to the “unknown” category.   A chained equation imputation technique was used to impute continuous variables with a proportion of missing data accounting for less than 10%. The MICE algorithm was implemented using a pre-existing package within the R statistical software. This technique creates an imputation model (i.e. regression model) for each variable with missing values. Incomplete variables, which act as dependent variables, are modeled using independent covariates. Among these covariates, a subset of 36 predictor variables were selected. Using the regression models, missing numeric values were imputed using a predictive mean matching technique. In addition, each imputed data set must undergo 10 iterations, in which the variable with the smallest number of missing values is regressed using the aforementioned methodology. This process is subsequently repeated using the variable with the second least amount of missing values using all complete predictors, including the newly imputed variable (Buuren & Groothuis-Oudshoorn,  67 2011). This cycle continues sequentially until all incomplete variables have been imputed, encompassing a single iteration. The imputed values following the 10th iteration as well as the non-missing data comprise 1 imputed data set. Multiple iterations are required during the imputation process in order to stabilize the predicted values such that the sequential order of imputation of incomplete variables (least to most) is inconsequential. The final output of this process results in 5 complete imputed sets of data for which all non-missing variables are identical but differ in imputed values for variables that were previously incomplete. These data sets were individually analyzed and final estimates were pooled using Rubin’s rules (as cited in Buuren & Groothuis-Oudshoorn, 2011). Hence, this approach allows repeated generation of reasonable imputed values so that variables need not be omitted during analysis, but makes clear the uncertainty associated with this processes by providing final results based on estimates pooled across all imputed data sets.         68 Table 3.4 Implemented imputation techniques  Level Variables Incomplete (%) Variable type Imputation method used Neonatal  Admission weight  0.70% Numeric Multiple imputation  Gestational age 0.11% Numeric Multiple imputation  Place of birth 1.99% Factor Missing indicator Maternal  Total length of stay 2.02% Numeric Multiple imputation  Antepartum length of stay 2.50% Numeric Multiple imputation  Number of previous term deliveries 0.03% Numeric Multiple imputation  Number of previous preterm deliveries 0.03% Numeric Multiple imputation  Number of previous spontaneous abortions 0.84% Numeric Multiple imputation  Place of delivery 2.03% Factor Missing indicator  ‡‡Length of time from rupture of membrane to delivery of 1st baby (i.e. ROM) 18.1% Factor Missing indicator                                                       ‡‡ Approximately 60% of missing values were associated with caesarean section delivery, without the presence of labour. Therefore, it is likely that these records never experienced rupture of membranes, resulting in blank values. Taking this into consideration, the percent of missing values was 7.24%.   69 Chapter 4: Results  4.1 Study Population Characteristics  Table 4.1 provides a compendium of maternal and neonatal demographic characteristics as well as obstetric and neonatal outcome characteristics of infant-mother pairs. There were a total of 463,376 neonates in the Perinatal Data Registry, of which 255 (0.06%) were GBS cases and the remaining 463,121 (99.9%) were considered non-GBS controls. The average gestation and birth weight of neonatal GBS cases was 37.2 weeks and 3118 grams respectively. Their average length of hospital stay was 245.6 hours. For neonatal controls the average gestation, birth weight and length of stay were 38.6 weeks, 3389 grams and 70.4 hours, respectively. There was a higher proportion of premature infants among GBS cases compared to controls (25.5% vs. 9.54%).  There were a total of 456,241 parturients with live births in British Columbia from 2004 through 2014. Among these, 248 (0.05%) were classified as mothers of an infant with GBS and 455,993 (99.9%) were controls. For case mothers, the average age at delivery was 29.7 years old and for control mothers the average age at delivery was 30.7 years old. The age range among all parturients was 11 to 59 years old. In addition, case mothers had overall lengths of hospital stay that were longer, with an average total stay of 88.9 hours in which the average antepartum and postpartum periods accounted for 23.4 and 66 hours respectively. Control mothers had an antepartum and postpartum average length of stay of 12.6 hours and 51.4 hours. Their overall average length of stay was 63.9 hours. Compared to control mothers, there was also a higher proportion of case mothers with positive GBS culture results (51.2% vs. 20%) and prolonged rupture of membranes (19.4% vs. 11.2%).     70 Table 4.1 Descriptive characteristics of the study population classified into GBS infant-mother pairs and non-GBS infant-mother pairs Characteristics  GBS Cases  Non-GBS Controls Neonatal Characteristics and Outcomes Total  255 (0.06%) 463 121 (99.9%) Neonatal sex Female Male  109 (42.7%) 146 (57.3%)   225 387 (48.7%) 237 734 (51.3%)  Mean gestational age  37.2 38.6 Term status Premature Term  65 (25.5%) 189 (74.1%)  44 206 (9.54%) 418 412 (90.3%) Term Status (detailed)  Extremely preterm (<28) Very preterm (28 to <32) Moderate preterm (32 to <37) Term (37 to ≥42)   17 (6.66%) 8 (3.14%) 40 (15.7%) 189 (74.1%)   1887 (0.41%) 3492 (0.75%) 38 827 (8.38%) 418 412 (90.3%)  Weight at birth (grams) <2500g ≥2500g  3118 41 (16.1%)  214 (83.9%) 3389 25 632 (5.5%) 437 353 (94.4%)  Length of stay (hours) 245.6 70.4 Mean NICU days* II III  9.86 11.6  6.8 6.7 Number of infants Singletons Twins  241 (94.5%) 14 (5.5%)  448 737 (96.9%) 14 384 (3.1%)  Discharge to Adoption  Death Foster Care Home Other Hospital Unknown  0 8 (3.13%) 5 (1.96%) 173(67.8%) 69 (27.1%) 0  746 (0.16%)  872 (0.18%)  1991 (0.43%)  448 748 (96.9%) 10 758 (2.3%)  6 (0.001%)  Maternal demographic and obstetric characteristics Total 248 (0.05%) 455 993 (99.9%) Health Authority Fraser (FHA) Interior (IHA) Island (VIHA) Northern (NHA) Vancouver Coastal (VCH)  83 (33.5%) 30 (12.1%) 38 (15.3%) 47 (18.9%) 50 (20.2%)  178 307 (39.1%) 65 565 (14.4%) 66 023 (14.5%) 36 989 (8.1%) 109 109 (23.9%) Age at time of delivery <20 ≥20 29.7 12 (4.8%) 236 (95.2%) 30.7 13 329 (2.9%) 442 664 (97.1%)  71 Characteristics  GBS Cases  Non-GBS Controls Maternal demographic and obstetric characteristics Body Mass Index Normal Obese Overweight Underweight  88 (35.5%) 21 (8.47%) 41 (16.5%) 11 (4.4%)  195 797 (42.9%) 40 440 (8.9%) 66 158 (14.5%) 19 560 (4.3%)  GBS Screening No Unknown Yes  7 (2.8%) 35 (14.1%) 206 (83.1%)  11 996 (2.6%) 60 041 (13.2%) 383 956 (84.2%) GBS Culture Result  Negative  Positive Unknown  76 (30.6%) 127 (51.2%) 45 (18.1%)  283 117 (62.1%) 91 353 (20%) 81 523 (17.9%) Delivery Mode Caesarean section  Vaginal  74 (29.8%) 174 (70.2%)  140 349 (30.8%) 315 644 (69.2%)  Use of Midwife  Yes  No   38 (15.3%) 210 (84.7%)  59 943 (13.1%) 396 050 (86.9%)  Length of rupture of membranes (in hours) <18 hours  ≥18 hours  13.75 176 (70.9%) 48 (19.4%)  9.48 325 970 (71.5%) 50 984 (11.2%)  Length of stay (in hours) Total Antenatal Postpartum  88.9 23.4 66  63.9 12.6 51.4 Fetal surveillance monitoring All** Auscultation & External electronic monitoring  Auscultation only External electronic monitoring & Internal electronic monitoring External electronic monitoring only Internal electronic monitoring only No labour None  15 (6.04%) 55 (22.2%)  60 (24.2%) 8 (3.2%)  78 (31.5%) 0 14 (5.7%) 18 (7.3%)   19 924 (4.36%) 128 884 (28.3%)   105 634 (23.2%)  17 833 (3.9%)  94 250 (20.7%) 1582 (0.35%)  65085 (14.3%) 22 801 (5%)  Abbreviations: GBS= Group B Streptococcus and NICU= Neonatal Intensive Care Unit  Note: Totals may vary due to missing values and/or rounding  *Control mean NICU days are based on a subset of non-GBS controls. NICU II and III account for 7.55% and 3.20% of controls respectively.  ** The category “auscultation & internal electronic monitoring” is also included in this classification    72 4.2 The Burden of Neonatal GBS Disease  The burden of illness was assessed in terms of the incidence rate and case fatality of neonatal GBS disease§§. A total of 263 neonatal GBS cases were identified in the PDR data set that met the case definition during the 2004 to 2014 period. Cases occurring in 2004 (n=20) were excluded. Furthermore, based on the retrospective chart review, 8 neonatal GBS cases previously identified as GBS were excluded, resulting in 235 cases. The average incidence rate in BC from 2005 through 2014 was 0.54 per 1000 live births and ranged from 0.62 to 0.57 per 1000 live births during this period (Figure 4.1). The provincial incidence rate has shown a relatively stable trend, with the highest rate reported in 2006 (0.68 per 1000) and the lowest rate reported in 2012 (0.39 per 1000) (Figure 4.2). The incidence of neonatal GBS disease varied by health authority, with the highest average rate in NHA (1.26 per 1000) and VIHA neonates (0.56 per 1000) and the lowest average rate in FHA (0.45 per 1000), IHA (0.46 per 1000) and VCH neonates (0.49 per 1000) compared to overall rates in BC (Table 4.2). The incidence of neonatal GBS disease in VCH was reduced by 3% as 8 cases identified as GBS were miscoded in the PDR according to the chart review at the BC Women’s and Children’s Hospital. The adjusted average incidence rate in VCH is 0.50 per 1000 live births. The most common outcome in neonatal GBS cases based on ICD-10-CA codes was sepsis (62.9%) (Table 4.3). Among neonates with GBS, 13 cases were identified by ‘other’ specimens, and did not have an ICD-10-CA code.   Among neonatal GBS cases, there were 8 documented deaths in BC from 2005 through 2014. The overall case fatality of neonatal GBS disease was 3.4%. The case fatality varied by year and was the lowest in 2005, 2006, 2008 and 2012, in which no deaths were observed. The highest                                                 §§ Neonatal GBS disease was defined as the presence of disease in the first 28 days of life. Cases occurring from 29-89 days of life could not be assessed due to the scope of the data set.   73 case fatality was observed in 2007 (9.5%), and a declining trend was noted until 2013.  The case fatality in 2013 and 2014 was 5% and 8% respectively (Figure 4.3).  4.2.1 The burden of neonatal GBS disease by varying inclusion criteria   The provincial incidence of neonatal GBS disease was assessed using an alternative inclusion criterion. The GBS ICD-10-CA code, G00.2, was excluded to minimize potential misclassification arising from the inclusion of streptococcus group A (section 3.6.2). There were 9 cases associated with this diagnostic code, reducing the number of neonatal GBS cases to 226 in the years 2005 through 2014. The incidence of neonatal GBS disease ranged from 0.59 to 0.55 per 1000 live births during this time period (Table 4.2). The average incidence rate was 0.52 per 1000 live births. Compared to the all-inclusive case definition, this alternative GBS case definition resulted in marginal decreases in the annual incidence of neonatal GBS disease. The reduction in annual incidence rates ranged from 0.02 to 0.05 per 1000 live births, with an average of 0.02 per 1000 live births. However, the incidence remained unchanged in some years (i.e. 2006, 2009, 2011). This observed provincial incidence rate had a stable trend (Figure 4.4). Excluding the code G00.2, there were 7 documented deaths in BC from 2005 through 2014, resulting in an overall case fatality of 3.1%. A single neonatal death in 2007 was associated with the excluded diagnostic code. Therefore, the case fatality in 2007 decreased from 9.5% to 5.3%. The case fatality of neonatal GBS disease remained approximately unchanged for the remaining years (Figure 4.3).    74 Table 4.2 Incidence of neonatal group B streptococcal disease by health authority from 2005 to 2014, reported per 1000 live births  2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Fraser Health 0.19 0.68 0.42 0.53 0.40 0.59 0.52 0.29 0.41 0.47 Interior Health 0.35 0.17 0.16 0.15 0.76 0.16 1.11 0.33 0.49 0.97 Northern Health 2.35 1.17 1.96 0.55 0.81 1.43 1.45 0.87 0.88 1.15 Vancouver Coastal Health 0.41 0.89 0.1 0.68 0.76 0.39 0.29 0.56 0.38 0.46 Vancouver Island Health 1.35 0.5 0.8 0.94 0 0.62 0.47 0.16 0.48 0.32 Provincial* 0.62 0.68 0.48 0.57 0.51 0.55 0.61 0.39 0.46 0.57 Provincial**  0.59 0.68 0.44 0.55 0.51 0.53 0.61 0.36 0.41 0.55 *This provincial rate includes all group B streptococcus diagnostic codes including:  A40.1, B95.1, G00.2, J15.3, P23.3, and P36.0. ** This provincial rate excludes the group B streptococcus diagnostic code G00.2 and includes: A40.1, B95.1, J15.3, P23.3, and P36.0.                         Table 4.3 Outcomes of neonatal group B streptococcal disease as per the ICD-10-CA codes  ICD-10-CA Code Cases (%) A40.1-Sepsis due to streptococcus, group B 4.2% B95.1-Streptococcus, group B, as the cause of diseases 17.0% G00.2- Streptococcal meningitis (Streptococcus, Group A) (Streptococcus, Group B) 3.83% J15.3-Pneumonia due to Streptococcus, group B 0.85% P23.3-Congenital pneumonia due to streptococcus, group B 11.1% P36.0-Sepsis of newborn due to streptococcus, group B 62.9% Abbreviation: ICD-10-CA= International Classification of Disease, 10th Canadian Edition 75 Figure 4.1 The average annual incidence rate of neonatal group B streptococcal disease by health authority in British Columbia, 2005 through 2014.     Abbreviations: FHA= Fraser Health Authority, IHA= Interior Health Authority, NHA= Northern Health Authority, VCH= Vancouver Coastal Health Authority, VIHA= Vancouver Island Health Authority    00.20.40.60.811.21.4FHA IHA NHA VCH VIHA ProvincialAverage Incidence Rate per 1000 live birthsHealth Authority 76 Figure 4.2 The annual incidence rate of neonatal group B streptococcal disease from January 1, 2005 through December 31, 2014.    Abbreviations: FHA= Fraser Health Authority, IHA= Interior Health Authority, NHA= Northern Health Authority, VCH= Vancouver Coastal Health Authority, VIHA= Vancouver Island Health Authority     00.511.522.52005 2006 2007 2008 2009 2010 2011 2012 2013 2014Incidence Rate per 1000 Live Births YearFHAIHANHAVCHVIHAProvincial 77 Figure 4.3 The case fatality of neonatal group B streptococcal disease in British Columbia from January 1, 2005 through December 31, 2014.            Abbreviation: BC= British Columbia, G00.2= Streptococcal meningitis including Streptococcus group A and group B.         2005 2006 2007 2008 2009 2010 2011 2012 2013 2014BC (including G00.2) 0 0 9.52 0 4.35 4.16 3.7 0 5 8BC (excluding G00.2) 0 0 5.26 0 4.35 4.35 3.7 0 5.5 8.3012345678910Percent 78 Figure 4.4 The annual incidence rate of neonatal group B streptococcal disease from January 1, 2005 through December 31, 2014 in British Columbia using different case definition.    Abbreviation: BC= British Columbia, G00.2= Streptococcal meningitis including Streptococcus group A and group B.   00.10.20.30.40.50.60.70.82005 2006 2007 2008 2009 2010 2011 2012 2013 2014Incidence Rate per 1000 Live BirthsYearBC (including G00.2)BC (excluding G00.2) 79 4.3 Maternal and Neonatal Risk Factors and Neonatal GBS Disease   Neonatal GBS disease (i.e. 0-28 days of life) was associated with maternal, obstetric and neonatal characteristics in univariate and multivariable analyses (Table 4.4). Variables that were assessed in the univariate analyses included GBS culture result, length of time of rupture of membranes, gestational age, prematurity, spontaneous labour type, emergency caesarean section deliveries, maternal age, fetal monitoring, neonatal sex, maternal body mass index, young maternal age, neonatal birth weight, twins and birth order in twin delivery. In univariate analyses, the following variables were significantly associated with higher odds ratio of neonatal GBS disease: twins, birth order in twin delivery, neonatal birth weight and sex, and young maternal age. Normal maternal body mass index based on pre-pregnancy weight was associated with lower odds ratio of neonatal GBS disease in the univariate analysis. However, these associations did not persist in the multiple logistic model. Emergency caesarean delivery revealed no relationship with the odds ratio of neonatal GBS disease in the univariate analysis, nevertheless, this variable was kept in the multiple logistic model to account for potential confounding. Specifically, mode of delivery (e.g. vaginal vs. caesarean section) can be associated with type of labour and the duration of time of rupture of membranes, and should be controlled for in the model. Further, the literature revealed that caesarean section deliveries was an important risk factor of neonatal GBS disease. Further, due to small cell counts, the variable mode of delivery (detailed) was categorized into 3 levels: vaginal deliveries, emergency caesarean and elective caesarean. The original categorization of this variable can be found in Table 3.1. In addition, fetal monitoring was significantly associated with a decrease in the odds ratio of neonatal GBS disease in the univariate analysis. However, this variable was removed from the multiple regression model due to collinearity. Specifically, fetal monitoring contained a  80 category of “no labour”, which was highly correlated with the variable labour type causing a significant increase in standard error. In an effort to examine this risk factor, fetal monitoring was assessed as a dichotomous variable, comparing the presence and absence of fetal monitoring, but no significance was found.  Due to small cell size, the exclusive use of internal fetal monitoring could not be assessed.    The final multiple logistic model included GBS culture result, duration of time of rupture of membranes, gestational age, prematurity, labour type, mode of delivery and maternal age. In the multiple logistic regression, which grouped singleton and multiple births, a positive GBS culture result was a significant predictor of neonatal GBS disease. Neonates born to women with positive GBS culture results had higher odds ratio of neonatal GBS disease compared to neonates born to women with negative GBS culture results (adjusted OR: 5.1; 95%CI: 3.81-6.68). Unknown GBS culture result was also associated with higher odds ratio of neonatal GBS disease (adjusted OR: 1.55, 95%CI: 1.03-2.31). Compared to parturients with a rupture of membrane shorter than 18 hours, prolonged rupture of membranes (≥18 hours) significantly increased the odds ratio of neonatal GBS disease (adjusted OR: 1.68; 95%CI: 1.20-2.34). Moreover, neonates born via emergency caesarean delivery had higher odds ratio of neonatal GBS disease compared to neonates delivered vaginally (adjusted OR: 1.55, 95%CI: 1.13 to 2.11), after controlling for GBS colonization during pregnancy, length of rupture of membranes, labour type, maternal age and gestational age. Labour type was also a significant risk factor for neonatal GBS disease. Neonates born to parturients experiencing spontaneous labour had approximately double the odds ratio of neonatal GBS disease compared to neonates born to parturients with induced labour (adjusted OR: 1.89, 95% CI: 1.32-2.69).   81 Multiple logistic models also revealed the changing effect of gestational age on the odds ratio of neonatal GBS disease. Specifically, every 1-week increase in gestation was associated with a 17% decrease in the odds ratio of neonatal GBS disease (adjusted OR: 0.83, 95%CI: 0.81-0.86). Prematurity (<37 weeks) revealed a significant increase in the odds ratio of neonatal GBS disease among preterm infants compared to term infants (adjusted OR: 3.40, 95%CI: 2.47-4.68). Moreover, compared to full term infants, extremely preterm neonates (adjusted OR: 10.7, 95%CI: 6.61-17.2) and moderately preterm neonates (adjusted OR: 2.53, 95%CI: 1.76-3.66) had significantly higher odds ratio of neonatal GBS disease (Table 4.5). The odds ratio of neonatal GBS disease among post-term infants (≥42 weeks) compared to term infants could not be assessed due to small sample size. Lastly, a weak association was observed between maternal age at time of delivery and neonatal GBS disease. In this instance, for every year increase in maternal age at delivery, there was a 3% decrease in the odds ratio of neonatal GBS disease (adjusted OR: 0.97; 95%CI: 0.95-0.99). No significant interaction terms were found in the final multiple logistic model.             82 Table 4.4 Multiple logistic regression analysis of neonatal group B streptococcal disease among singletons and twins  Covariate Unadjusted odds ratio (95% CI) P value Adjusted odds ratio (95%CI)Y P value Group B streptococcus screening results Negative Positive Unknown   Ref 5.04 (3.8 to 6.7) 1.99 (1.3 to 2.8)    <0.001 <0.001   Ref 5.1 (3.81 to 6.68) 1.55 (1.03 to 2.31)    <0.001 0.03 Length of rupture of membranes <18 hours ≥18 hours No rupture Missing   Ref 1.74 (1.3 to 2.8) 0.38 (0.2 to 0.7) 1.02 (0.6 to 1.7)    <0.001 <0.001 0.91   Ref 1.68 (1.20 to 2.34) 0.79 (0.26 to 2.36) 0.61 (0.33 to 1.09)    0.002 0.68 0.09 Gestational age 0.84 (0.81 to 0.87) <0.001 0.83 (0.81 to 0.86) <0.001 Labour type Induced None Spontaneous  Ref 0.58 (0.33 to 1.04) 1.65 (1.17 to 2.33)   0.06 0.004  Ref 1.10 (0.39 to 2.88) 1.89 (1.32 to 2.69)   0.90 <0.001 Mode of delivery (detailed) Vaginal Elective caesarean Emergency caesarean    Ref 0.93 (0.25 to 1.72) 0.98 (0.75 to 1.27)   0.91 0.88  Ref 0.53 (0.19 to 1.53) 1.55 (1.13 to 2.11)   0.24 0.005 Maternal age 0.97 (0.94 to 0.99) 0.008 0.97 (0.95 to 0.99) 0.009 Fetal monitoring All Ausc and exfm Ausc and infm Ausc only Exfm and infm Exfm only No labour None   Ref 0.55 (0.31 to 1.00) 0.61 (0.08 to 4.7) 0.73 (0.41 to 1.31) 0.62 (0.27 to 1.45) 1.07 (0.61 to 1.90) 0.32 (0.15 to 0.65) 0.99 (0.49 to 2.00)   0.05 0.64 0.30 0.27 0.79 0.001 0.98  -  - Neonatal sex Female Male  Ref 1.26 (0.9 to 1.6)   0.06  -  - Maternal Body Mass Index Missing Normal Obese Overweight Underweight  Ref 0.72 (0.54 to 0.97) 0.79 (0.49 to 1.27) 0.96 (0.66 to 1.39) 0.86 (0.45 to 1.61)   0.03 0.33 0.85 0.63  -  -  83 Covariate Unadjusted odds ratio (95% CI) P value Adjusted odds ratio (95%CI)Y P value Young maternal age ≥20 <20  Ref 1.64 (0.92 to 2.94)   0.09  -  - Neonatal birth weight  0.99 (0.99 to 1.00) <0.001 - - Twins  1.8 (1.05 to 3.09) 0.03 - - Birth order of twins 2.04 (1.01 to 4.14) 0.04 - - Abbreviations: Ausc= Auscultation, Exfm= External electronic monitoring, Infm= Internal electronic monitoring -: Denotes variables not in the final model due to lack of statistical significance  Y Odds ratio from multiple logistic regression model adjusted for GBS results, length of rupture of membranes, gestational age, labour type, mode of delivery(detailed) and maternal age.           Table 4.5 Multiple logistic regression analysis of neonatal group B streptococcal disease with different gestational age cutoffs   Unadjusted odds ratio (95% CI) P Value Adjusted odds ratio (95%CI) P Value Gestational age 0.84 (0.81 to 0.87) <0.001 0.83 (0.81 to 0.86) * <0.001 Term status Term  Preterm  Ref 3.24 (2.44 to 4.29)  <0.001  Ref 3.40 (2.47 to 4.68) **  <0.001 Term status (weeks) Term (³37 weeks) Moderately preterm (32 to 37 weeks) Extremely preterm (<32 weeks)  Ref 2.27 (1.6 to 3.2) 10.2 (6.7 to 15.5)   <0.001 <0.001  Ref 2.53 (1.76 to 3.66) *** 10.7 (6.61 to 17.2)     <0.001 <0.001 Abbreviation: CI=Confidence Interval * Odds ratio from multiple logistic regression model adjusted for GBS results, length of rupture of membranes, gestational age, labour type, mode of delivery (detailed) and maternal age. ** Odds ratio from multiple logistic regression model adjusted for GBS results, length of rupture of membranes, term status, labour type, mode of delivery (detailed) and maternal age. *** Odds ratio from multiple logistic regression model adjusted for GBS results, length of rupture of membranes, term status (weeks), labour type, mode of delivery (detailed) and maternal age.  84 4.3.1 Gestation specific analyses  Gestation specific analyses were also conducted. Among singleton births, similar risk factors were found to be associated with neonatal GBS disease, varying in the magnitude of association (Appendix F, Table F.1). Significant risk factors found to increase the odds ratio for neonatal GBS disease included positive and unknown GBS results at delivery, prolonged rupture of membranes, spontaneous labour, emergency caesarean section deliveries, preterm infants, extremely preterm and moderately preterm infants. Factors associated with a decrease in the odds of neonatal GBS disease included increase in gestational age and maternal age.    Among the twin only multiple analysis, gestational age was the only significant risk factor associated with neonatal GBS disease in both univariate and multiple regression models (Appendix F, Table F.2). Specifically, increase in gestational age was associated with a decrease in the odds ratio for neonatal GBS disease. In addition, extremely preterm infants had increased odds ratio of neonatal GBS disease compared to full term infants. The final multiple regression model was adjusted for colonization status, duration of time of rupture of membranes and spontaneous labour. Although these variables were not significant, they are known risk factors from the literature and were observed to be highly associated with neonatal GBS disease in the previous models and therefore should be controlled for potential confounding. Low statistical power due to small sample size resulted in the absence of emergence of independent risk factors within this stratified model.           85 4.4  Descriptive Statistics  4.4.1 Proportion of maternal group B streptococcus screening  Among parturients in BC, screening for GBS increased from 76.8% to 86.6% from 2004 through 2014, with an overall increase of approximately 10%. The proportion of parturients who were screened for GBS varied by year and health authority (Figure 4.5). The most substantial change in screening was documented among parturients residing in NHA; screening for GBS increased by 25.2% from 2004 through 2014. Nonetheless, the proportion of maternal screening among NHA residents remains the lowest in the province (p<0.001). Contrastingly, parturients from VIHA had the highest annual screening proportion compared to the remaining health authorities (p<0.001). Comparatively, in recent years, screening proportions among parturients from IHA have increased; as such, IHA is one of the health authorities with the highest screening proportions compared to BC overall (p<0.001). From 2004 through 2014, screening in VIHA and IHA increased by 12.7% and 14.4% respectively. During this time period, maternal screening for GBS increased by 8.4% among VCH residents, however annual screening remains below the provincial screening proportion (p<0.001). Lastly, FHA had the smallest increase in screening (4.9%) in approximately 10 years, but had an annual screening proportion that was equal-to or above the provincial proportion (p<0.001).    86 Figure 4.5 The proportion of parturients screened for group B streptococcus from April 1, 2004 through December 31, 2014.***  Abbreviations: FHA= Fraser Health Authority, IHA= Interior Health Authority, NHA= Northern Health Authority, VCH= Vancouver Coastal Health Authority, VIHA=          Vancouver Island Health Authority                                                  *** Proportions for 2004 are based on data from April 1st 2004 through December 31st 2004 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014FHA 81.6 83.0 84.1 84.0 86.2 87.3 87.3 87.0 86.1 86.9 86.5IHA 75.7 80.8 82.5 85.3 86.1 85.7 85.9 87.7 88.2 88.2 90.1NHA 57.1 65.3 72.3 73.7 78.3 79.0 81.7 82.7 81.8 82.0 82.3VCH 75.7 78.7 78.5 80.5 83.4 83.6 83.4 84.4 84.1 84.1 84.0VIHA 77.9 84.1 85.3 86.0 86.7 87.7 87.8 88.9 89.0 90.7 90.7Provincial 76.8 80.3 81.7 82.8 85.0 85.6 85.8 86.4 86.0 86.5 86.60.020.040.060.080.0100.0Percent 87 4.4.2 Proportion of GBS colonization during pregnancy   Among parturients who were screened for GBS, the frequency of GBS colonization during pregnancy appeared stable in BC (Figure 4.6). Specifically, the proportion of GBS colonization during pregnancy ranged from 25.3% to 22.5% from 2004 through 2014. A decrease in the proportion of positive GBS culture results was observed in parturients residing in VIHA (4.5%) and FHA (4.4%). Annually, FHA had proportions of positive GBS results that were below or equal to the provincial proportion; as such, FHA was one of the health authorities with the lowest proportion of parturients with GBS colonization (p<0.001). VIHA documented the highest annual proportion in BC (p<0.001). Among screened parturients residing in IHA, GBS colonization in pregnancy ranged from 23.8% to 22.1% from 2004 through 2014. IHA had annual proportions that were below provincial proportions and had the lowest colonization proportion compared to other health authorities (p<0.001). The proportional trend in NHA has ranged during this time period (22.7%-22.4%), with observed increases from 2011 through 2013. Lastly, the annual proportion of GBS colonization during pregnancy in VCH was higher than annual provincial proportions; these ranged from 23.9% to 23.5% from 2004 through 2014. Consequently, VCH is second only to VIHA in the proportion of maternal positive GBS culture results compared to the remaining health authorities (p<0.001).   88 Figure 4.6 The proportion of positive group B streptococcus culture results among screened parturients from April 1, 2004 through December 31, 2014.††† Abbreviations: FHA= Fraser Health Authority, IHA= Interior Health Authority, NHA= Northern Health Authority, VCH= Vancouver Coastal Health Authority, VIHA=    Vancouver Island Health Authority                                                  ††† Proportions for 2004 are based on data from April 1st 2004 through December 31st 2004 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014FHA 25.9 24.8 24.9 24.0 22.7 22.2 22.7 23.2 23.1 22.3 21.4IHA 23.8 23.5 23.5 21.8 21.0 22.7 23.3 23.0 22.8 22.3 22.1NHA 22.7 22.8 22.1 24.0 21.0 22.0 22.5 24.5 24.9 23.5 22.4VCH 23.9 24.7 23.1 24.4 25.7 25.2 24.5 25.2 24.3 23.2 23.5VIHA 28.5 27.8 25.7 24.9 25.5 26.6 26.5 26.6 26.0 25.1 24.1Provincial 25.3 24.9 24.2 23.9 23.4 23.6 23.8 24.2 23.9 23.0 22.50.05.010.015.020.025.030.0Percent 89 4.4.3 Proportion of parturients with risk factors  Among parturients in BC, the presence of risk factors ranged from 6.2% to 4.8% from 2004 through 2014, with an overall decrease of 1.4% (Figure 4.7). These risk factors include unknown GBS status and preterm birth (<37 weeks) or PROM (≥18 hours). The proportion of parturients with risk factors also varied by year and health authority. During this time period, there was a decrease of 2.4%, 2.1%, 2.1% and 2.0% in the presence of risk factors in parturients residing in NHA, IHA, VIHA and VCH respectively. Annually, parturients from VIHA had the lowest proportion of risk factors in BC (p<0.001). The proportion of parturients from IHA with risk factors was below or equal to the annual proportions in BC (p<0.001). The presence of risk factors was highest among parturients from NHA and VCH compared to the remaining health authorities and were annually above the provincial proportions of risk factors (NHA: p<0.001, VCH: p<0.001). The presence of risk factors among parturients from FHA appeared stable from 2004 through 2014 (5.5%-5.3%). The annual proportions in this health authority were annually below provincial proportions of risk factors (p<0.001).   The presence of risk factors was also analyzed separately. On average, the presence of risk factors among parturients was predominantly associated with preterm birth and unknown GBS status (Appendix D, Figure D.1). Among parturients with risk factors, the provincial proportion of parturients with unknown GBS status and preterm birth gradually increased from 79.8% to 90.8% from 2004 through 2014 (Appendix D, Figure D.2). The provincial proportion of parturients with prolonged rupture of membranes and unknown GBS status decreased from 20.2% to 9.2% during this time period (Appendix D, Figure D.3).  90 Figure 4.7 Proportion of parturients with risk factors* from April 1, 2004 through December 31, 2014.‡‡‡  Abbreviations: FHA= Fraser Health Authority, IHA= Interior Health Authority, NHA= Northern Health Authority, VCH= Vancouver Coastal Health Authority, VIHA=     Vancouver Island Health Authority                                                  ‡‡‡ Proportions for 2004 are based on data from April 1st 2004 through December 31st 2004 *Risk factors include: unknown GBS status and preterm birth (<37 weeks) or prolonged rupture of membranes (≥18 hours) 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014FHA 5.5 5.7 5.5 5.3 4.8 4.6 4.6 4.6 5.2 4.9 5.3IHA 6.1 5.2 5.9 4.9 4.6 5.0 4.8 4.4 5.0 4.5 4.0NHA 7.8 7.8 7.3 6.3 6.1 5.9 5.3 4.8 5.7 5.3 5.4VCH 6.9 6.4 6.4 6.4 5.5 4.9 5.3 5.2 5.3 5.4 5.0VIHA 6.1 5.0 5.1 4.8 5.0 4.7 4.3 4.3 4.2 4.0 4.0Provincial 6.2 5.9 5.9 5.5 5.1 4.9 4.8 4.7 5.1 4.9 4.90.02.04.06.08.010.0Percent  91 4.4.4 Proportion of antibiotic administration among at-risk parturients   There has been a continuous increase in the administration of antibiotics among parturients with positive GBS culture results or other GBS risk factors in BC. From 2004 through 2014, the provincial proportion of antibiotic administration among at-risk parturients increased from 75.5% to 87.2% respectively, an overall increase of 11.7%. The proportion of maternal antibiotic administration varied by year and health authority (Figure 4.8). The most substantial increase of 26.5% was observed among IHA residents, who had significantly lower proportions in earlier years compared to the rest of BC (p<0.001). VIHA also had a substantial increase of 17.5% in the administration of maternal antibiotics during labour and delivery. However, their proportions were regularly below the provincial average for antibiotic administration (p<0.001). FHA and VCH had similar increases of 8.7% and 8.8% respectively. FHA observed annual proportions that were over the average in BC and in recent years, had the highest proportion of antibiotic administration compared to other health authorities (p<0.001). VCH also observed high proportions of antibiotic administration during labour and delivery (p<0.001), ranging from 76.7% to 85.6%, but these have since declined in recent years. Parturients from NHA had the smallest increase in administration of antibiotics of 1.2%. Further, proportions of antibiotic administration were annually below the provincial average (p<0.001). NHA documented an increase in the proportion of antibiotic administration until 2011, but the proportions have since declined. The administration of antibiotic was higher among parturients with GBS colonization compared to parturients with the presence of other risk factors. From 2004 through 2014, the proportion of antibiotic administration ranged from 80.6% to 88% among parturients with GBS colonization and 59.5% to 84% among parturients with other risk factors (Appendix E). 92 Figure 4.8 The proportion of maternal antibiotic administration among at-risk parturients from April 1, 2004 through December 31, 2014.§§§                                                  §§§ Proportions for 2004 are based on data from April 1st 2004 through December 31st 2004 Abbreviations: FHA= Fraser Health Authority, IHA= Interior Health Authority, NHA= Northern Health Authority, VCH= Vancouver Coastal Health, VIHA= Vancouver Island Health Authority    FHA IHA NHA VCH VIHA Provincial2004 81.8 61.8 73.4 76.7 69.3 75.52005 85.9 61.9 72.3 82.3 78.8 79.72006 87.2 72.5 74.7 81.8 78.7 81.72007 85.9 75.8 75.4 82.7 81.1 82.22008 85.7 80.6 73.9 88.4 83.8 84.52009 86 80.9 75.1 88.6 84.2 84.82010 87.1 84.3 77.4 86.9 87.3 862011 88.3 88 81.8 89.4 87.8 87.92012 88.1 87.2 77 87.3 85.3 86.42013 89.4 87.2 77.9 85 87.3 86.82014 90.6 88.2 74.5 85.6 86.8 87.20102030405060708090100Percent 93 4.5 Absence of Antibiotic Administration among At-Risk Parturients  The absence of antibiotic administration among at-risk parturients was associated with demographic, maternal and obstetric characteristics (Table 4.6). At-risk parturients were characterized as having either positive GBS culture result or unknown GBS status and preterm birth or PROM. Because of the large sample size and subsequent statistical power, negligible changes among covariates resulted in statistical significance. However, such statistical significance occasionally detected main effects that were not epidemiologically meaningful.  Maternal age at delivery (adjusted OR: 1.01; 95%CI: 1.01-1.02) and maternal length of hospital stay (adjusted OR: 1.00; 95%CI: 1.00-1.00) were among the variables that were statistically significant. Nonetheless, these predictors were found to have an odds ratio of 1, suggesting these factors do not impact the odds of antibiotic administration among at-risk parturients. These variables were kept in the final multiple regression model as they significantly improved model fit.   In the multiple logistic regression, vaginal delivery among at-risk parturients was a significant predictor of the absence of antibiotic administration. At-risk parturients with spontaneous vaginal delivery without the use of obstetric instruments had higher odds ratio for not receiving antibiotics compared to women undergoing an elective primary caesarean section (adjusted OR: 1.87; 95%CI: 1.65-2.13). In addition, the use of a vacuum during vaginal parturition was significantly associated with the absence of antibiotic administration (adjusted OR: 1.18, 95%CI: 1.02-1.37). Conversely, compared to elective primary caesarean sections, parturients undergoing an emergency primary caesarean section delivery were more likely to receive antibiotics (adjusted OR: 0.81, 95%CI: 0.72 -0.91). Labour type was an additional predictor of absence of  94 antibiotic administration. Specifically, the absence of labour (adjusted OR: 3.52, 95%CI: 3.22-3.85) and spontaneous labour (adjusted OR: 1.65, 95%CI: 1.58-1.73) were significantly associated with lack of antibiotic administration compared to induced labour. Moreover, the absence of antibiotic administration among at-risk parturients varied by health authority. Compared to at-risk parturients in Fraser Health, at-risk parturients from Interior Health (adjusted OR: 1.72; 95%CI: 1.64-1.80), Island Health (adjusted OR: 1.21; 95%CI: 1.15-1.28), Northern Health (adjusted OR: 2.24, 95%CI: 2.11-2.39) and Vancouver Coastal (adjusted OR: 1.07, 95%CI: 1.02-1.12) were significantly less likely to receive antibiotics.   Further, this study revealed differences in the administration of antibiotics among provider types. Specifically, at-risk parturients who use midwives were less likely to receive antibiotics compared to women who did not use a midwife (adjusted OR: 1.75, 95%CI: 1.68-1.83). Differences among 10 provider types (listed in Table 3.1) were assessed in the unadjusted model. However, the only significant association was observed among midwives and midwife trainees. Therefore, a dichotomous variable denoting midwife use was used in the multiple regression model to improve model fit.                    95 Table 4.6 Multiple logistic regression analysis of characteristics associated with absence of antibiotic administration among at-risk parturients  Variable  Adjusted Odds Ratio 95% Confidence Interval P value Health Authority Fraser Interior Island Northern Vancouver Coastal   Ref 1.72 1.21 2.24 1.07  -  1.64         1.80  1.15         1.28  2.11         2.38 1.02         1.12   <0.001 <0.001 <0.001 0.002 Delivery Mode 2 Elective primary Elective repeat Emergency primary Emergency repeat Forceps Forceps and vacuum Spontaneous Vacuum  Ref 1.05 0.81 0.93 1.13 0.98 1.87 1.18  - 0.93         1.18 0.72         0.91 0.82         1.06 0.96         1.33 0.72         1.32 1.65         2.13 1.02         1.37   0.37 <0.001 0.33 0.12 0.9 <0.001 0.02 Labour type Induced None Spontaneous  Ref 3.52 1.65  - 3.22       3.85 1.58       1.73   <0.001 <0.001 Total Length of maternal stay 1.00 1.00       1.00 <0.031 Maternal age 1.01 1.01       1.02 <0.001 Midwife use No Midwife Midwife  Ref 1.75  - 1.68       1.83   <0.001           96 4.6 Characteristics Associated with False Negative Results   Maternal characteristics associated with false negative results were explored in univariate and multiple logistic regression analyses. False negative results were defined as the maternal screening for GBS with negative screening culture, and GBS disease outcome in the infant. As a result of a small sample size (n=77) and low statistical power, few variables were significantly associated with this outcome. In the multiple logistic regression model, maternal characteristics found to be associated with false negative results were demographic and obstetric characteristics (Table 4.7). The odds ratio for false negative results varied by health authority; compared to parturients residing in Fraser Health, a significantly higher odds ratio for false negative results was found among parturients residing in Island Health Authority (adjusted OR: 2.01, 95%CI: 1.04-3.84) and Northern Health Authority (adjusted OR: 3.79, 95%CI: 1.93-7.43). Among obstetric characteristics, prolonged rupture of membranes was significantly associated with a higher odds ratio for false negative outcomes (adjusted OR: 2.84, 95%CI: 1.71-4.66) compared to parturients who experienced duration of rupture of membranes less than 18 hours. Lastly, women with preterm birth had significantly higher odds ratio for false negative results compared to women who gave birth at term (adjusted OR: 2.81, 95%CI: 1.44-5.51). Maternal age at delivery was not significantly associated with false negative results, however, this variable was kept in the final model as it improved model fit. Moreover, spontaneous labour type also improved model fit but was removed from the final multiple regression model due to multicollinearity. No significant interaction terms were found.           97 Table 4.7 Multiple logistic regression analysis of maternal characteristics associated with false negative results  Variable Adjusted Odds Ratio 95% Confidence Interval P value Length of rupture of membranes <18 hours ≥18 hours No rupture Missing  Ref 2.84 0.45 0.94  - 1.71         4.66 0.80         1.89 0.29         3.03   <0.001 0.28 0.92 Health Authority Fraser Interior Island Northern Vancouver Coastal   Ref 1.39 2.01 3.79 1.07  - 0.67           2.91 1.04           3.84 1.93           7.43 0.53           2.15   0.36 0.03 <0.001 0.83 Term status Term (³37 weeks) Preterm (<37 weeks)  Ref 2.81  - 1.44           5.51   0.002 Maternal age 0.97 0.93           1.01 0.30  98 4.7 Guideline Failures and Lack of Adherence among EO GBS Cases  4.7.1 Participant characteristics among chart review sample   In BC, there were a total of 263 neonatal GBS cases (i.e. 0-28 days) from 2004 through 2014 identified in the PDR data set, among which 47 cases originated from the BC Women’s and Children’s Hospital. These cases were identified based on their place of birth, readmission and/or transfer for care to the BC Women’s and Children’s Hospital. Among these 47 cases, 9 cases were excluded following chart review for the following reasons:  6 were caused by other pathogens such as Staphylococcus aureus (n=2) and Streptococcus bovis (n=4); 2 had no signs of illness; 1 neonatal chart could not be located. The remaining 38 neonatal GBS cases were classified as early-onset (n=25, including one set of twins), defined as illness in the first week of life, and late-onset (n=13) defined as illness occurring between 7-28 days of life. These classifications were based on the date of onset of signs indicative of GBS disease and/or date of laboratory results. The remainder of the analysis was restricted to the 25 neonates with early onset disease.    The most common clinical presentation of early-onset GBS disease among these 25 cases was sepsis (68%), followed by meningitis (12%) and pneumonia (8%). The mean gestational age was 34.9 weeks. Among the women (n=24) with neonates with early-onset GBS disease, bacteriuria caused by GBS was present in 16.6% (n=4) of maternal hospital records at time of delivery. Furthermore, chorioamnionitis was documented in 41.6% (n=10) of parturitions. There were no documented maternal refusals for IAP. Table 4.8 provides a detailed description of this sample.     99 Table 4.8 Characteristics of early-onset group B streptococcal disease in infant-mother pairs  Neonatal Characteristics and Outcomes Total  25 Percentage (%) Mean Gestational Age 34.9 - Culture/ isolation positive specimen from* Blood Cerebrospinal fluid Urine Respiratory culture   24 16 4 2  100 66.6 16.6 8.33 Diagnosis Sepsis due to streptococcus, group B (A40.1) Streptococcus, group B, as the cause of diseases(B95.1) Streptococcal meningitis (G00.2) Congenital pneumonia due to streptococcus, group B (P23.3) Sepsis of newborn due to streptococcus, group B (P36.0)  1 2 3 2 17  4 8 12 8 68  Outcome of Hospitalization Survived Died   24 1   96 4 Maternal demographic and obstetric characteristics Total** 24 Percentage (%) Maternal group B streptococcus screening No Unknown  Yes  2 4 18  8.33 16.6 75 Maternal group B streptococcus screening result  Negative Positive UnknownY   5 13 6  20.8 54.1 25 Signs of infection during labour and delivery Chorioamnionitis None  10 13  41.6 54.2 Bacteriuria caused by group B streptococcus  Yes No  4 20  16.6 83.3 * Totals and percentages do not add up to 100% as more than one specimen type may be collected per case  ** Total includes one mother that gave birth to twins.  Y Unknown culture results are due to a lack of maternal screening or unknown screening status         100 4.7.2 Overview  Failures due to lack of adherence to the guidelines were the most predominant failure associated with EO GBS disease (n=13, 52%). This overarching category was further stratified according to deviations at the screening and prophylaxis level resulting in EO GBS cases. This includes lack of adherence to GBS screening (i.e. premature screening) (n=1), failure to treat (i.e. absence of IAP administration) (n=4) and deviation from the recommended chorioamnionitis treatment regimen (i.e. suboptimal treatment) (n=8). The remaining 48% (n=12) of EO GBS cases were irreducible and were associated with failures of the guidelines. These subcategories include screening failures (i.e. false negative test result) (n=3), and treatment/prophylaxis failures (n=9). Treatment/prophylaxis failures were further stratified according to chorioamnionitis treatment failure (n=3) and IAP failure (n=6), administered according to the presence of GBS indicators such as preterm birth, positive GBS colonization or PROM. The following section provides a comprehensive description of the failure types, defined as failure of the guidelines or failure due to lack of adherence to the guidelines, and are outlined based on the screening and prophylaxis segments of the guidelines.    4.7.3 Screening There were a total of 24 women who gave birth to 25 neonates with EO GBS disease at the BC Women’s and Children’s Hospital (Figure 4.9). Among these 24 women, 75% (n=18) of parturients, inclusive of preterm and term births, were screened either during the antenatal period or at time of delivery. Among all screened parturients, screening test dates were available for all with the exception of one maternal record. Appropriate screening was examined based on term status and is outlined below.  101 Failure of the screening component of the guidelines and screening tests The percentage of women with preterm birth (<37 weeks) was 37.5% (n=9). Among this subset, 44.4% (n=4) were not screened at time of delivery. The gestational age of preterm infants ranged from 24 to 34 weeks, occurring before the recommended screening period (35-37 weeks). The remaining subset of women with preterm birth (n=5) were appropriately screened at admission for delivery (n=2) or in the weeks prior to delivery (n=3). All 5 maternal culture results were positive for GBS colonization.   Parturients with term birth comprised 62.5% (n=15). A total of 86.6% (n=13) of women with term deliveries had a known screening status at delivery. Specifically, 69.2% (n=9) of women with screening status at delivery were appropriately screened. Among appropriately screened women with term deliveries, 33.3% (n=3) of neonatal GBS cases were classified as a screening failure. This failure type was the result of appropriate screening completed at 35-37 weeks gestation resulting in a false negative result (section 3.8.1). Because these women had negative GBS screens and the absence of other risk factors, GBS prophylaxis was not indicated, and such cases were not preventable by current screening and prophylaxis guidelines.   Failure to adhere to the screening guidelines   Among women with term births (n=15), 40% (n=6) of women were classified as a failure to adhere to the screening guidelines as a result of inappropriate screening due to lack of screening (n=2) or premature screening (n=4). Among women with premature screening, maternal GBS culture results were documented as positive (n=3) and negative (n=1). This single negative GBS  102 culture result was classified as a failure to adhere to the screening guidelines, as the negative predictive value of the screening test decreases when maternal culture is extracted more than 5 weeks before delivery. Therefore, this neonatal GBS case was attributed to lack of adherence to screening guidelines, as a result of premature screening.   4.7.4 Prophylaxis administration   Prophylaxis recommendations differ according to the presence of various maternal indicators at delivery (Figure 4.10). Maternal treatment or prophylaxis was indicated in 83.3% (n=20) of deliveries.  Failure of the guidelines (i.e. GBS despite adherence to guidelines)   The first indication, chorioamnionitis, occurred in 50% (n=10) of parturients. As per the SOGC guidelines, all cases of chorioamnionitis were treated with antibiotics. However, 30% (n=3) of neonatal GBS cases occurred due to treatment failure. Specifically, the parturients were treated appropriately with a broad-spectrum antibiotic with coverage for GBS at the recommended dosage and frequency according to the BC Women’s Hospital prescriber’s order form. These include broad-spectrum agents for gram positive pathogens such as cefazolin (n=2) and clindamycin (n=1). Antibiotic agents for gram negative pathogens were not collected in our study. Nonetheless, despite appropriate treatment, neonates developed EO GBS disease. The sensitivity of GBS isolates to antibiotics was not assessed in this study.  There were 10 parturients with other GBS indications for IAP administration according to the SOGC guidelines. These included positive GBS culture results (n=7), presence of GBS bacteriuria in the current pregnancy and/or previous infant with GBS disease (n=2), and presence  103 of premature birth and unknown GBS status (n=1). Among women with these GBS indications, the largest portion of neonatal EO GBS cases (60%, n=6) occurred due to a prophylaxis failure. In these circumstances, women were provided with the appropriate antibiotic regimen according to current guidelines. Antibiotic agents included penicillin (n=3), cefazolin (n=2) and ampicillin (n=1). Ampicillin was given based on the 1997 guidelines and prior to the September 2004 guidelines update; therefore, this antibiotic agent was an acceptable option.   Failure to adhere to the guidelines  The remaining women with chorioamnionitis (n=7) were treated using a narrow spectrum antibiotic or administered a differing dosage and/or frequency of a broad-spectrum antibiotic not included on the BC Women’s Hospital prescriber’s order form. These include penicillin, cefoxitin and ampicillin. Therefore, the remaining neonatal EO GBS cases (n=8) were attributed to a failure to adhere to the guidelines due to suboptimal treatment of chorioamnionitis resulting in GBS cases; this total is inclusive of one set of twins of whom both developed EO GBS. There are alternate regimens which can be used for the treatment of chorioamnionitis. However, this study used the BC Women’s Hospital prescriber’s order forms to infer suboptimal treatment of chorioamnionitis.   Neonatal GBS cases attributed to a failure to treat due to the absence of prophylaxis despite clear indication for IAP comprised 40% (n=4) of cases. The predominant factor associated with missed chemoprophylaxis included precipitous delivery after hospital admission (n=3). Only one case was associated with failure to administer IAP despite sufficient time between admission and delivery.  104 Figure 4.9  Maternal screening algorithm identifying guideline failures and lack of adherence.    Abbreviations: EO= early-onset, GBS= group B streptococcus, IAP= intrapartum antibiotic prophylaxis   Failure of the screening guidelines: This is defined as the inherent limitation of the guidelines. For preterm infants, absence of screening was associated with preterm birth occurring before the recommended screening period of 35-37 weeks. Among term infants, this relates to a failure of screening tests to detect maternal GBS colonization (i.e. false negative test result) resulting in 3 cases of neonatal GBS disease.    Lack of adherence to screening guidelines: This is defined as a lack of adherence to screening guidelines due to the inappropriate screening of women. This resulted in 1 case of neonatal EO GBS disease due to premature screening.  105 Figure 4.10 Maternal prophylaxis algorithm identifying guideline failures and lack of adherence.   Abbreviations: EO= early-onset, GBS= group B streptococcus, SOGC= Society of Obstetricians and Gynaecologists of Canada, IAP= intrapartum antibiotic prophylaxis  Treatment/prophylaxis failure: prophylaxis failure refers to the effectiveness of prophylaxis in parturients with GBS indications. This resulted in 6 EO GBS cases among women treated with IAP for GBS indications. Treatment failure refers to chorioamnionitis treatment, which resulted in 3 EO GBS cases.  Suboptimal treatment: Refers to a deviation of the treatment regimens in terms of dosage, frequency or antibiotic agent. This resulted 8 neonatal EO GBS cases from women with chorioamnionitis.  Lack of treatment: Refers to the absence of treatment of women with GBS indications for prophylaxis. This resulted in 4 cases of neonatal EO GBS.    106 Chapter 5: Discussion 5.1  Objective 1: The Incidence and Case Fatality Rate of Neonatal GBS Disease in BC from 2005 through 2014  5.1.1 Incidence   This study examined the provincial burden of neonatal GBS disease in neonates aged 0-28 days of life from 2005 through 2014. This study found an incidence rate that ranged from 0.62 to 0.57 per 1000 live births during this time period. These results were inconsistent with other studies as rates were higher in BC compared to other Canadian provinces and overall incidence in the United States (Davies et al., 2001c; Davies et al., 2001a; SOGC, 2004; Hamada et al., 2008; Health & Schuchat, 2007; Le Doare & Heath, 2013; Jordan et al., 2008). Provincial data from Alberta reported an increasing incidence rate for neonatal GBS disease (0-89 days), ranging from 0.15 to 0.39 per 1000 live births from 2003 through 2013 (Alhhazmi & Hurteau & Tyrrell, 2016). Although, similar to our study, the incidence of EO GBS disease in the United States has reportedly reached a plateau since 2008, it remains lower than the rate in BC at approximately 0.24 per 1000 live births. The incidence of late-onset GBS disease in the US is 0.27 per 1000 live births (CDC, 2010, 2015). Both Canadian and American recommendations stipulate the use of universal antenatal screening at 35-37 weeks gestation as well as maternal chemoprophylaxis. There are a number of possible explanations for the observed differences in incidence rates that are unrelated to difference in prevention guidelines. Firstly, differences in the case definition of neonatal GBS disease impacted the incidence rates observed in our study. Our study assessed neonatal GBS disease in infants up to 28 days of life, while other studies provide separate incidence rates for early onset (<7 days) and late-onset (7-89 days). Alternatively, case definitions that combine EO and LO cases, include overall incidence through 89 days of age.   107 Secondly, intrinsic population differences may result in variances in maternal colonization rates as well as the presence of risk factors. There was a high proportion of GBS colonization among parturients in our study, which ranged from 22.5% to 25.3% during the study period. Additionally, virulence of strains and levels of maternal and infant anti-GBS antibodies can influence disease manifestation (Le Doare & Health, 2013; Edwards & Nizet & Baker, 2016). The incidence rates observed in our study are consistent with rates reported in countries that follow a risk-based approach such as the United Kingdom. For instance, the incidence of neonatal EO GBS disease reported in the UK increased from 0.42 to 0.70 per 1000 live births between 1991 and 2010 (Lamagni et al., 2013). Nonetheless, the average incidence rate in BC (0.54 per 1000) was consistent with the global mean average (0.53 per 1000) reported by Edmond and colleagues (2012). Our study indicates that neonatal GBS disease remains a public health issue, which has implications for neonatal morbidity and mortality.   Our study also revealed regional differences in the incidence of neonatal GBS disease. The average incidence was lowest in FHA (0.45 per 1000 live births) and highest in NHA (1.26 per 1000). The heterogeneity in health authorities is likely a result of various factors. These include differences in maternal screening, risk factors in the pregnant population and GBS colonization during pregnancy.  Specifically, parturients in FHA had high rates of maternal screening and low rates of GBS colonization during pregnancy. In contrast, antenatal screening was lowest in NHA resulting in missed opportunities for prevention in terms of identifying colonized parturients for maternal chemoprophylaxis. Other contributing factors to high incidence rates in the north are limited health care resources and hospitals, particularly tertiary care centres, large geographic  108 distances between health care centers and lack of health care personnel (Riddle & Hutcheon & Dahlgren, 2015; Northern Health, 2016).  5.1.2 Case fatality  Our study assessed the case fatality of neonatal GBS disease in BC. The cause of neonatal death is not documented in the PDR, therefore, neonatal GBS cases in which death was an outcome were inferred to be caused by GBS disease.  The overall number of deaths among neonates with GBS disease in the 10-year period was low, with 8 cases identified in the PDR data set. As a result, annual case fatality rates ranged from 0 to 8% from 2005 through 2014. This range is consistent with data described in the literature. Studies in the US reported a range of 5-9% in EO GBS case fatalities, and 4.7% among LO GBS cases (Phares et al., 2008). The case fatality in other Canadian provinces ranged from 2% to 13.6% in Alberta and 12% in Ontario (Davies et al., 2001c; Porta & Rizzolo, 2015; Adair et al., 2003; Hamada et al., 2008). However, our study reported an overall case fatality in British Columbia (3.4%) that was significantly lower than the reported global case fatality (9.6%) (Edmond et al., 2012). Variance in case fatality can be attributed to differences in prematurity rate and population characteristics. Specifically, preterm infants with GBS have a higher case fatality rate than term cases (Phares et al., 2008). In addition, neonatal GBS case fatality varies based on ethnicity in some regions of the world; case fatality is higher among black infants than white infants in the US. Race may be a proxy for SES and access to health care (Phares et al., 2008). However, our study did not collect information on ethnicity or examine case fatality by term status; therefore, we cannot evaluate this in our study.     109 5.2 Objective 1.1: Examine the Association Between Maternal and Neonatal Risk Factors, and Neonatal GBS Disease Outcome   In our study, numerous maternal and neonatal risk factors were observed to be significantly associated with neonatal GBS disease. These include positive GBS culture results, prolonged rupture of membranes, younger gestational age, spontaneous labour, emergency cesarean section delivery and younger maternal age. These risk factors are consistent with data from the literature (Schuchat, 1999; Melin, 2011; Schuchat et al., 1990; Schuchat, 1998; Edwards & Nizet & Baker, 2016).    Our study revealed extreme prematurity (<32 weeks) and GBS colonization during pregnancy as the strongest predictors of neonatal GBS disease; the latter is a prerequisite for the vertical transmission of GBS (Schrag et al., 2000; Sherman et al., 2012; Melin, 2011). The association between gestational age and neonatal GBS disease suggests insufficient levels of neonatal anticapsular antibodies that are similar to the maternal colonizing strain due to the disruption of maternal IgG antibody transfer (Schuchat et al., 1990). Our study’s findings also suggest prematurity may be a proxy for lack of maternal GBS screening and as a result unknown maternal GBS colonization at time of delivery. These results give further credence for the need for additional prevention methods for preterm infants who are at increased risk for neonatal morbidity and mortality due to GBS and whose mothers are less likely to have received screening and IAP.   This study also identified unknown GBS screening status as a significant risk factor for neonatal GBS disease. Unknown GBS screening in our data is likely a surrogate for absence of GBS screening and/or inadequate prenatal care. Inadequate prenatal care is a known risk factor for EO  110 GBS disease (Schuchat et al., 1994; Schuchat, 1998; Adair et al., 2003; Edwards & Nizet & Baker, 2016). However, adequacy of prenatal care was not examined in this study as this variable was not collected.   PROM is a well-documented risk factor for neonatal GBS disease. Our study found infants born to women who experienced PROM had an odds ratio of 1.68 for neonatal GBS disease compared to infants born to women without PROM. Prolonged rupture of membranes prior to delivery suggests increased neonatal exposure to GBS in utero. In our study, the presence of PROM was also more frequent among women with neonates with GBS (19.4%) than women with non-GBS neonates (11.2%). This finding is similar to what has been reported in the literature (Health et al., 2009; Schrag et al., 2002).   Our analysis also revealed an increase in the odds ratio of neonatal GBS disease with emergency caesarean delivery. Our analysis controlled for risk factors such as PROM and spontaneous labour. Therefore, the relationship observed between emergency caesarean deliveries and neonatal GBS disease was not confounded by factors such as absence of labour and/or PROM, which influences neonatal exposure to GBS in utero. Few studies have explored the association between route of delivery and neonatal GBS disease and those that have report conflicting findings. Some sources report caesarean section deliveries as risk factors, while other studies have reported a higher risk for GBS among vaginal deliveries (Hickman & Rench & Ferrieri & Baker, 1999; Edwards & Nizet & Baker, 2016). Nonetheless, infants delivered via caesarean deliveries are still at risk for infection, as GBS is able to cross-intact membranes (Royston & Geoghegan, 1985). The association found in our study likely suggests the presence of  111 complications or advanced infection in utero and the relationship between mode of delivery and neonatal GBS disease should be investigated further.  Young maternal age, a known risk factor, did not emerge as an independent predictor in our study. In our study, 2.9% of women were considered to be of young maternal age. Due to the small sample size, our study had low statistical power and was unable to detect this effect. However, an increase in maternal age at delivery was observed to decrease the odds of disease, albeit this was a weak association.  Compared to induced labour, spontaneous labour was found to be associated with higher odds ratio for neonatal GBS disease. However, few studies have documented such a relationship (Hannah et al., 1997). Induction of labour is likely indicative of a shorter labour duration, specifically a shorter length of rupture of membranes, reducing neonatal exposure to GBS in utero (SOGC, 2013, Hannah et al., 1997).    5.3 Objective 2: Maternal GBS Screening   The SOGC first recommended routine maternal screening for GBS in 2004. This resulted in a change from one of two alternative approaches to prevention, which included either a risk based or screening based approach, to a universal screening approach (SOGC, 2004). Following this iteration of clinical practice guidelines, antenatal GBS screening in BC increased by approximately 10%. By 2014, up to 87% of parturients were being screened for GBS provincially. This latest screening proportion found in our study is consistent with data in the literature. However, the magnitude of this change (i.e. 10%) was lower than findings reported in  112 other studies. Following congruent revisions to guidelines to a universal screening approach in the United States, antenatal screening in the US increased from 48.1% to 85% from 1998 to 2004 respectively (Van Dyke et al., 2009). An additional study conducted in the US reported a maternal GBS screening rate of 93% between 2002 and 2006 (Schrag & Verani, 2013). In the absence of articulated targets for achievements and based on our study’s findings, it can be inferred that universal screening has been adequately accepted given its implementation provincially. Nonetheless there remains a gap in maternal screening likely as a result of inadequate antenatal care and preterm births. Our study revealed that among unscreened parturients in BC in 2014, 61% had infants that were born before 35 weeks gestation. The proportion of unscreened women decreased to 24% when infants were born from 35 to 37 weeks gestation, the recommended screening period. Women with term infants (≥38 weeks) had the lowest unscreened proportion (9%). Our findings highlight the limitations of screening among preterm infants born before the recommended screening period and the reducible proportion of missed opportunities for screening among term infants. Therefore, uptake in maternal screening is influenced by term status. Absence of GBS screening can also relate to the presence of GBS bacteriuria during the current pregnancy. According to the guidelines, these women would be considered colonized at parturition and would not require additional screening in the 3rd trimester (SOGC, 2012). GBS bacteriuria is present in up to 4% of pregnancies (Wood & Dillon, 1981), and may account for the absence of or unknown screening status among parturients in this study. Nonetheless, this variable is not collected by the PDR and could not be assessed in this study.   There was high adherence to screening guidelines in IHA and VIHA, as screening was observed in up to 90% of parturients in recent years, indicating further improvement is possible. Despite  113 overall high screening proportions, improvements are still required. Regions such as NHA had the lowest proportion of GBS screening among parturients. Factors associated with failure to screen can include lack of or inadequate prenatal care, history of drug use, previous delivery of a live infant, lack of access to health care and preterm birth (Pulver et al., 2009; Van Dyke et al., 2009, CDC, 2010; Riddle & Hutcheon & Dahlgren, 2015). Variances in health authorities can be due to differences related to practice and/or differences in the population of parturients. Other explanations involve poor-resource settings resulting from the vast geographic area (Riddle & Hutcheon & Dahlgren, 2015; Northern Health, 2016). Our findings suggest challenges remain in fully implementing universal screening, resulting in missed opportunities for prevention, particularly in certain regions in BC. This has serious implications for the burden of neonatal illness and warrants further investigation. Additional research is also needed to assess adherence to timing of screening, an essential component to the accuracy of maternal colonization status.    5.4 Objective 3: Describe the Proportion of Parturients with IAP Indicators and therefore Examine the Proportion of At-Risk Parturients in BC   This study found a high at-risk population in British Columbia. These results revealed an average of 29% of parturients in BC have risk factors. Further, the most predominant risk factor found in this study was maternal colonization with GBS; an average of 24% of parturients in BC are colonized with GBS. The remaining proportion of parturients (5%) have unknown GBS status and preterm birth or PROM at term.     114 5.4.1 Positive culture results  The literature shows GBS colonization during pregnancy varies due to differences in population characteristics (Regan et al., 1991; Edwards & Nizet & Baker, 2016; Money et al., 2008, Young et al., 2011). Our study found the overall colonization among parturients in BC ranged from 22.5% to 25.3% from 2004 through 2014. This is in contrast with a recent study set in British Columbia that reported a maternal colonization rate of 30% (Money et al., 2008). This discrepancy in GBS colonization relates to differences in sample characteristics and data sources. Specifically, Money et al. (2008) were limited to recruitment from one clinical setting. Our study is more representative of actual trends in BC as provincial data were used to calculate the proportion of GBS colonization. Another single centre study conducted in Ottawa observed a colonization rate of 9.5% among pregnant women (Allan et al., 1999), while a study conducted using a cohort of pregnant women from the Calgary Health Region had a colonization rate of 19.5% (Spaetgens et al., 2002).  Regional differences in maternal colonization were observed. The highest colonization proportions were among parturients from VIHA (28.5%-24.1%), while the lowest was among IHA residents (23.8%-22.1%). The heterogeneity within and across provinces can be attributed to the transient nature of colonization, differences in sampled body sites, population characteristics as well as laboratory procedures performed (i.e. culture methods) (Melin, 2011; Edwards & Nizet & Baker, 2016; Regan et al., 1991; Money et al., 2008). However, due to the limitations of the data, differences in culture methods used by laboratories in each health authorities could not be assessed. This study revealed a high at-risk population among parturients in British Columbia, as maternal colonization is a prerequisite for vertical transmission.   115 5.4.2 Presence of GBS risk factors  Preterm birth and prolonged rupture of membranes have been identified as concomitant risk factors in neonates with EO GBS disease (Edwards & Nizet & Baker, 2016). Canadian guidelines stipulate the administration of IAP based on the presence of one of these risk factors in conjunction with unknown GBS status. In our study, an average of approximately 5% of parturients in BC had one of these risk factors during parturition; this includes unknown GBS status and preterm birth or PROM. The combination of these risk factors ranged from 6.21% to 4.85% from 2004 through 2014. Reduction in the presence of risk factors was attributed to increases in maternal GBS screening. Risk factors such as PROM and preterm births are only considered indicators for prophylaxis in the presence of unknown GBS status at delivery.   Regional differences in the presence of maternal GBS risk factors were also observed. Specifically, a higher proportion of women in NHA had the presence of risk factors. The presence of risk factors was lowest among parturients from VIHA. These regional variances are due to differences in the proportion of screening in these health authorities. Maternal screening for GBS was lowest in NHA and highest in VIHA. This can be reflective of deviances in screening adherence among health authorities, differences in adequacy of prenatal care or reflective of maternal acceptance of screening.   Our study revealed that the presence of risk factors among unscreened parturients was predominantly associated with preterm birth. This is consistent with data from the literature, as preterm birth among unscreened parturients was more frequent than PROM (Bianco et al., 2016). The proportion of unscreened women with preterm birth in BC increased from 79.8% to 90.8%  116 from 2004 through 2014, respectively. As previously discussed, preterm births contribute to the gap in screening uptake, particularly when birth occurs before the recommended screening period. Our study further confirms the inherent limitation of GBS screening among preterm births, previously established in the literature. These findings have serious public health implication for neonatal morbidity and mortality, as preterm infants have a higher risk of neonatal GBS disease and mortality (Mario & Valenzuela & Vasquez & Illanes, 2013; Hanson & VandeVusse, 2010). Additional preventive methods such as rapid screening at delivery in addition to improved prophylaxis implementation, are warranted to target this subgroup of infants.   Our study revealed that the remaining unscreened women with risk factors were associated with PROM. Prolonged rupture of membranes is considered a risk factor among term deliveries, as PROM is more predominant among GBS colonized parturients compared to non-colonized parturients (Heath et al., 2009; Schrag et al., 2002). From 2004 through 2014, the proportion of unscreened women with PROM decreased from 20.2% to 9.2% respectively. This decline coincided with increases in maternal GBS screening. Therefore, there are more opportunities for screening during the recommended period (35-37 weeks) among infants born at term, which further contributes to the lower proportion of this risk factor compared to preterm birth and unknown GBS status. Our finding also suggests continued adherence to GBS screening guidelines.      117 5.5 Objective 4: Antibiotic Administration among At-Risk Parturients   The present study also assessed the proportion of at-risk parturients who received antibiotics during parturition. Antibiotic administered during labour was used as a proxy to assess IAP administration in the analysis of the PDR data set, which did not contain information about the ‘reason’ for antibiotic use nor the details of which antibiotics were administered. This at-risk population consisted of parturients with the presence of positive GBS screening results or unknown GBS status and the presence of either preterm birth or PROM at term. Results indicated approximately a 12% increase in the proportion of antibiotic administration in at-risk parturients from 2004 through 2014. Currently, 87.2% of at-risk parturients in BC are administered antibiotics during parturition.  Regional differences among antibiotic administration were observed. The proportion of antibiotic administration was lower among at-risk women from NHA. Moreover, in recent years, antibiotic administration among this group of women appeared to decrease. At-risk women from FHA had the highest proportion of antibiotic administration. Our study revealed regional differences in the time between hospital admission and delivery. Specifically, the mean time from admission to delivery was 9.6 hours and 12.5 hours in NHA and FHA respectively. Differences in lengths of antepartum stay can be associated with restricted opportunities for the administration of antibiotics. Length of maternal hospital admission before delivery is a documented factor associated with lack of IAP (Van Dyke et al., 2009; Stoll et al., 2011; Schrag & Verani, 2013). These suggest differences are related to access to health care or birthing preferences including maternal refusal of antibiotics. Such findings require further investigation.   118 Our study revealed gaps remain in antibiotic administration. Women with risk factors including unknown GBS status and preterm birth or PROM had a lower proportion of antibiotic administration compared to women with positive GBS culture results. From 2004 through 2014, women with positive GBS culture results had higher proportions of antibiotics administration (80.6%-88%) compared to women with risk factors such as unknown GBS status and preterm birth or PROM (59.5%-84%). Nonetheless, both groups observed increasing proportions in antibiotic administration. Indicators for maternal prophylaxis have evolved in the GBS prevention guidelines over time. Prior to a universal screening approach, a study conducted from 1998 through 1999, found 61% of women with risk factors such as preterm birth and PROM, were provided IAP, and 89% of GBS colonized women received IAP (Schrag et al., 2002). Following the universal screening era, 85.1% of women received prophylaxis based on GBS positive status at delivery or in the presence of risk factors (Schrag & Verani, 2013). Specifically, a range of 76% to 87.5% of GBS colonized women received IAP, while 66% of women with other risk factors received IAP (Stoll et al., 2011; Pulver et al., 2009). Our results supported these findings. This requires better implementation of current guidelines, particularly for women who are unscreened and present with risk factors, leading to missed opportunities for prophylaxis. 5.6 Objective 5: Maternal Factors Associated with Absence of Antibiotic Administration among At-Risk Parturients   Our study revealed factors such as health authority, spontaneous labour, vaginal delivery and midwife assisted delivery were associated with absence of antibiotic administration during parturition.     119 Despite statistical significance of length of maternal hospital stay (OR: 1.00, 95% CI:1.00- 1.00), which encompasses both antenatal and postpartum stays, the odds ratio suggests that this factor does not affect antibiotic administration. The variable, antenatal length of stay, which assesses the time between admission and delivery of infant, was not significantly associated with the absence of antibiotic administration. This finding was inconsistent with other studies, as short length of stay was significantly associated with missed opportunities for maternal chemoprophylaxis. Specifically, increase in length of stay between admission and delivery was associated with increase in antibiotic administration (Stoll et al., 2011; Van Dyke et al., 2009). A potential explanation for the lack of evidence found in this study is due to the use of a proxy variable for IAP administration and the potential for misclassification of exposure status to be included in this variable.   Our study suggested that women who deliver vaginally, including spontaneous vaginal birth and delivery with the assistance of a vacuum, were less likely to receive antibiotics compared to women with caesarean sections. Vaginal deliveries may occur rapidly, which minimizes the opportunity for antibiotic administration. Our study also found that women undergoing an emergency primary caesarean section were more likely to receive antibiotics. Emergency caesarean section may also suggest the presence of complications or advanced infection in utero, leading to IAP administration. Previous studies had similar findings and reported women with vaginal deliveries were less likely to receive IAP (Stokholm et al., 2013; Goins et al., 2010). Alternatively, the relationship between mode of delivery and absence of antibiotic administration may relate to the standard of care of caesarean section deliveries. Specifically, antibiotics such as  120 cefazolin are used for the prevention of wound infections. Therefore, the relationship that was observed in our study may be unrelated to IAP administration for GBS.   Additional predictors found to be significantly associated with higher odds ratio for no antibiotic administration during parturition include absence of labour and spontaneous labour. The absence of labour suggests an absence of rupture of membranes, and in such instances maternal chemoprophylaxis is not indicated (Goins et al. 2010, SOGC, 2013; Bienenfeld & Rodriguez-Riesco & Heyborne, 2016). However, findings from the literature report that women with induced labour are less likely to receive antibiotics, contrary to what was observed in our study (Bienenfeld & Rodriguez-Riesco & Heyborne, 2016). It is possible that spontaneous labour in our study was a proxy for rapidity of delivery ensuing in inadequate time for IAP administration. Our study did not find maternal age to be associated with absence of antibiotic administration, as shown in the literature (Stokholm et al., 2013; Goins et al. 2010).   Regional differences were also observed in risk factors for no IAP. Specifically, at-risk parturients residing in IHA, VIHA, VCH, and NHA were significantly less likely to receive antibiotics, compared to women residing in FHA. These differences are suggestive of variances in practices between health authorities.   Lastly, differences in antibiotic administration were observed between providers. Specifically, women who used a midwife were less likely to receive antibiotics compared to women who did not use a midwife during parturition. In contrast, a study conducted by Paccine and Wiesenfeld (2013) in the US revealed the absence of association between IAP administration and provider  121 type. Nonetheless, differences in labour management are documented in the literature between midwives and physicians. Women who use midwives are characterized as having a shorter first stage of labour and reduction in oxytocin use (Bodner-Adler et al., 2004, Chambliss et al., 1992).    5.7  Objective 6: Investigate the Characteristics of Women with False Negative Results    According to the literature, a significant proportion (61-82%) of neonatal GBS cases occur due to false negative test results. However, there is a gap in the literature that assesses maternal characteristics associated with false negative results. Our study revealed demographic and obstetrics characteristics associated with false negative results. Specifically, parturients residing in NHA and VIHA had a significantly higher odds ratio of false negative test results compared to parturients residing in FHA. These findings suggest variances in the method of collection and testing of specimens in these health authorities. A swab of both the vaginal and rectum results in higher sensitivity than sampling only from the vagina. Further, the use of selective broth culture medium yields better results than nonselective blood agar (El Aila et al., 2010; Philipson & Palermino & Robinson, 1995). However, laboratory techniques could not be assessed with the available data from this study. Alternatively, our findings may suggest differences in timing of screening, as sensitivity and specificity of screening tests decrease if they are taken more than 5 weeks before delivery (Yancey et al., 1996; Hanson & VandeVusse, 2010). Further studies are needed to assess whether differences in timing, laboratory and culture procedures cause false negative results in BC. Adding a variable in the PDR to collect the date of screening would enable the determination of whether screening timing is the cause.      122 Factors such as preterm birth and prolonged rupture of membranes were also associated with higher odds ratio for false negative GBS screening results. The literature reports that PROM is more frequent among GBS colonized parturients than non-colonized mothers (Heath et al., 2009; Schrag et al., 2002). Further, maternal GBS colonization at time of delivery has been positively associated with preterm delivery (Valkenburg-Van Den Berg et al., 2009). Therefore, the presence of these risk factors could indicate the presence of maternal colonization during delivery. This warrants further investigation and may necessitate additional screening among women with such indicators during parturition. Further, this highlights the importance of providing IAP to women with preterm birth, who are less likely to be screened. Additional studies are required to assess the validity of such findings, which may have implications for future guidelines as well as clinical management.   5.8 Objective 7: Examine the Application of the SOGC Guidelines and Identify Gaps and Limitations of these Recommendations The results from the retrospective chart review that was a part of this study suggest that the guidelines are an effective prevention strategy, given the limited number of neonatal GBS cases at the BC Women’s and Children’s Hospital. However, they also highlighted failures due to a lack of adherence to these guidelines resulting in 52% of EO cases as well as failure of the guidelines, resulting in the remaining 48% of EO cases.   123 5.8.1 Screening Failure of the screening guidelines and screening tests The initial preventive measure in clinical practice guidelines is universal maternal screening at 35 to 37 weeks gestation. Our study found that 44% (n=4/9) of women with preterm birth were not screened. In these instances, the gestational age ranged from 24-34 weeks. Therefore, the absence of screening in these cases was related to the occurrence of delivery prior to the recommended screening period. This is a limitation of the guidelines that has been highlighted by other studies (Pulver et al., 2009; Stoll et al., 2011; Van Dyke et al., 2009). Specifically, neonatal delivery before 34 weeks gestation is a significant risk factor associated with lack of prenatal screening (Van Dyke et al., 2009). Ergo, premature birth limits opportunities for prevention; this shortcoming has important implications for clinical practice and warrants supplementary attention.  An additional shortcoming highlighted by this study was the failure of the screening test, categorized under failure of the guidelines. Among the subset of cases at the BC Women’s and Children’s Hospital, a total of 33.3% (n=3/9) of neonatal GBS cases were among women with negative screening results. In other studies, infants with GBS disease born to women with negative screening cultures ranged from 61% to 82% (Pulver et al., 2009; Puopolo & Madoff & Eichenwald, 2005; Van Dyke et al., 2009). Factors that impact GBS cultures include inadequate sampling, delays in processing, suboptimal laboratory technique, use of non-selective media and recent maternal antibiotic use (Stoll et al., 2011; Goins et al., 2010). Alternatively, due to the transient nature of colonization, this can be suggestive of late colonization following screening. Therefore, appropriately screened women who were true negatives at the time of screening were colonized subsequent to being screened. Lastly, these differences can be due to higher screening  124 proportions found in other populations (Pulver et al., 2009). Additional research is necessary to investigate adherence to laboratory specimen collection and transport in BC and consequently the origin of false negative results. Our findings emphasize that there are some unavoidable limitations to the guidelines that result in residual cases. These gaps can be remedied with the help of maternal immunization or with screening tests with a higher sensitivity.   Failure to adhere to screening guidelines  Among term infants, lack of adherence to the screening guidelines was due to either (1) a lack of screening (n=2) or premature screening (n=4). A single neonatal GBS case was associated with a negative maternal screening culture result on a specimen collected prior to 35 weeks gestation. The sensitivity and specificity of the screening test decreases when screening is completed more than 6 weeks before delivery, as previously described (Hanson & VandeVusse, 2010; Yancey et al., 1996). This emphasizes the importance of screening within the appropriate time period.    5.8.2 Prophylaxis  Failure of the guidelines (i.e., GBS despite adherence to guidelines)  Up to sixty percent of neonatal GBS disease at BC Women’s and Children’s Hospital occurred in women who had been appropriately managed according to guidelines (“failure of guidelines”). Therefore, there was a failure of the guidelines due to treatment and/or prophylaxis failure. This includes chorioamnionitis treatment (n=3/10) and prophylaxis of women with indications for GBS prevention (n=6/10) such as positive GBS culture results, GBS bacteriuria in the current pregnancy and unknown GBS status and preterm birth. The effectiveness of IAP decreases in the  125 presence of intrapartum fever and when administered less than 2 hours before delivery (Lin et al., 2001; Schrag & Verani, 2013). Additional influencing factors include neonatal infection in utero, resulting in the ineffectiveness of IAP, as well as GBS strain resistance. Studies in the US have revealed an increase in antibiotic resistance among GBS isolates. Specifically, clindamycin is an antibiotic agent recommended for use among penicillin allergic parturients who are at risk for anaphylaxis, however, resistance to clindamycin ranges from 4% to 33%. (Blaschke et al., 2009; Murdoch & Reller, 2001; Phares et al., 2008; Castor et al., 2008; Stoll et al., 2011). Reassuringly, no cases of resistance to penicillin have been reported, the first line agent in maternal chemoprophylaxis. Results from our study revealed only one case that received clindamycin, while the remaining women were administered penicillin, ampicillin or cefazolin. This study did not collect information on GBS strain sensitivity to specific antibiotics. However, the failures of the guidelines identified in this study appear to be unrelated to antibiotic resistance and are most likely related to the lack of effectiveness of IAP. Effectiveness of the guidelines could be improved by additional prevention methods such as immunization. A larger study using chart review data in British Columbia would also provide a more robust estimate of effectiveness of the guidelines including IAP. Such a study should collect more details including total duration of prophylaxis and strain sensitivity to antibiotics.   Failure to adhere to the guidelines 	In our study, a failure of adherence to guidelines due to absence of IAP was observed in 40% (n=4/10) of neonatal GBS cases. The predominant explanation for absence of use of IAP was precipitous delivery in which the women gave birth within an hour of hospital admission. In such instances, the physicians’ ability to administer maternal chemoprophylaxis is impeded. Findings  126 from the literature revealed that 24%-34% of neonatal GBS cases are due to the absence of IAP (Stoll et al., 2011). It has been indicated that the interval of time between admission and delivery affects the capacity for IAP. Specifically, deliveries occurring less than 4 hours after hospital admission result in lower odds of chemoprophylaxis administration (Van Dyke et al., 2009; Stoll et al., 2011). As well, the duration of prophylaxis for the reduction of vertical transmission has been debated. Prophylaxis is effective at reducing the vertical transmission of GBS if provided ≥4 hours before delivery; there is some limited conferred protection if prophylaxis is administered ≥2 hours before delivery (Lin et al., 2001; de Cueto et al., 1998). Thus, timing of delivery limits the capacity to implement guidelines effectively. These factors further contribute to the residual burden of illness.   Failure to adhere to the guidelines due to suboptimal treatment of chorioamnionitis resulted in 8 neonatal GBS cases. Specifically, there was a deviation from the BC Women’s prescriber’s order form. However, the antibiotic regimens listed on the prescriber’s order form are not part of the SOGC guidelines but a local hospital guideline and there are other potential antibiotic regimens to treat chorioamnionitis. Given the non-specificity of the SOGC guidelines for chorioamnionitis treatment, this study used the antibiotic regimens on the BC Women’s prescriber’s order form to infer nonadherence and thus suboptimal treatment. Our study found that the majority of chorioamnionitis cases were provided with alternate treatment regimens; all antibiotic agents covered GBS. These included ampicillin, penicillin and cefoxitin. This may suggest the misclassification of the 8 GBS cases due to suboptimal treatment; rather, these cases may instead be attributed to a treatment failure. This finding requires further investigation. Factors contributing to inadequate and suboptimal prophylaxis include lack of susceptibility testing of  127 GBS cultures, preterm deliveries, penicillin allergies, as well as lack of recommended prenatal serologic testing for other infectious diseases, having intact membranes at admission, induced labour and prior vaginal birth (Bienenfeld & Rodriguez-Riesco & Heyborne, 2016; Goins et al., 2010). Therefore, the proportions of EO GBS cases occurring due to suboptimal/nonadherence to guidelines are preventable. However, the feasibility of prevention is contingent on the ability to maximize adherence to recommendation practices.   Extrapolation The results from the retrospective chart review found that early-onset GBS disease represented 65.7% of cases. If this proportion was applied to the estimates of the incidence of neonatal GBS from the PDR data set, this would decrease the GBS case count (n=255) to 168 EO cases in BC from April 1st 2004 through December 31st 2014. Our findings also indicated that approximately half of EO cases (n=84) were associated with lack of adherence to the guidelines. However, the retrospective chart review was conducted on neonatal GBS cases born, transferred and/or readmitted to the only hospital that provides tertiary care for high risk pregnancies and neonates as well as the local population. The patterns of practice and guideline adherence in this tertiary care hospital may not be representative of practice throughout the province or reflective of smaller, community hospitals. Additionally, as the confidence intervals around the rates based on the small numerator endpoints in the chart review would be wide, no attempt was made to extrapolate the observations at this one hospital to estimates of preventability of GBS province-wide.    128 5.9 Limitations  This study had several limitations. This study found that the PDR data is not designed to effectively evaluate the implementation and effectiveness of the GBS prevention guidelines. The PDR does not contain information about date of onset of illness in neonates with GBS; therefore, early onset cases cannot be identified. As a result, this study was unable to reliably assess the incidence of early onset GBS cases, and the estimates based on infants diagnosed within the first 28 days of life overestimate the incidence of early onset disease.  Other limitations from the PDR data include the absence of cause of neonatal death. Therefore, it was inferred that neonatal GBS cases in which death was an outcome were due to GBS disease. The variable denoting maternal antibiotic administration during labour does not include the reason for treatment; antibiotics may be administered for other non-GBS indications such as chorioamnionitis. However, our study limited the use of this variable to a specific subset of women with risk factors for GBS. While the PDR includes results of maternal vagino-rectal GBS screening, it does not contain information on date(s) or gestational week of screening, which are an important component of the screening guidelines. Further, this study may have underestimated adherence to the screening guidelines, as the presence of GBS bacteriuria in the current pregnancy is not recorded in the PDR. Women with GBS bacteriuria do not require vagino-rectal screening as they are considered to be colonized (section 2.10.5). Dates associated with neonatal admissions and discharges were not requested from the PDR. Ergo, neonatal GBS cases with numerous hospital admissions and discordant ICD-10-CA codes could not be arranged sequentially, and we were unable to determine if GBS disease was due to the reoccurrence of disease or a single manifestation ensuing in multiple transfers. Nonetheless, multiple transfers with different diagnostic codes resulted in 9 cases in this data set.   129 Our study exclusively used data on live births and excluded stillborn neonates and their respective mothers. As a result, our study’s GBS incidence rate is not comparable to incidence rates that include all births, inclusive of stillbirths and live births. Data on stillbirths were excluded as the PDR does not contain information about the presence of GBS from placenta culture or other sterile sites sampled from the stillborn that could identify GBS as a factor.  An additional limitation of our incidence calculation was the inclusion of positive infectious agent culture found in ‘other’ specimen. This criterion includes GBS culture results from both sterile and non-sterile sites. This could have resulted in the inclusion of cases that were not truly GBS disease cases, therefore increasing the incidence rate. However, these records only represented 5% of cases.   This study also assumed that mothers who had negative GBS screening results (termed ‘false negative’ in this study) were not subsequently colonized with GBS. Our assumptions about the means by which such cases can be prevented, for at least a portion of such cases, may be incorrect, as we have assumed that all such women were colonized at the time of screening but colonization was missed due to sample collection or transport methods, rather than colonization after the screening.   Data from January 1st 2004 through March 31st 2004 were not retrieved for this study, as the PDR started collecting GBS screening data as of April 1st 2004. This may have led to the underestimation of the proportions of GBS screening, maternal colonization, presence of risk factors and administration of antibiotics for the year 2004.    130 5.10 Implications for Policy and Recommendations  GBS preventive guidelines are not expected to prevent all cases of neonatal early-onset GBS disease due to inherent limitations such as late maternal colonization and precipitous labour, and are inefficacious against late-onset manifestation of disease. The results from this study highlight the remaining challenges and limitations of universal screening and maternal chemoprophylaxis in British Columbia. Despite the successful implementation of the guidelines and evidence of improvements in the frequency of screening during pregnancy, some gaps remain. Problems with adherence can be due to lack of familiarity or confusion regarding the current guidelines, deviation from screening and prophylaxis regimens, and lack of access to maternal screening and culture status, including date of screening, which may not always be available to the provider at time of delivery. This underscores the need to educate health professionals about the risk factors for neonatal GBS disease, and the importance of adherence to recommendations to reduce the residual burden of illness. To address the vulnerability of unscreened pregnancies, future iterations of guidelines should consider the benefit of using rapid GBS diagnostic tests for women presenting in labour prior to being screened. This would aid in determining the need for maternal prophylaxis and further reduce the incidence of neonatal GBS disease.   There are also inherent limitations to the guidelines that impede health professionals’ ability to provide care such as maternal refusal of prophylaxis and precipitous delivery, subsequently impacting timing and dosage of IAP. These limitations are due to factors that the guidelines cannot account for. Therefore, supplementary strategies are required to manage these circumstances. In the future, these limitations could potentially be addressed with maternal immunization.  If used in conjunction with screening and IAP, maternal immunization would be  131 a beneficial strategy in the prevention of both early- and late-onset GBS and in reduction of adverse neonatal outcomes; a vaccine could prevent cases missed due to false negative laboratory results, precipitous deliveries resulting in inadequate prophylaxis administration, late maternal colonization following screening and preterm births occurring prior to screening. Maternal immunization could also reduce cases occurring due to the gaps in adherence to SOGC guidelines as well as limitations of the guidelines relating to IAP effectiveness. Maternal vaccines may also have the potential to prevent stillbirths and preterm labour related to GBS colonization (Porta & Rizzolo, 2015; Bernardini et al., 2016).   This study has created a strong impetus for research on a larger scale to investigate these findings with improvements to study design and methodology. Firstly, studies assessing the burden of neonatal GBS disease should differentiate between early-and late-onset disease. To accomplish this, the use of different data sources should be considered. Future studies should consider utilizing data from the Discharge Abstract Database and Medical Services Plan, available from Population Data BC. These data sources collect information on all infants, including infants aged 0-89 days old. Variables that are collected include diagnostic codes, which can be used to identify neonatal GBS cases, as well as service dates or hospital admission dates, which can be used as proxies for date of illness onset. Data relating to service dates can also be used as proxies to determine dates associated with maternal GBS screening and address questions related to adherence to the GBS guidelines. A retrospective chart review in more hospitals is necessary to accurately assess implementation of guidelines, limitations of the guidelines and missed opportunities for prevention due to gaps in adherences among practices in British Columbia. Future studies should also investigate factors associated with IAP failure to help guide future  132 iterations of the SOGC guidelines. Additional research is also required to assess reasons for regional differences in incidence rates, maternal screening and antibiotic administration. Economic studies are also warranted to assess supplementary or alternate prevention modalities such as rapid diagnostic testing during labour and maternal vaccination. In addition, to improve clinical practice guidelines, health professionals should be surveyed to identify barriers to guideline adherence and reasons for nonadherence.  Future study of the effectiveness of the neonatal GBS prevention program in British Columbia would also benefit from modification to the variables related to GBS prevention in the PDR. Specifically, as GBS remains one of the leading infectious cause associated with neonatal morbidity and mortality rates, the addition of GBS related variables to the PDR would further support PSBC’s objective of improving the health outcomes in women and newborns. Such variables include date of maternal GBS screening, date of GBS onset, specification of type of antibiotic or reason for antibiotic use at delivery as well as cause of death and presence of GBS bacteriuria during pregnancy.          133 Chapter 6: Conclusion  Significant progress has been made in reducing the burden of neonatal group B streptococcal disease in high-income countries. Despite this success, GBS remains a leading pathogen in neonatal morbidity and mortality. The main objective of this study was to quantify the burden of illness, frequency of screening, prevalence of maternal colonization, risk factors for neonatal GBS, as well as examine the implementation of the guidelines, and identify their limitations within a Canadian setting. The overall goal was to facilitate a discussion about the future of GBS prevention and the need for additional clinical measures to further reduce neonatal disease caused by this pathogen. This research was accomplished using data from the Perinatal Data Registry and researcher collected data from a retrospective chart review as well as applying quantitative methods such as Poisson and multiple logistic regression. This study was the first to use data from the Perinatal Data Registry to quantify the burden of neonatal GBS disease and examine the implementation of the SOGC preventive guidelines in British Columbia.   The first project objective found a stable provincial incidence rate of GBS disease among neonates aged 0-28 days over a 10-year period with an attendant low mortality. This study also found numerous risk factors that were significantly associated with increasing the odds ratio for neonatal GBS disease. These include GBS colonization during pregnancy, unknown colonization status, prolonged rupture of membranes, spontaneous labour, emergency caesarean delivery, younger maternal age and preterm birth. In addition, this study found that antenatal screening for GBS in BC increased over time, suggesting improvements in adherence to universal GBS screening recommendations. There was also a high at-risk population in BC compared to other Canadian provinces, predominantly associated with GBS colonization during pregnancy. An  134 additional finding was an increase in antibiotic administration among at-risk parturients in BC. This study found prolonged rupture of membranes and preterm birth were associated with false negative results; regional differences were also observed in regard to false negative results. This study also entailed a chart review of neonatal GBS cases and their mothers at a large tertiary care hospital and found that about half of the cases were associated with non-adherence to the screening and prophylaxis guidelines.  Although current guidelines are not expected to prevent all neonatal GBS cases, these findings suggest opportunities for improvement. This study recommends the need of a supplementary preventive approach to further reduce the burden of illness such as rapid GBS diagnostic tests among women presenting in labour who have not been screened during the antepartum period. Recommendations include further study to identify reasons for guideline nonadherence. Lastly, this thesis recommends the addition of GBS related variables to the PDR such as the presence of GBS bacteriuria during pregnancy, the date of maternal GBS screening, the reason for maternal antibiotic use at delivery, and for the infant, date of GBS onset, and cause of death.   This study further contributes to the literature in the field, particularly within a Canadian setting. The findings from this research will aid policy makers in decision-making regarding the need of supplementary preventive approaches. Future studies would benefit from the use of additional data sources such as the Discharge Abstract Database and Medical Services Plan. Future studies using these data sources can subsequently distinguish early- and late-onset GBS cases and include cases occurring from birth up to 89 days of age. Economic studies are warranted to assess the cost effectiveness of supplementary prevention measures.   135  In summary, the prevention of neonatal GBS disease remains a complex problem influenced by various factors relating to the organism’s virulence and pathogenesis, population characteristics as well as limitations of prevention guidelines involving false-negative culture results and GBS strain resistance to antibiotic agents (Vornhagen & Waldorf & Rajagopal, 2017). Current prevention strategies are incapable of eliminating the burden of neonatal GBS disease. Supplementary preventive measures such as additional screening should be considered in specific circumstances to prevent residual neonatal GBS cases. The future of GBS prevention will require targeted interventions. Maternal immunization remains the ultimate preventive strategy for both developing and developed countries, with the potential to supplement guideline limitations. This strategy will require a compendium of efforts including advocacy from women and health providers and appropriate implementation.                  136 References  Adair, C. E., Kowalsky, L., Quon, H., Ma, D., Stoffman, J., McGeer, A., … & Davies, H. D. (2003). Risk factors for early-onset group B streptococcal disease in neonates: a population-based case–control study. Canadian Medical Association Journal, 169(3), 198–203.  Adams, W. G., Kinney, J. S., Schuchat, A., Collier, C. L., Papasian, C. J., Kilbride, H. W., … & Broome, C. V. (1993). Outbreak of early-onset Group B Streptococcal sepsis. Pediatr Infect Dis J, 12(7), 565–570.  Aharoni, A., Potasman, I., Levitan, Z., Golan, D., & Sharf, M. (1990). Postpartum maternal group B streptococcal meningitis. Review of Infectious Diseases, 12(2), 273–276.   Ahmadzia, H. K., & Heine, R. P. (2014). Diagnosis and management of group B streptococcus in pregnancy. Obstetrics and Gynecology Clinics of North America, 41(4), 629–647.   Alfa, M. J., Sepehri, S., De Gagne, P., Helawa, M., Sandhu, G., & Harding, G. K. M. (2010). Real-time PCR assay provides reliable assessment of intrapartum carriage of group B Streptococcus. Journal of Clinical Microbiology, 48(9), 3095–3099.   Alhhazmi, A., Hurteau, D., & Tyrrell, G. J. (2016). Epidemiology of invasive group B streptococcal disease in Alberta, Canada, from 2003 to 2013. Journal of clinical microbiology, 54(7), 1774-1781.  Allen, U., Nimrod, C., MacDonald, N., Toye, B., Stephens, D., & Marchessault, V. (1999). Relationship between antenatal group B streptococcal vaginal colonization and premature labour. Paediatrics & child health, 4(7), 465-469.  Baker, C. J. (2013). The spectrum of perinatal group B streptococcal disease. Vaccine, D3–D6.   Baker, C. J., & Barrett, F. F. (1973). Transmission of group B streptococci among paturient women and their neonates. The Journal of Pediatrics, 83(6), 919–925.  Baker, C. J., & Barrett, F. F. (1974). Group B streptococcal infections in infants: the importance of the various serotypes. Jama, 230(8), 1158-1160.  Baker, C. J., Barrett, F. F., Gordon, R. C., & Yow, M. D. (1973). Suppurative meningitis due to streptococci of lancefield group B: A study of 33 infants. The Journal of Pediatrics, 82(4), 724–729.   Baker, C. J., Carey, V. J., Rench, M. A., Edwards, M. S., Hillier, S. L., Kasper, D. L., & Platt, R. (2014). Maternal antibody at delivery protects neonates from early onset group B streptococcal disease. Journal of Infectious Diseases, 209(5), 781–788.   Baker, C., Carey, V., Edwards, M., Ewell, M., Ferrieri, P., Ferris, D., … & Hillier, S. (2009). Women receiving group B Streptococcus serotype III tetanus toxoid (GBS III-TT) vaccine have  137 reduced vaginal and rectal acquisition of GBS type III. In Annual meeting of the Infectious Diseases Society of America, Philadelphia, Pennsylvania.  Baker, C. J., & Edwards, M. S. (2003). Group B streptococcal conjugate vaccines. Archives of Disease in Childhood, 88(5), 375–378.   Baker, C. J., Edwards, M. S., & Kasper, D. L. (1978). Immunogenicity of polysaccharides from type III, group B Streptococcus. Journal of Clinical Investigation, 61(4), 1107–1110.   Baker, C., Goroff, D., Alpert, S., Crockett, V. A., Zinner, S. H., Evrard, J. R., … McCormack, W. M. (1977). Vaginal colonization with group B streptococcus: a study in college women. Journal of Infectious Diseases, 135(3), 392–397.   Baker, C. J., Paoletti, L. C., Rench, M. a, Guttormsen, H. K., Carey, V. J., Hickman, M. E., & Kasper, D. L. (2000). Use of capsular polysaccharide-tetanus toxoid conjugate vaccine for type II group B Streptococcus in healthy women. Journal of Infectious Diseases, 182(4), 1129–38.   Baker, C. J., Paoletti, L. C., Wessels, M. R., Guttormsen, H. K., Rench, M. a, Hickman, M. E., & Kasper, D. L. (1999). Safety and immunogenicity of capsular polysaccharide-tetanus toxoid conjugate vaccines for group B streptococcal types Ia and Ib. Journal of Infectious Diseases, 179(1), 142–150.   Baker, C. J., Rench, M. A., & McInnes, P. (2003). Immunization of pregnant women with group B streptococcal type III capsular polysaccharide-tetanus toxoid conjugate vaccine. Vaccine, 21(24), 3468-3472.  BC Centre for Disease Control and Prevention. (2012). Burden of Neonatal Group B Streptococcal Disease in British Columbia. Unpublished raw data.    Beardsall, K., Thompson, M. H., & Mulla, R. J. (2000). Neonatal group B streptococcal infection in South Bedfordshire, 1993-1998. Archives of Disease in Childhood - Fetal and Neonatal Edition, 82(3), F205–F207.  Bedford, H., Louvois, J. De, Halket, S., Peckham, C., Hurley, R., & Harvey, D. (2001). Meningitis in infancy in England and Wales: follow up at age 5 years. BMJ, 323(September), 1–5.  Bennett, D. A. (2001). How can I deal with missing data in my study?. Australian and New Zealand journal of public health, 25(5), 464-469.  Berardi, A., Lugli, L., Rossi, C., China, M., Chiossi, C., Gambini, L., ... & Ferrari, F. (2011). Intrapartum antibiotic prophylaxis failure and group-B streptococcus early-onset disease. The Journal of Maternal-Fetal & Neonatal Medicine, 24(10), 1221-1224.  Berardi, A., Rossi, C., Lugli, L., Creti, R., Bacchi Reggiani, M. L., Lanari, M., … &Ferrari, F. (2013). Group B Streptococcus Late-Onset Disease: 2003-2010. Pediatrics, 131(2), e361–e368.   138 Bernardini, R., Aufieri, R., Detcheva, A., Recchia, S., Cicconi, R., Amicosante, M., ... & Orefici, G. (2016). Neonatal protection and preterm birth reduction following maternal group B streptococcus vaccination in a mouse model. The Journal of Maternal-Fetal & Neonatal Medicine, 1-7.  Bianco, A., Larosa, E., Pileggi, C., Pavia, M., & Collaborative Working Group. (2016). Appropriateness of Intrapartum Antibiotic Prophylaxis to Prevent Neonatal Group B Streptococcus Disease. PloS one, 11(11), e0166179.  Bienenfeld, S., Rodriguez-Riesco, L. G., & Heyborne, K. D. (2016). Avoiding Inadequate Intrapartum Antibiotic Prophylaxis for Group B Streptococci. Obstetrics & Gynecology, 128(3), 598-603.  Blaschke, A. J., Pulver, L. S., Korgenski, E. K., Savitz, L. A., Daly, J. A., & Byington, C. L. (2010). Clindamycin-resistant group B Streptococcus and failure of intrapartum prophylaxis to prevent early-onset disease. The Journal of pediatrics, 156(3), 501-503.  Bliss, S. J., Manning, S. D., Tallman, P., Baker, C. J., Pearlman, M. D., Marrs, C. F., & Foxman, B. (2002). Group B Streptococcus colonization in male and nonpregnant female university students: a cross-sectional prevalence study. Clinical Infectious Diseases : An Official Publication of the Infectious Diseases Society of America, 34(2), 184–190.   Bloom, S. L., Cox, S. M., Bawdon, R. E., & Gilstrap, L. C. (1996). Ampicillin for neonatal group B streptococcal prophylaxis: How rapidly can bactericidal concentrations be achieved? American Journal of Obstetrics and Gynecology, 175(4), 974–976.   Blumberg, H. M., Stephens, D. S., Modansky, M., Erwin, M., Elliot, J., Facklam, R. R., ... & Farley, M. M. (1996). Invasive group B streptococcal disease: the emergence of serotype V. Journal of Infectious Diseases, 173(2), 365-373.  Bobitt, J. R., & Ledger, W. J. (1978). Amniotic Fluid Analysis. Its role in maternal and neonatal infection. Obstetrics and Gynecology, 51(1), 56–62.  Bodner-Adler, B., Bodner, K., Kimberger, O., Lozanov, P., Husslein, P., & Mayerhofer, K. (2004). Influence of the birth attendant on maternal and neonatal outcomes during normal vaginal delivery: a comparison between midwife and physician management. Wiener Klinische Wochenschrift, 116(11-12), 379-384.  Boyer, K. M., Gadzala, C. A., Burd, L. I., Fisher, D. E., Paton, J. B., & Gotoff, S. P. (1983). Selective intrapartum chemoprophylaxis of neonatal group B streptococcal early-onset disease. I. Epidemiologic rationale. Journal of Infectious Diseases, 148(5), 795–801.   Boyer, K. M., Gadzala, C. A., Kelly, Peggy, D., Burd, L. I., & Gotoff, S. P. (1983). Selective intrapartum chemoprophylaxis of neonatal group B streptococcal early-onset disease. II. Predictive Value of Prenatal Cultures. Journal of Infectious Diseases, 148(5), 802–809.    139 Boyer, K. M., Gadzala, C. A., Kelly, Peggy, D., & Gotoff, S. P. (1983). Selective intrapartum chemoprophylaxis of neonatal group B streptococcal early-onset disease. III. Interruption of Mother-to-Infant Transmission. Journal of Infectious Diseases, 148(5), 810–816.   Braun, T. I., Pinover, W., & Sih, P. (1995). Group B Streptococcal Meningitis in a Pregnant Woman Before the Onset of Labor. Clinical Infectious Diseases, 21, 1042–1043.  Buser, G. L. (2017). Notes from the Field: Late-Onset Infant Group B Streptococcus Infection Associated with Maternal Consumption of Capsules Containing Dehydrated Placenta—Oregon, 2016. MMWR. Morbidity and Mortality Weekly Report, 66.  Buuren, S., & Groothuis-Oudshoorn, K. (2011). mice: Multivariate imputation by chained equations in R. Journal of statistical software, 45(3).  Camacho-Gonzalez, A., Spearman, P. W., & Stoll, B. J. (2013). Neonatal infectious diseases: evaluation of neonatal sepsis. Pediatric Clinics of North America, 60(2), 367.  Canadian Institute for Health Information. (2012). International statistical classification of diseases and related health problems, Tenth Revision, Canada (ICD-10-CA). (Vol. 1)  Castor, M. L., Whitney, C. G., Como-Sabetti, K., Facklam, R. R., Ferrieri, P., Bartkus, J. M., ... & Schrag, S. J. (2009). Antibiotic resistance patterns in invasive group B streptococcal isolates. Infectious diseases in obstetrics and gynecology, 2008.  Centers for Disease Control and Prevention (CDC). (1996). Prevention of perinatal group B streptococcal disease: a public health perspective. MMWR. Morbidity and mortality weekly report. Recommendations and Reports. (Vol. 45).   CDC. (2002). Prevention of perinatal group B streptococcal disease. Revised guidelines from CDC. MMWR. Morbidity and mortality weekly report. Recommendations and reports (Vol. 51).  CDC. (2009). Trends in perinatal group B streptococcal disease - United States, 2000-2006. MMWR. Morbidity and mortality weekly report (Vol. 58).   CDC. (2010). Prevention of Perinatal Group B Streptococcal Disease. Revised Guidelines from CDC, 2010. MMWR. Morbidity and mortality weekly report (Vol. 59).  CDC. (2015). Active Bacterial Core Surveillance (ABCs) Report Emerging Infections Program Network Group B Streptococcus, provisional, 2014. Retrieved May 2016, from https://www.cdc.gov/abcs/reports-findings/survreports/gbs14.pdf     Chambliss, L. R., Daly, C., Medearis, A. L., Ames, M., Kayne, M., & Paul, R. (1992). The role of selection bias in comparing cesarean birth rates between physician and midwifery management. Obstetrics and Gynecology, 80(2), 161-165.  Chen, V. L., Avci, F. Y., & Kasper, D. L. (2013). A maternal vaccine against group B  140 Streptococcus: Past, present, and future. Vaccine, 31(S4), D13–D19.  Colombo, D. F., Lew, J. L., Pedersen, C. A., Johnson, J. R., & Fan-Havard, P. (2006). Optimal timing of ampicillin administration to pregnant women for establishing bactericidal levels in the prophylaxis of Group B Streptococcus. American Journal of Obstetrics and Gynecology, 194(2), 466–470.   Davies, H., Miller, M., Faro, S., Gregson, D., Kehl, S. C., & Jordan, J. A. (2004). Multicenter study of a rapid molecular-based assay for the diagnosis of group B Streptococcus colonization in pregnant women. Clinical Infectious Diseases, 39, 1129–1135.   Davies, D. H., Adair, C. E., Schuchat, A., Low, D. E., Sauve, R. S., & McGeer, A. (2001a). Physicians’ prevention practices and incidence of neonatal group B streptococcal disease in 2 Canadian regions. Canadian Medical Association Journal, 164(4), 479–485.  Davies, H. D., Adair, C., McGeer, A., Ma, D., Robertson, S., Mucenski, M., ... & Baker, C. J. (2001b). Antibodies to capsular polysaccharides of group B Streptococcus in pregnant Canadian women: relationship to colonization status and infection in the neonate. Journal of infectious diseases, 184(3), 285-291.  Davies, H. D., Miller, M. A., Faro, S., Gregson, D., Kehl, S. C., & Jordan, J. A. (2004). Multicenter study of a rapid molecular-based assay for the diagnosis of group B Streptococcus colonization in pregnant women. Clinical Infectious Diseases, 39(8), 1129-1135.  Davies, D. H., Raj, S., Adair, C., Robinson, J., & McGeer, A. (2001c). Population-based active surveillance for neonatal group B streptococcal infections in Alberta, Canada: implications for vaccine formulation. Pediatr Infect Dis J, 20(9), 860–863.  Davis, R. L., Hasselquist, M. B., Cardenas, V., Zerr, D. M., Kramer, J., Zavitkovsky, A., & Schuchat, A. (2001). Introduction of the new Centers for Disease Control and Prevention group B streptococcal prevention guideline at a large West Coast health maintenance organization. American Journal of Obstetrics and Gynecology, 184(4), 603–610.   De Cueto, M., Sanchez, M. J., Sampedro, A., Miranda, J. A., Herruzo, A. J., & Rosa-Fraile, M. (1998). Timing of intrapartum ampicillin and prevention of vertical transmission of group B streptococcus. Obstetrics & Gynecology, 91(1), 97-102.  Denny, F. W. (2000) History of Hemolytic Streptococci and Associated Diseases. In D. L. Stevens & E. L. Kaplan (eds.), Streptococcal Infections. Clinical Aspects, Microbiology and Molecular Pathogenesis (pp.1-18). New York, NY: Oxford University Press, Inc.   Deutscher, M., Lewis, M., Zell, E. R., Taylor Jr, T. H., Van Beneden, C., & Schrag, S. (2011). Incidence and Severity of Invasive Streptococcus pneumoniae, Group A Streptococcus, and Group B Streptococcus Infections Among Pregnant and Postpartum Women. Clinical infectious diseases, 53(2), 114-123.   141 Dillon, H. C., Gray, E., Pass, M. a, & Gray, B. M. (1982). Anorectal and vaginal carriage of group B streptococci during pregnancy. Journal of Infectious Diseases, 145(6), 794–799.   Dillon, H. C., Khare, S., & Gray, B. M. (1987). Group B streptococcal carriage and disease: A 6-year prospective study. The Journal of Pediatrics, 110(1), 31–36.   Donders, G. G., Halperin, S. A., Devlieger, R., Baker, S., Forte, P., Wittke, F., … Dull, P. M. (2016). Maternal Immunization With an Investigational Trivalent Group B Streptococcal Vaccine A Randomized Controlled Trial. Obstet Gynecol, 127(2), 213–21.   Dong, Y., & Peng, C.Y.J. (2013). Principled missing data methods for researchers. SpringerPlus, 2(1), 222.  Doran, K. S., Chang, J. C. W., Benoit, V. M., Eckmann, L., & Nizet, V. (2002). Group B streptococcal beta-hemolysin/cytolysin promotes invasion of human lung epithelial cells and the release of interleukin-8. Journal of Infectious Diseases, 185(2), 196–203.   Easmon, C. S., Hastings, M. J., Clare, A. J., Bloxham, B., Marwood, R., Rivers, R. P., & Stringer, J. (1981). Nosocomial transmission of group B streptococci. British Medical Journal, 283, 459–461.   Easmon, C. S., Hastings, M. J., Deeley, J., Bloxham, B., Rivers, R. P., & Marwood, R. (1983). The effect of intrapartum chemoprophylaxis on the vertical transmission of group B streptococci. British Journal of Obstetrics and Gynaecology, 90(7), 633–635.  Edmond, K. M., Kortsalioudaki, C., Scott, S., Schrag, S. J., Zaidi, A. K., Cousens, S., & Heath, P. T. (2012). Group B streptococcal disease in infants aged younger than 3 months: Systematic review and meta-analysis. The Lancet, 379(9815), 547–556.   Edwards, M. S. (2008). Group B streptococcal conjugate vaccine. A timely concept for which the time has come. Human Vaccines, 4(6), 444–448.  Edwards, M. S., Nizet, V., & Baker, C. J. (2016). Group B Streptococcal Infections. In C. B. Wilson, V. Nizet, Y. A. Maldonado, J. S. Remington & J. O. Klein (eds.), Remington and Klein’s Infectious Diseases of the Fetus and Newborn Infant (pp.411-456). Philadelphia, PA: Elsevier Saunders.    El Aila, N. A., Tency, I., Claeys, G., Saerens, B., Cools, P., Verstraelen, H., ... & Vaneechoutte, M. (2010). Comparison of different sampling techniques and of different culture methods for detection of group B streptococcus carriage in pregnant women. BMC infectious diseases, 10(1), 285.  Ferrieri, P., & Blair, L. L. (1977). Pharyngeal carriage of group B streptococci : Detection by Three Methods. Journal of Clinical Microbiology, 6(2), 136–139.  Fossieck, B., & Parker, R. H. (1974). Neurotoxicity during intravenous infusion of penicillin. A  142 review. Journal of Clinical Pharmacology, 14(10), 504–12.   Franciosi, R., Knostman, J., & Zimmerman, R. (1973). Group B streptococcal neonatal and infant infections. The Journal of Pediatrics, 82(4), 707–718.   Freeman, R. M., Ingram, D. L., Gross, I., Ehrenkrenz, R. A., Warshaw, J. B., & Baltimore, R. S. (1981). A half century of neonatal sepsis at Yale. Am J Dis Child, 135, 140.  Fry, R. M. (1938). Fatal Injections by Haemolytic Streptococcus Group B. The Lancet, 231(5969), 199–201.  Gardner, S. E., Yow, M. D., Leeds, L. J., Thompson, P. K., Mason, E. O., & Clark, D. J. (1979). Failure of penicillin to eradicate group B streptococcal colonization in the pregnant woman. Am J Obstet Gynecol, 135(8), 1062-1065.  Garland, S. M., & Fliegner, J. R. (1991). Group B streptococcus (GBS) and neonatal infections: the case for intrapartum chemoprophylaxis. The Australian & New Zealand Journal of Obstetrics & Gynaecology, 31(2), 119–122.   Gibson, R. L., Nizet, V., & Rubens, C. E. (1999). Group B streptococcal β-hemolysin promotes injury of lung microvascular endothelial cells. Pediatric research, 45, 626-634  Godambe, S., Shah, P. S., & Shah, V. (2005). Breast Milk As a source of late onset neonatal sepsis. The Pediatric Infectious Disease Journal, 24(4), 375–377.   Goins, W. P., Talbot, T. R., Schaffner, W., Edwards, K. M., Craig, A. S., Schrag, S. J., ... & Griffin, M. R. (2010). Adherence to perinatal group B streptococcal prevention guidelines. Obstetrics and gynecology, 115(6), 1217.  Hall, R. T., Barnes, W., Krishnan, L., Harris, D. J., Rhodes, P. G., Fayez, J., & Miller, G. L. (1976). Antibiotic treatment of parturient women colonized with group B streptococci. American journal of obstetrics and gynecology, 124(6), 630-634.  Hamada, S., Vearncombe, M., McGeer, A., & Shah, P. S. (2008). Neonatal group B streptococcal disease: incidence, presentation, and mortality. The Journal of Maternal-Fetal & Neonatal Medicine, 21(1), 53–57.   Hannah, M. E., Ohlsson, A., Wang, E. E., Matlow, A., Foster, G. A., Willan, A. R., ... & TermProm Study Group. (1997). Maternal colonization with group B Streptococcus and prelabor rupture of membranes at term: the role of induction of labor. American journal of obstetrics and gynecology, 177(4), 780-785.  Hanson, L., & VandeVusse, L. (2010). Group B streptococcus intrapartum prophylaxis guidelines adherence: A perinatal risk management issue. The Journal of perinatal & neonatal nursing, 24(2), 100-103.   143 Harrison, L. H., Dwyer, D. M., & Johnson, J. A. (1995). Emergence of Serotype V group Group B Streptococcal Infection among Infants and Adults.  Journal of Infectious Diseases, 171(2), 513–513.   Harrison, L. H., Elliott, J. A., Dwyer, D. M., Libonati, J. P., Ferrieri, P., Billmann, L., ... & Maryland Emerging Infections Program. (1998). Serotype distribution of invasive group B streptococcal isolates in Maryland: implications for vaccine formulation. Journal of Infectious Diseases, 177(4), 998-1002.  Heath, P. T. (2011). An update on vaccination against group B streptococcus. Expert Review of Vaccines, 10(5), 685–694.   Heath, P. T., Balfour, G. F., Tighe, H., Verlander, N. Q., Lamagni, T. L., & Efstratiou,  a. (2009). Group B streptococcal disease in infants: a case control study. Archives of Disease in Childhood, 94(9), 674–80.   Heath, P. T., Balfour, G., Weisner, A. M., Efstratiou, A., Lamagni, T. L., Tighe, H., … Group, P. G. B. S. W. (2004). Group B streptococcal disease in UK and Irish infants younger than 90 days. The Lancet, 363(9405), 292–294.   Heath, P. T., & Schuchat, A. (2007). Perinatal group B streptococcal disease. Best Practice & Research Clinical Obstetrics & Gynaecology, 21(3), 411–424.   Hensler, M. E., Miyamoto, S., & Nizet, V. (2008). Group B streptococcal B-Hemolysin/cytolysin directly impairs cardiomyocyte viability and function. PLoS ONE, 3(6), e2446.   Hickman, M. E., Rench, M. A., Ferrieri, P., & Baker, C. J. (1999). Changing epidemiology of group B streptococcal colonization. Pediatrics, 104(2), 203-209.  James, D. C. (2001). Maternal Screening and Treatment for Group B Streptococcus. Journal of Obstetrics, Gynecologic and Neonatal Nursing, 30(6), 659–666.  Jefferies, A. L. (2017). Management of term infants at increased risk for early-onset bacterial sepsis. Paediatrics & Child Health, 22(4), 223-228.  Jordan, H. T., Farley, M. M., Craig, A., Mohle-Boetani, J., Harrison, L. H., Petit, S., … Schrag, S. J. (2008). Revisiting the need for vaccine prevention of late-onset neonatal group B streptococcal disease: a multistate, population-based analysis. The Pediatric Infectious Disease Journal, 27(12), 1057–64.   Joshi, A. K., Chen, C. I., & Turnell, R. W. (1987). Prevalence and significance of group B Streptococcus in a large obstetric population. CMAJ: Canadian Medical Association Journal, 137(3), 209.  Jurgens, D., Sterzik, B., & Fehrenbach, F. J. (1987). Unspecific binding of group B streptococcal cocytolysin (CAMP factor) to immunoglobulins and its possible role in pathogenicity. The  144 Journal of Experimental Medicine, 165(3), 720–732.  Kasper, D. L., Paoletti, L. C., Wessels, M. R., Guttormsen, H. K., Carey, V. J., Jennings, H. J., & Baker, C. J. (1996). Immune response to type III group B streptococcal polysaccharide-tetanus toxoid conjugate vaccine. The Journal of Clinical Investigation, 98(10), 2308–14.   Kenyon, S. L., Taylor, D. J., & Tarnow-Mordi, W. (2001). Broad-spectrum antibiotics for spontaneous preterm labour: The ORACLE II randomised trial. Lancet, 357(9261), 989–994.   Kobayashi, M., Vekemans, J., Baker, C. J., Ratner, A. J., Le Doare, K., & Schrag, S. J. (2016). Group B Streptococcus vaccine development: present status and future considerations, with emphasis on perspectives for low and middle income countries. F1000Research, 5.  Koenig, J. M., & Keenan, W. J. (2009). Group B Streptococcus and Early-Onset Sepsis in the Era of Maternal Prophylaxis. Pediatric Clinics of North America, 56(3), 689–708.   Konrad, G., & Katz, A. (2007). Epidemiology of early-onset neonatal group B streptococcal infection. Implication for screening. Canadian Family Physician • Le Médecin de Famille Canadien, 53, 1054–1055.   Kotiw, M., Zhang, G. W., Daggard, G., Reiss-Levy, E., Tapsall, J. W., & Numa, A. (2003). Late-onset and recurrent neonatal Group B streptococcal disease associated with breast-milk transmission. Pediatric and Developmental Pathology : The Official Journal of the Society for Pediatric Pathology and the Paediatric Pathology Society, 6(3), 251–6.   Kwatra, G., Cunnington, M. C., Merrall, E., Adrian, P. V., Ip, M., Klugman, K. P., ... & Madhi, S. A. (2016). Prevalence of maternal colonisation with group B streptococcus: a systematic review and meta-analysis. The Lancet Infectious Diseases, 16(9), 1076-1084.  Lachenauer, C. S., Kasper, D. L., Shimada, J., Ichiman, Y., Ohtsuka, H., Kaku, M., … Madoff, L. C. (1999). Serotypes VI and VIII predominate among group B streptococci isolated from pregnant Japanese women. Journal of Infectious Diseases, 179(4), 1030–3.   Lamagni, T. L., Keshishian, C., Efstratiou, A., Guy, R., Henderson, K. L., Broughton, K., & Sheridan, E. (2013). Emerging Trends in the Epidemiology of Invasive Group B Streptococcal Disease in England and Wales, 1991-2010. Clinical Infectious Diseases, 57(5), 682–688.  Lancefield, R. C., & Hare, R. (1935). The serological differentiation of pathogenic and non-pathogenic strains of hemolytic streptococci from parturient women. The Journal of Experimental Medicine, 61(3), 335.  Le Doare, K., & Heath, P. T. (2013). An overview of global GBS epidemiology. Vaccine, 31(S4), D7–D12.   Leroux-Roels, G., Maes, C., Willekens, J., De Boever, F., de Rooij, R., Martell, L., ... & Dull, P. (2016). A randomized, observer-blind Phase Ib study to identify formulations and vaccine  145 schedules of a trivalent Group B Streptococcus vaccine for use in non-pregnant and pregnant women. Vaccine, 34(15), 1786-1791.  Lim, D. V, Morales, W. J., Walsh, A. F., & Kazanis, D. (1986). Reduction of morbidity and mortality rates for neonatal group B streptococcal disease through early diagnosis and chemoprophylaxis. J Clin Microbiol, 23(3), 489–492.   Lin, F. Y. C., Brenner, R. A., Johnson, Y. R., Azimi, P. H., Philips, J. B., Regan, J. A., … Clemens, J. D. (2001). The effectiveness of risk-based intrapartum chemoprophylaxis for the prevention of early-onset neonatal group B streptococcal disease. American Journal of Obstetrics and Gynecology, 184(6), 1204–1210.   Lin, F. Y., Clemens, J. D., Azimi, P. H., Regan, J. A., Weisman, L. E., Philips  3rd, J. B., … Ferrieri, P. (1998). Capsular polysaccharide types of group B streptococcal isolates from neonates with early-onset systemic infection. J Infect Dis, 177(3), 790–792.   Lin, F.-Y. C., Weisman, L. E., Troendle, J., & Adams, K. (2003). Prematurity is the major risk factor for late-onset group B streptococcus disease. Journal of Infectious Diseases, 188(2), 267–271.   Lin, W.-J., Walthers, D., Connelly, J., Burnside, K., Jewell, K., Kenney, L. J., & Rajagopal, L. (2009). Threonine phosphorylation prevents promoter DNA binding of the Group B Streptococcus response regulator CovR. Molecular Microbiology, 71(6), 1477–1495.   Main, E. K., & Slagle, T. (2000). Prevention of early-onset invasive neonatal group B streptococcal disease in a private hospital setting: the superiority of culture-based protocols. American journal of obstetrics and gynecology, 182(6), 1344-1354.  Manning, S. D., Neighbors, K., Tallman, P. a, Gillespie, B., Marrs, C. F., Borchardt, S. M., … Foxman, B. (2004). Prevalence of group B streptococcus colonization and potential for transmission by casual contact in healthy young men and women. Clinical Infectious Disease, 39, 380–388.   Marió, M. S., Valenzuela, I., Vásquez, A. E., & Illanes, S. E. (2013). Prevention of early-onset neonatal group B Streptococcal disease. Reviews in obstetrics and gynecology, 6(2), 63.  Marques, M. B., Kasper, D. L., Pangburn, M. K., & Wessels, M. R. (1992). Prevention of C3 deposition by capsular polysaccharide is a virulence mechanism of type III group B streptococci. Infection and Immunity, 60(10), 3986–3993.  Matorras, R., García-Perea, A., Omeñaca, F., Diez-Enciso, M., Madero, R., & Usandizaga, J. A. (1991). Intrapartum chemoprophylaxis of early-onset group B streptococcal disease. European Journal of Obstetrics and Gynecology and Reproductive Biology, 40(1), 57–62.    146 McDonald, H. M., & Chambers, H. M. (2000). Intrauterine infection and spontaneous midgestation abortion: is the spectrum of microorganisms similar to that in preterm labor?. Infectious diseases in obstetrics and gynecology, 8(5-6), 220-227.  Melin, P. (2011). Neonatal group B streptococcal disease: From pathogenesis to preventive strategies. Clinical Microbiology and Infection, 17(9), 1294–1303.   Mifsud, A. J., Efstratiou, A., Charlett, A., & McCartney, A. C. (2004). Early-onset neonatal group B streptococcal infection in London: 1990-1999. BJOG: An International Journal of Obstetrics and Gynaecology, 111(9), 1006–1011.   Mohle-Boetani, J. C., Schuchat, A., Plikaytis, B. D., Smith, D., & Broome, C. V. (1993). Comparison of prevention strategies for neonatal group B streptococcal infection: A population-based economic analysis. Journal of the American Medical Association, 270(12), 1442–1448.  Money, D., Dobson, S., Cole, L., Karacabeyli, E., Blondel-hill, E., Milner, R., & Thomas, E. (2008). An Evaluation of a Rapid Real Time Polymerase Chain Reaction Assay for Detection of Group B Streptococcus as Part of a Neonatal Group B Streptococcus Prevention Strategy. Journal of Obstetrics and Gynaecology Canada, 30(9), 770–775.   Moses, L. M., Heath, P. T., Wilkinson,  a R., Jeffery, H. E., & Isaacs, D. (1998). Early onset group B streptococcal neonatal infection in Oxford 1985-96. Archives of Disease in Childhood - Fetal and Neonatal Edition, 79(2), F148–9.  Mullaney, D. M. (2001). Group B Streptococcal Infections in Newborns. Journal of Obstetrics, Gynecologic and Neonatal Nursing, 30(6), 649–658.  Murdoch, D. R., & Reller, L. B. (2001). Antimicrobial susceptibilities of group B streptococci isolated from patients with invasive disease: 10-year perspective. Antimicrobial agents and chemotherapy, 45(12), 3623-3624.  Nizet, V. (2002). Streptococcal beta-hemolysins: genetics and role in disease pathogenesis. Trends in Microbiology, 10(12), 575–80.   Nizet, V., Ferrieri, P., & Rubens, C. E. (2000). Molecular Pathogenesis of Group B Streptococcal Disease in Newborns. In D. L. Stevens & E. L. Kaplan (eds.), Streptococcal Infections. Clinical Aspects, Microbiology and Molecular Pathogenesis (pp.180- 182). New York, NY: Oxford University Press, Inc.   Nizet, V., Gibson, R. L., Chi, E. Y., Framson, P. E., Hulse, M., & Rubens, C. E. (1996). Group B streptococcal beta-hemolysin expression is associated with injury of lung epithelial cells. Infection and Immunity, 64(9), 3818–26.   Northern Health. (2016). Chief medical health officer’s health status report on child health. Retrieved May 2016, from: https://www.northernhealth.ca/Portals/0/About/Community_Accountability/documents/Northern-Health-CMHO.pdf   147 Olver, W. J., Bond, D. W., Boswell, T. C., & Watkin, S. L. (2000). Neonatal group B streptococcal disease associated with infected breast milk. Archives of Disease in Childhood. Fetal and Neonatal Edition, 83(1), F48–F49.  Our Story. (2017). Retrieved November 2016, from http://www.bcwomens.ca/about/who-we-are/our-unique-role/our-story   Our Unique Role. (2017). Retrieved November 2016, from http://www.bcwomens.ca/about/who-we-are/our-unique-role   Paccione, K. A., & Wiesenfeld, H. C. (2013). Guideline adherence for intrapartum group B streptococci prophylaxis in penicillin-allergic patients. Infectious diseases in obstetrics and gynecology, 2013.  Paoletti, L. C., & Kasper, D. L. (2003). Glycoconjugate vaccines to prevent group B streptococcal infections. Expert Opinion on Biological Therapy, 3(6), 975–84.   Paoletti, L. C., Rench, M. a, Kasper, D. L., Molrine, D., & Baker, C. J. (2001). Effects of Alum Adjuvant or a Booster Dose on Immunogenicity during Clinical Trials of Group B Streptococcal Type III Conjugate Vaccines. Infection and Immunity, 69(11), 6696–6701.   Parriott, A. M., Brown, J. M., & Arah, O. A. (2014). Predischarge postpartum methicillin resistant Staphylococcus aureus infection and group B Streptococcus carriage at the individual and hospital levels. Infectious diseases in obstetrics and gynecology, 2014.  Perinatal Data Registry. (2017).  Retrieved November 2016, from  http://www.perinatalservicesbc.ca/health-professionals/data-surveillance/perinatal-data-registry   Perinatal Services BC. (2014). British Columbia Perinatal Data Registry Reference Manual Version 6.01. Retrieved November 2016, from http://www.perinatalservicesbc.ca/Documents/Data-Surveillance/PDR/Resources-Coders/PDR_ReferenceManual_2014.pdf  Persson, K., Bjerre, B., Elfström, L., Polberger, S., & Forsgren, A. (1986). Group B streptococci at delivery: high count in urine increases risk for neonatal colonization. Scandinavian journal of infectious diseases, 18(6), 525-531.  Phares, C. R., Lynfield, R., Farley, M. M., Mohle-Boetani, J., Harrison, L. H., Petit, S., … Schrag, S. J. (2008). Epidemiology of Invasive Group B Streptococcal Disease in the United States, 1999-2005. JAMA : The Journal of the American Medical Association, 299(17), 2056–2065.  Philipson, E. H., Palermino, D. A., & Robinson, A. (1995). Enhanced antenatal detection of group B streptococcus colonization. Obstetrics & Gynecology, 85(3), 437-439.  Pintye, J., Saltzman, B., Wolf, E., & Crowell, C. S. (2015). Risk Factors for Late-Onset Group B  148 Streptococcal Disease Before and After Implementation of Universal Screening and Intrapartum Antibiotic Prophylaxis. Journal of the Pediatric Infectious Diseases Society, piv067.   Porta, K., & Rizzolo, D. (2015). Preventing group B streptococcal infection in newborns. Journal of the American Academy of Physician Assistant, 28(3), 24–29.   Pulver, L. S., Hopfenbeck, M. M., Young, P. C., Stoddard, G. J., Korgenski, K., Daly, J., & Byington, C. L. (2009). Continued early onset group B streptococcal infections in the era of intrapartum prophylaxis. Journal of Perinatology, 29(1), 20.  Puopolo, K. M., Madoff, L. C., & Eichenwald, E. C. (2005). Early-onset group B streptococcal disease in the era of maternal screening. Pediatrics, 115(5), 1240-1246.  Public Health Agency of Canada. (2009). Case Definitions for Communicable Diseases under National Surveillance CCDR.   Rajagopal, L. (2009). Understanding the regulation of Group B Streptococcal virulence factors. Future Microbiol., 4(2), 201–221.   Rallu, F., Barriga, P., Scrivo, C., Martel-Laferrière, V., & Laferrière, C. (2006). Sensitivities of antigen detection and PCR assays greatly increased compared to that of the standard culture method for screening for group B streptococcus carriage in pregnant women. Journal of clinical microbiology, 44(3), 725-728.  Ramamurthy, R. S., Pyati, S. P., & Pildes, R. S. (1979). Penicillin Prophylaxis For Neonatal Group-B Streptococcal Infection. The Lancet, 314(8136), 246–247.  Regan, J. A., Klebanoff, M. A., & Nugent, R. P. (1991). The Epidemiology of Group B Streptococcal Colonization in Pregnancy. Obstetrics and Gynecology, 77(4), 604–610.  Ring, A., Brraun, J. S., Pohl, J. J., Nizet, V., Stremmel, W., Shenep, J. L., … Shened, T. L. (2002). Group B Streptococcal B-Hemolysin Induces Mortality and Liver Injury in Experimental Sepsis. Journal of Infectious Diseases, 185(12), 1745–1753.  Rouse, D. J., Andrews, W. W., Lin, F. Y. C., Mott, C. W., Ware, J. C., & Philips, J. B. (1998). Antibiotic susceptibility profile of group B streptococcus acquired vertically. Obstetrics & Gynecology, 92(6), 931-934.  Royal College of Obstetricians and Gynaecologists. (2012). The Prevention of Early-onset Neonatal Group B Streptococcal Disease. Green-top Guidelines (No.36). Retrieved September 2017, from: https://www.rcog.org.uk/globalassets/documents/guidelines/gtg_36.pdf   Royston, D., & Geoghegan, F. (1985). Amniotic Fluid Injection with Intact Membranes in Relation to Stillborns. Obstetrics and Gynecology, 65(5), 745–746.  Schafer, J. L. (1999). Multiple imputation: a primer. Statistical methods in medical  149 research, 8(1), 3-15.  Schrag, S. J., & Verani, J. R. (2013). Intrapartum antibiotic prophylaxis for the prevention of perinatal group B streptococcal disease: Experience in the United States and implications for a potential group B streptococcal vaccine. Vaccine, 31(S4), D20–D26.   Schrag, S. J., Zell, E. R., Lynfield, R., Roome, A., Arnold, K. E., Craig, A. S., … Schuchat, A. (2002). A population-based comparison of strategies to prevent early-onset group B streptococcal disease in neonates. The New England Journal of Medicine, 347(4), 233–239. http://doi.org/10.1056/NEJMoa020205  Schrag, S. J., Zywicki, S., Farley, M. M., Reingold, A., Harrison, L. H., Lefkowitx, L. B., … Schuchat, A. (2000). Group B Streptococcal Disease in the Era of Intrapartum Antibiotic Prophylaxis. The New England Journal of Medicine, 342(1), 15–20.  Schubert, A., Zakikhany, K., Schreiner, M., Frank, R., Spellerberg, B., Eikmanns, B. J., & Reinscheid, D. J. (2002). A fibrinogen receptor from group B Streptococcus interacts with fibrinogen by repetitive units with novel ligand binding sites. Molecular Microbiology, 46(2), 557–569.   Schuchat,  A. (1998). Epidemiology of group B streptococcal disease in the United States: shifting paradigms. Clinical Microbiology Reviews, 11(3), 497–513.  Schuchat, A. (1999). Group B Streptococcus. The Lancet, 353(9146), 51–56.  Schuchat, A., Deaver-Robinson, K., Plikaytis, B. D., Zangwill, K., Mohle-Boetani, J., & Wenger, J. D. (1994). Multistate case-control study of maternal risk factors for neonatal Group B Streptococcal disease. Pediatr Infect Dis J, 13(7), 623–629.  Schuchat, A., Oxtoby, M., Cochi, S., Sikes, R., Hightower, A., Plikaytis, B., & Broome, C. (1990). Population-based risk factors for neonatal group B streptococcal disease: results of a cohort study in metropolitan Atlanta. Journal of Infectious Diseases, 162(3), 672–677.  Schuchat, A., & Wenger, J. D. (1994). Epidemiology of Group B Streptococcal Disease. Risk factors, prevention strategies and vaccine development. Epidemiologic Reviews, 16(2), 374–402.  Schuchat, A., Zywicki, S. S., Dinsmoor, M. J., Mercer, B., Romaguera, J., Sullivan, M. J. O., … Levine, O. S. (2000). Risk Factors and Opportunities for Prevention of Early-Onset Neonatal Sepsis: A multicenter Case-Cotnrol study. Pediatrics, 105(1), 21–26.  Seaward, P. G. R., Hannah, M. E., Myhr, T. L., Farine, D., Ohlsson, A., Wang, E. E., & Hodnett, E. (1998). International Multicenter Term PROM Study: Evaluation of predictors of neonatal infection in infants born to patients with premature rupture of membranes at term. American Journal of Epidemiology, 179(3), 635–639  Sherman, K., Whitehead, S., Blondel-hill, E., Wagner, K., & Cheeptham, N. (2012). Penicillin  150 susceptibility and macrolide-lincosamide- streptogramin B resistance in group B Streptococcus isolates from a Canadian hospital. Canadian Journal of Infectious Diseases and Medical Microbiology, 23(4), 196–198.  Society of Obstetricians and Gynaecologists of Canada (SOGC). (1994). National consensus statement on the prevention of early-onset group B streptococcal infections in the newborn. Journal of the Society of Obstetricians and Gynaecologists of Canada and Canadian Pediatric Society, 16, 2271-8  SOGC. (1997). Statement on the prevention of early-onset group B streptococcal infections in the newborn. Journal of Obstetrics and Gynaecology Canada, 19, 751-8.  SOGC. (2004). The prevention of early-onset neonatal group B streptococcal disease. Journal of Obstetrics and Gynaecology Canada, 26(9), 826-832.  SOGC. (2012). Management of group B streptococcal bacteriuria in pregnancy. Journal of Obstetrics and Gynaecology Canada, 34(5), 482-486  SOGC. (2013).The prevention of early-onset neonatal group B streptococcal disease. Journal of Obstetrics and Gynaecology Canada, 35(10), 939-948.  Spaetgens, R., DeBella, K., Ma, D., Robertson, S., Mucenski, M., & Dele Davies, H. (2002). Perinatal antibiotic usage and changes in colonization and resistance rates of group B streptococcus and other pathogens. Obstetrics and Gynecology, 100(3), 525–533.   Stokholm, J., Schjørring, S., Pedersen, L., Bischoff, A. L., Følsgaard, N., Carson, C. G., ... & Bisgaard, H. (2013). Prevalence and predictors of antibiotic administration during pregnancy and birth. PLoS One, 8(12), e82932.  Stoll, B. J., Hansen, N. I., Sánchez, P. J., Faix, R. G., Poindexter, B. B., Van Meurs, K. P., … Higgins, R. D. (2011). Early onset neonatal sepsis: the burden of group B Streptococcal and E. coli disease continues. Pediatrics, 127(5), 817–26.   Stringer, J., Crees-Morris, J. A., Young, S. E. J., & Mayon-White, R. T. (1981). Group B streptococcal systemic disease in Great Britain. Journal of Infection, 3(4), 385–391.  Turrentine, M. A., & Ramirez, M. M. (2008). Recurrence of group B streptococci colonization in subsequent pregnancy. Obstetrics & Gynecology, 112(2, Part 1), 259-264.  Valkenburg-Van Den Berg, A. W., Sprij, A. J., Dekker, F. W., Dörr, P., & Kanhai, H. H. (2009). Association between colonization with Group B Streptococcus and preterm delivery: a systematic review. Acta obstetricia et gynecologica Scandinavica, 88(9), 958-967.  Van Dyke ,MK; Phares, CR; Lynfield, R. et al. (2009). Evaluation of Universal Antenatal Screening for Group B Streptococcus. The New England Journal of Medicine, 360(25), 2626–2636.  151  Verani, J. R., & Schrag, S. J. (2010). Group B streptococcal disease in infants: Progress in prevention and continued challenges. Clinics in Perinatology, 37(2), 375–392.   Vornhagen, J., Waldorf, K. M. A., & Rajagopal, L. (2017). Perinatal Group B Streptococcal Infections: Virulence Factors, Immunity, and Prevention Strategies. Trends in Microbiology.  Watt, J. P., Schuchat, A., Erickson, K., Honig, J. E., Gibbs, R., & Schulkin, J. (2001). Group B streptococcal disease prevention practices of obstetrician-gynecologists. Obstetrics and Gynecology, 98(1), 7–13.  Weisner, A. M., Johnson, A. P., Lamagni, T. L., Arnold, E., Warner, M., Heath, P. T., & Efstratiou, A. (2004). Characterization of group B streptococci recovered from infants with invasive disease in England and Wales. Clinical Infectious Diseases, 38(9), 1203–8.   Wessels, M. R. (1997). Biology of streptococcal capsular polysaccharides. Journal of Applied Microbiology, 83, 20S–31S.  Weston, E. J., Pondo, T., Lewis, M. M., Martell-Cleary, P., Morin, C., Jewell, B., … Schrag, S. J. (2011). The Burden of Invasive Early-Onset Neonatal Sepsis in the United States, 2005-2008. Pediatr Infect Dis J, 30(11), 937–941.  Wood, E. G., & Dillon, H. C. (1981). A prospective study of group B streptococcal bacteriuria in pregnancy. American journal of obstetrics and gynecology, 140(5), 515-520.  World Health Organization. (2016). Preterm Birth. Retrieved November 2016, from http://www.who.int/mediacentre/factsheets/fs363/en/   Yancey, M. K., Duff, P., Clark, P., Kurtzer, T., Frentzen, B. H., & Kubilis, P. (1994). Peripartum Infection Associated with Vaginal Group B Streptococcal Colonization. Obstetrics and Gynecology, 84(5), 816–819.  Yancey, M. K., Schuchat, A., Brown, L. K., Ventura, V. L., & Markenson, G. R. (1996). The Accuracy of Late Antenatal Screening Cultures in Predicting Genital Group B Streptococcal Colonization at Delivery. Obstetrics and Gynecology, 88(5), 811–815.  Young, B. C., Dodge, L. E., Gupta, M., Rhee, J., & Hacker, M. (2011). Evaluation of a rapid, real-time intrapartum group B Streptococcus assay. American Journal of Obstetrics and Gynecology, 205(4), 1–12.   Zaleznik, D., Rench, M., Hillier, S., Krohn, M. A., Platt, R., Lee, M.-L. T., … & Baker, C. J. (2000). Invasive disease due to group B Streptococcus in pregnant women and neonates from diverse population groups. Clinical Infectious Diseases, 30(2), 276–281.  Zhou, X. H., Zhou, C., Lui, D., & Ding, X. (2014). Applied missing data analysis in the health sciences. Hoboken, New Jersey: John Wiley & Sons.   152 Appendix A: Hospital Chart Extraction Form   INSTRUCTIONS This form will be completed through a maternal/infant pair chart review/ abstraction for every mother and infant, who had an ICD-10-CA code of one or more of: A40.1, B95.1, P36.0, G00.2, J15.3, P23.3 and/or had a positive infectious agent culture (blood or other).  Items that are available from the Perinatal Services BC Registry (in italics) will be inserted onto the form and verified against the information on the chart; only items not available from the PDR will be retrieved from the chart.  [TO BE INSERTED BY INVESTIGATOR PRIOR TO FORM COMPLETION] STUDY ID: [RANDOMLY GENERATED NUMBER]  MOTHER’S INFORMATION A. DEMOGRAPHIC INFORMATION Date of Birth (MM/YYYY): ETHNICITY: □ White/Caucasian □ Arab, Middle Eastern or North African (Armenian, Egyptian, Iranian, Lebanese, Moroccan) □ Black (West Indies, Caribbean, African) □ Chinese □ Filipino □ First Nations, Inuit or Métis □ Japanese □ Korean □ Latin American □ South Asian (East Indian, Pakistani, Punjabi, Sri Lankan) □ South East Asian (Cambodian, Indonesian, Laotian, Vietnamese) □ Other, specify: ____________________ □ Unknown           A.1.  Obstetrical History Perinatal Complications: Previous infant with GBS disease Other If other, please specify: B. MATERNAL TESTING  Urine C&S result: GBS Screening (35-37 weeks):  Yes      No    Unknown If yes: Date of screening (DD/MM/YYYY):   Results of Screening:    Positive GBS culture      Negative GBS culture     Unknown  C. LABOUR AND DELIVERY INFANT DELIVERY DATE (DD/MM/YYYY): GESTATIONAL AGE AT BIRTH (WEEKS):  Were any signs or diagnoses compatible with infection present at the time of delivery?    No  Fever  Highest temperature recorded:    Endometritis   Urinary tract infection   Sepsis   Pneumonia  153 Were any signs or diagnoses compatible with infection present at the time of delivery?   Other Please specify:                Were antibiotics administered during delivery admission?   Yes      No     Unknown (coded as Blank) Were intrapartum antibiotics administered for GBS: Yes  If yes, please specify (name, dosage and frequency):   Penicillin G; 5 million units intravenously (IV), followed by 2.5-3.0 million units every 4 hours [until delivery]   Cefazolin, 2g IV (initial dose), followed by 1g IV every 8 hours [until delivery]  Vancomycin, 1 g IV every 12 hours [until delivery]  Clindamycin, 900 mg IV every 8 hours [until delivery]       Other, please specify:      No If no (was a reason provided?):  Was there documentation of offering/ recommendation of antibiotics for GBS?  Yes  No   Was there documentation of mother’s refusal of antibiotics for GBS?                 Yes                  No   If yes, what reason was provided?    Newborn Information ICD-10-CA Code:  Date of Onset of illness compatible with GBS (DD/MM/YYYY): Date of Specimen Collection (DD/MM/YYYY): Date of Laboratory Results (DD/MM/YYYY): Specimens positive for GBS (check all that apply):  Culture/ isolation from  blood  cerebrospinal fluid  surface/ skin swab (specify):    DNA / PCR detection from  blood  cerebrospinal fluid  Antigen detection from  blood  cerebrospinal fluid  Other, please specify:  Outcome of Hospitalization: Survived Died     Was GBS cause of death?  Yes    No  Not specified Unknown Date completed (MM/DD/YYYY):              154 Appendix B: Inclusion Criteria Diagram                                   Maternal and neonatal records from April 1st 2004 through December 2014 466 083 Eligible maternal and neonatal records  463 384  Final maternal and neonatal records 463 376 Multiples (>2)                                                    279  Out of province/country maternal residence     2413 Neonatal sex classified as                                  7                “Unknown” or “Other”  Blank values for maternal                        4 GBS screening variable   Blank values for provider type                4 variable   155 Appendix C: Algorithm   Figure C.1 Algorithm for the identification of at-risk parturients.     Abbreviations: GBS= Group B streptococcus, PROM= prolonged rupture of membranes               156 Appendix D: Proportion of Parturients with GBS Risk Factors   Figure D.1 Average proportion of parturients with risk factors in British Columbia.       Abbreviations: GBS= Group B streptococcus, PROM= prolonged rupture of membranes   157 Figure D.2 The proportion of parturients with unknown GBS status and preterm birth among parturients with risk factors, British Columbia April 1, 2004 through December 31, 2014.****                                                 **** Proportions for 2004 are based on data from April 1st 2004 through December 31st 2004 Abbreviations: FHA= Fraser Health Authority, IHA= Interior Health Authority, NHA= Northern Health Authority, VCH= Vancouver Coastal Health Authority, VIHA= Vancouver Island Health Authority  2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014FHA 82.2 86.8 90.2 91.0 90.5 93.4 93.4 93.7 95.0 95.2 92.8IHA 82.0 87.7 89.7 90.5 91.2 91.0 86.3 91.7 87.4 89.2 89.9NHA 67.9 66.9 73.9 79.6 77.6 79.9 84.7 86.1 84.6 87.1 83.6VCH 78.4 84.0 85.5 86.6 88.3 92.2 90.3 90.9 88.8 90.0 88.1VIHA 83.3 89.1 88.3 93.5 90.2 89.5 94.1 92.3 92.7 96.4 95.9Provincial 79.8 84.2 87.0 89.0 88.7 90.9 90.8 91.9 91.2 92.5 90.80.020.040.060.080.0100.0Percent 158 Figure D.3 The proportion of parturients with unknown GBS status and prolonged rupture of membranes among parturients with risk factors; British Columbia April 1, 2004 through December 31, 2014.††††                                                  †††† Proportions for 2004 are based on data from April 1st 2004 through December 31st 2004 Abbreviations: FHA= Fraser Health Authority, IHA= Interior Health Authority, NHA= Northern Health Authority, VCH= Vancouver Coastal Health Authority, VIHA= Vancouver Island Health Authority   2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014FHA 17.8 13.2 9.8 9.0 9.5 6.6 6.6 6.3 5.0 4.8 7.2IHA 18.0 12.3 10.3 9.5 8.8 9.0 13.7 8.3 12.6 10.8 10.1NHA 32.1 33.1 26.1 20.4 22.4 20.1 15.3 13.9 15.4 12.9 16.4VCH 21.6 16.0 14.5 13.4 11.7 7.8 9.7 9.1 11.2 10.0 11.9VIHA 16.7 10.9 11.7 6.5 9.8 10.5 5.9 7.7 7.3 3.6 4.1Provincial 20.2 15.8 13.0 11.0 11.3 9.1 9.2 8.1 8.8 7.5 9.20.05.010.015.020.025.030.035.0Percent 159 Appendix E: Proportion of Parturients with Antibiotic Administration  Figure E.1 The proportion of maternal antibiotic administration during labour among GBS colonized parturients from April 1, 2004 through December 31, 2014.‡‡‡‡                                                 ‡‡‡‡ Proportions for 2004 are based on data from April 1st 2004 through December 31st 2004 Abbreviations: FHA= Fraser Health Authority, IHA= Interior Health Authority, NHA= Northern Health Authority, VCH= Vancouver Coastal Health Authority, VIHA= Vancouver Island Health Authority  FHA IHA NHA VCH VIHA Provincial2004 85.2 66.3 82.6 82.9 75.9 80.62005 89.2 63.1 80.2 86.3 83.6 83.52006 89.8 76.1 83.8 85.2 83.7 85.52007 88.3 79 80.6 86.8 86 85.72008 87.4 82.3 79.4 90.7 86.8 86.92009 87.8 83.1 79.7 90.3 86.4 86.92010 88 86.6 82.2 88.4 88.4 87.52011 89.9 89.4 84.2 90.5 88.6 89.32012 88.9 89 79.6 88.5 87.1 87.82013 90.5 88.7 78.9 85.8 87.9 87.82014 91.6 88.9 77.1 85.9 87.1 880102030405060708090100Percent 160 Figure E.2  The proportion of maternal antibiotic administration among parturients with risk factors from April 1, 2004 through December 31, 2014.§§§§                                                  §§§§ Proportions for 2004 are based on data from April 1st 2004 through December 31st 2004 Abbreviations: FHA= Fraser Health Authority, IHA= Interior Health Authority, NHA= Northern Health Authority, VCH= Vancouver Coastal Health Authority, VIHA= Vancouver Island Health Authority  FHA IHA NHA VCH VIHA Provincial2004 69.1 48.4 57.9 60.6 45.2 59.52005 74.1 57.3 57 70.3 56.3 66.92006 76.9 60.6 54.7 72.2 57.4 68.62007 76.4 63.7 60.6 70.2 58.8 69.32008 78.7 74.2 58.9 79.6 70.9 75.22009 78.4 72.4 61.7 81.1 73.6 75.82010 83.6 74.6 60.7 81.3 81.7 79.42011 81.3 81.2 71.5 85.1 83.1 81.72012 85.1 79.8 67.7 82.4 75.1 80.92013 85.3 80.3 74.2 82.1 84.3 82.72014 87 85 65.6 84.2 85.4 840102030405060708090100Percent 161 Appendix F: Multiple Logistic Regression Analysis of Neonatal GBS Disease    Table F.1 Singletons only: multiple logistic regression analysis of neonatal GBS disease  Covariate Adjusted odds ratio 95% Confidence Interval P value GBS results Negative Positive Unknown  Ref 5.32  1.76  - 3.97    7.09 1.17    2.69   <0.001 0.006 Length of rupture of membranes <18 hours ≥18 hours No rupture  Missing  Ref 1.74 0.87 0.65  - 1.23    2.44 0.22    3.42 0.35    1.20   0.001 0.85 0.17 Gestational age 0.84  0.81    0.87 <0.001 Term status Term (³37 weeks) Preterm (<37 weeks)  Ref 3.45  - 2.48   4.85   <0.001 Term status (weeks) Term (³37 weeks) Moderately preterm (32 to 37 weeks) Extremely preterm (<32 weeks)  Ref 2.73 10.6  - 1.86    4.01 6.11   18.1    <0.001 <0.001 Labour type Induced None Spontaneous  Ref 0.76 1.94  - 0.21    2.66 1.35    2.79   0.67 <0.001 Mode of delivery (detailed) Vaginal Elective caesarean Emergency caesarean   Ref 0.63 1.55   - 0.19   2.18 1.12    2.13   0.47 0.007 Maternal age 0.97 0.95   0.99 0.008        162 Table F.2 Twins only: multiple logistic regression analysis of neonatal GBS disease Covariate Adjusted odds ratio  95% Confidence Interval P value GBS results Negative Positive Unknown  Ref 1.63 0.39  - 0.39       6.82 0.10       1.46   0.5 0.16 Length of rupture of membranes <18 hours ≥18 hours No rupture  Missing  Ref 0.74 0.55 0.31  -  0.11        4.80                0.12        2.43  0.03        2.66   0.75 0.43 0.43 Gestational age 0.78 0.69        0.87 <0.001 Term status Term (³37 weeks) Preterm (<37 weeks)  Ref 2.29  - 0.24        21.5   0.46 Term status (weeks) Term (³37 weeks) Moderately preterm (32 to 37 weeks) Extremely preterm (<32 weeks)  Ref 1.11 11.2  - 0.24        5.10 2.55        49.8   0.88 0.001 Mode of delivery Vaginal Caesarean section   Ref 2.03  - 0.54        7.61   0.28         

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0362004/manifest

Comment

Related Items