UBC Faculty Research and Publications

Physician's manual reporting underestimates mortality: evidence from a population-based HIV/AIDS treatment… Au-Yeung, Christopher G; Anema, Aranka; Chan, Keith; Yip, Benita; Montaner, Julio S; Hogg, Robert S Oct 25, 2010

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

Item Metadata


52383-12889_2010_Article_2553.pdf [ 194.46kB ]
JSON: 52383-1.0224012.json
JSON-LD: 52383-1.0224012-ld.json
RDF/XML (Pretty): 52383-1.0224012-rdf.xml
RDF/JSON: 52383-1.0224012-rdf.json
Turtle: 52383-1.0224012-turtle.txt
N-Triples: 52383-1.0224012-rdf-ntriples.txt
Original Record: 52383-1.0224012-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessPhysician’s manual reporting underestimatesmortality: evidence from a population-based HIV/AIDS treatment programChristopher G Au-Yeung1, Aranka Anema1,2, Keith Chan1, Benita Yip1, Julio SG Montaner1,2, Robert S Hogg1,3*AbstractBackground: In clinical and cohort research, mortality estimates are often derived from manual reports generatedby physicians or electronic reports from vital event registries. We examined the rate of underreporting of deaths bymanual methods as compared with electronic reports from a vital event registry.Methods: The retrospective analyses included deaths among participants registered in an observational cohortwho initiated highly-active antiretroviral therapy (HAART) between August 1, 1996 and June 30, 2006. Deaths wereroutinely reported manually by physicians and through annual electronic record linkages with a population-basedvital event registry. Multivariate logistic regression was carried out to assess independent predictors of deathreporting by manual methods.Results: Of the 3,116 individuals included in the analyses, 622 (20.0%) died during follow-up. Manual reporting byphysicians only identified 377 (60.6%), while electronic linkages captured 598 (96.1%) of all deaths. Multivariateanalysis indicated that deaths among individuals with lower CD4 cell count, higher HIV plasma viral load, a historyof injection drug use, and under the care of an HIV-experienced physicians were more likely to be reportedmanually. Furthermore, non-accidental deaths were more likely to be reported manually, and manual reporting ofdeaths increased over time.Conclusions: Relying only on manual reports to ascertain deaths significantly underestimates the total number ofdeaths in the population. This can generate important biases when evaluating the impact of therapeuticinterventions in the populational setting.BackgroundAccurate estimates of mortality are necessary for HIVsurveillance, including assessments of antiretroviral treat-ment programs [1-4]. Several methods can be used torecord deaths in a population, including physician report-ing, vital statistics, and hospital registries [1,2]. Sampling-based approaches, verbal autopsies, morgue and burialdata provide additional proxies, particularly in settingswhere vital event data may not be available [5-9].Reporting methods for HIV deaths vary in terms of theirsensitivity and specificity [10-13]. Autopsy and chartreviews, for example, have better specificity than deathcertificates since they describe the underlying causes ofdeath in detail [3,4]. In clinical and cohort research, mor-tality estimates are often derived from manual reports gen-erated by physicians [10] or electronic reports from vitalevent registries [14,15]. Previous studies of the HIV-posi-tive and HIV-negative population have found varyingdegrees of agreement between deaths reported manuallyby physicians and population census data [16-20].In this analysis, data derived from a population-basedcohort of HIV-infected individuals on HAART in BritishColumbia (BC), Canada, are used to compare the num-ber of deaths reported manually by physicians to thenumber of deaths reported electronically by a vital eventregistry.MethodsThis study is based on HIV-positive men and women atleast 18 years of age in the HAART Observational* Correspondence: bobhogg@cfenet.ubc.ca1British Columbia Centre for Excellence in HIV/AIDS, Vancouver, BritishColumbia, CanadaFull list of author information is available at the end of the articleAu-Yeung et al. BMC Public Health 2010, 10:642http://www.biomedcentral.com/1471-2458/10/642© 2010 Au-Yeung et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.Medical Evaluation and Research (HOMER) cohort.HOMER is a prospective, observation population-basedcohort of individuals enrolled in a province-wide HIV/AIDS Drug Treatment Program (DTP) in BC, Canada.The HOMER cohort includes data on individuals whoentered the DTP antiretroviral-naïve to HAARTbetween August 1, 1996 and June 30, 2006, with follow-up to June 30, 2007.The HIV/AIDS DTP and the HOMER cohort havebeen approved by the University of British Columbia(UBC) Research Ethics Board at its St Paul’s Hospitalsite. Individuals in HOMER were not required to pro-vide written informed consent for the purposes of theanalyses presented herein. These administrative analysesoccur within the context of a universal health care sys-tem in which individuals receive medical care, laboratorymonitoring, and HAART free of charge.If a death occurs among an individual enrolled in theDTP, the HAART prescribing and/or primary care phy-sician is responsible to report the death to the program(manual reporting). A record linkage between the DTPand the BC Vital Statistics death registry is also per-formed annually to capture all deaths among individualsenrolled in the DTP (electronic reporting). In additionto these death data, the HOMER cohort captures clini-cal indicators and socio-demographic informationamong individuals enrolled in the DTP.Analyses included bivariate comparisons of individualswho died and did not die and multivariate logisticregression to examine factors associated with manualmethods of reporting deaths among the deceased. Thefollowing covariates were considered in these analyses:physician practice setting (dichotomized into rural andurban categories according to Statistics Canada Censusdenominations [21,22]); HIV experience among physi-cians (defined as the number of HIV-positive individualsthe physician had previously treated at the time apatient was enrolled into the DTP [5]); calendar year ofdeath, cause of death, age, sex, history of injection druguse, adherence to HAART (defined as the number ofdays of antiretroviral medications dispensed divided bythe number of days of follow-up during first year oftreatment, and expressed as a percent [23]), and baselineCD4 cell count and HIV plasma viral load. CD4 cellcount was recorded using flow cytometry and fluores-cent monoclonal antibody analysis (Beckman Coulter,Inc., Mississauga, Ontario, Canada), and HIV viral loadwas measured using the Roche Amplicor Monitor assay(Roche Diagnostics, Laval, Quebec, Canada) using eitherthe standard method or the ultrasensitive adaptation.All analyses are conducted using SAS version 9.1.3(SAS, Cary, North Carolina, United States). Statisticaltests are two-sided and p-values of less than 0.05 areconsidered statistically significant.Ethical StatementEthical approval was requested and obtained for thisstudy. The HIV/AIDS Drug Treament Program and theHOMER cohort have been approved by the Universityof British Columbia (UBC) Research Ethics Board at itsSt Paul’s Hospital site.ResultsBetween August 1, 1996 and June 30, 2006, 622 (20.0%)deaths occurred among the 3,116 individuals inHOMER. Manual reporting identified 377 (60.6%), whileelectronic linkages captured 598 (96.1%) of all deaths.Manual reporting identified 377 (60.6%), while electro-nic linkages captured 598 (96.1%) of all deaths. Manualreporting alone captured only 24 (3.9%) deaths, whilevital statistics alone captured 245 (39.4%) of all deaths.The remaining 353 (56.8%) deaths were reported byboth sources. Based on a review of the principal causeof death, the vast majority of deaths (535 [86%]) weredeemed non-accidental (i.e. HIV-related (373 [60%]), cir-culatory system-related (35 [5.6%]), malignancy-related(21 [3.4%]), hepatitis-related (19 [3%]), respiratory-system related (15 [2.4%]), digestive system related(14 [2.2%]), according to the International StatisticalClassification of Diseases and Related Health Problemscoding system [24]. The remaining 87 (14.0%) deathswere deemed accidental (i.e. accidental poisoning by andexposure to noxious substances (60 [9.6%]), intentionalself-harm (16 [2.6%]).In bivariate analyses, deaths were more likely to occuramong individuals who were older (median age: 41versus 39; p < 0.001), had a history of injection drug use(37.3% versus 27.4%; p < 0.001), had a less HIV-experi-enced physician (median number of HIV-positivepatients: 30 versus 70; p < 0.001), had an AIDS definingillness at baseline (17.7% versus 14.3%; p = 0.038), had alower CD4 cell count at baseline (median CD4: 140cells/mm3 versus 200 cells/mm3; p < 0.001), had aplasma HIV viral load ≥100,000 copies/mL at baseline(65.9% versus 53.0%; p < 0.001), and had lower adher-ence to HAART (61.7% versus 39.2% of patients whoare < 95% adherent to HAART; p < 0.001).Table 1 shows the results of the multivariate logisticregression examining factors associated with manualmethods of reporting deaths among the deceased.Deaths among those with lower CD4 cell count, higherHIV plasma viral load, a history of injection drug use,and under the care of an HIV-experienced physicianwere more likely to be reported manually. Furthermore,non-accidental deaths were more likely to be reportedmanually, and manual reporting of deaths increasedover time. Of the 245 patients captured by vital statis-tics, 16 (6.5%) were lost to follow-up (not seen over 6months) versus 4(1.1%) of the 377 patients captured byAu-Yeung et al. BMC Public Health 2010, 10:642http://www.biomedcentral.com/1471-2458/10/642Page 2 of 5manual reporting. Using a lost to follow-up definition of“not seen in over 12 months”, vital statistics recorded 6(2.4%) patients versus 1(0.3%) by manual reporting. Thenumber of deaths captured by vital statistics comparedto manual reporting was still significantly different afteraccounting for loss to follow-up (p < 0.05).DiscussionOur results demonstrate that manual methods of deathreporting by physicians underestimated the total numberof deaths in the population by 40%. Annual electroniclinkages with the vital statistics death registry captured96% of deaths, suggesting an improved death registrysensitivity. The remaining 4% of deaths not identifiedthrough this linkage may be attributed to missing dataor discrepancies in the patient identifiers being matched.While electronic linkages identify a high proportion ofthe mortality cases, the disadvantage is that there mustbe accurate, up-to-date demographic information suchas names, birthdates, and health care numbers for thematching process. Even though manual reporting onlycovers about 60% of all deaths in our cohort, any patientidentifier discrepancies can be easily clarified with thephysician office staff members.Our findings are similar to those identified by theUnited States Centres for Disease Control, which foundthat only 54% of deaths had been manually reported tothe District of Columbia between 2000 and 2005 [17].In this study, the higher proportion of deaths capturedby electronic methods as compared to manual methodssuggests that electronic record linkage is essential toaccurately ascertain deaths among persons with HIV[17]. The improved rate of manual reporting in ourstudy could be a consequence of its longer duration(1996-2006) and enhanced physician training for manualreporting over time. Improvements in reporting canenhance the accuracy of HIV prevalence estimates anddistribution of HIV treatment and prevention resourcesin regions with the highest burden [17].Studies from the United States suggest that underre-porting may be due to lack of physician knowledgeabout the administrative process for reporting deaths,and intentional non-reporting to protect patient confi-dentiality [16]. It is promising to note that manuallyreporting of deaths has increased over time in ourcohort. Also of interest, HIV-experienced physicianswere more likely to manually report deaths in our study.This suggests that educational initiatives may be able toenhance manual death reporting when vital registrationsystems are not available or accurate, such as inresource-limited settings. Additional physician or healthworking training for those caring for HIV-infected indi-viduals may be required to effectively quantify the actualnumber of deaths.Access to vital registration records could be an invalu-able resource to effectively ascertain the number ofdeaths among patients enrolled in a population basedHIV/AIDS treatment program. Cohort studies relyingsolely on manually reported data may be significantlyunderestimating actual mortality. In addition, patientsrecorded as lost to follow up may not be captured as adeath by vital registration systems. Thus, this data lim-itation calls for studies to analyze separate outcomesbased on the time to loss to follow up and the time todeath.Our study has some strengths and limitations.Strengths include having a large observational databaseand using physician reported data in conjunction withvital statistics records to obtain socio-demographic andclinical information on patients. However, the HOMERcohort provides varying amounts of detail on eachpatient’s clinical and socio-demographic characteristics.Specifically, patient ethnicity was unknown for morethan half of the 622 deaths. Therefore, we were unableTable 1 Unadjusted and adjusted models showing independent predictors of manual reporting of deathsVariable Unadjusted AdjustedOR (95% CI) OR (95% CI)Physician practice location Urban versus rural 0.79 (0.47 to 1.32) –Physician HIV-experience per 10 patient increase 1.06 (1.01 to 1.11) 1.02 (1.00 to 1.04)Year of death per 1 year increase 1.18 (1.11 to 1.26) 1.17 (1.09 to 1.25)Cause of death Non-accidental versus Accidental 2.23 (1.41 to 3.53) 1.69 (1.03 to 2.78)Age per 10 year increase 0.99 (0.85 to 1.17) –Sex Female versus Male 0.84 (0.56 to 1.27) –History of injection drug use Yes versus No 1.57 (1.12 to 2.20) 1.49 (1.03 to 2.15)Baseline CD4 cell count (Cells/mm3) per 100 cells/mm3increase0.88 (0.81 to 0.95) 0.90 (0.83 to 0.98)Baseline plasma viral load(Log10 copies/mL)per Log10 copies/mL increase 1.29 (0.99 to 1.69) 1.38 (1.04 to 1.85)Adherence ≥ 95% versus < 95% 1.44 (1.03 to 2.02) 1.41 (0.99 to 2.01)OR, odds ratio; CI, confidence interval.Au-Yeung et al. BMC Public Health 2010, 10:642http://www.biomedcentral.com/1471-2458/10/642Page 3 of 5to compare differences between the number of HIVdeaths reported by physicians and the death registrystratified by patient ethnicity. Inclusion of this variablewould have been relevant, given that studies in the Uni-ted States have found disagreement between physicianand vital statistic reporting of deaths by ethnic group[18-20]. We also realize that confounder variables suchas socio-economic status may have influenced ourresults, however, this classification is not made for indi-viduals in the HOMER cohort.ConclusionOur results demonstrate that physicians’ manual HIVdeath reports underestimated mortality among indivi-duals on HAART in BC by 40%. Cohorts relying onlyon physician reported deaths are potentially underre-porting the total number of deaths among HIV-infected individuals. This represents a major limitationfor studies using exclusively physicians’ manual HIVdeath reports to ascertain mortality rates, as this prac-tice can generate important biases when evaluating theimpact of therapeutic interventions in the populationalsetting. The use of both manually reported data sys-tematically cross-checked and supplemented with vitalregistration records is ideal for capturing the majorityof deaths in an observational cohort of HIV-infectedpatients.AcknowledgementsThe authors express their gratitude to Eric Druyts, Svetlana Draskovic,Elizabeth Ferris, Kelly Hsu, and Peter Vann for research and administrativeassistance.Author details1British Columbia Centre for Excellence in HIV/AIDS, Vancouver, BritishColumbia, Canada. 2Faculty of Medicine, University of British Columbia,Vancouver, British Columbia, Canada. 3Faculty of Health Sciences, SimonFraser University, Burnaby, British Columbia, Canada.Authors’ contributionsCGA, AA, KC, BY, JSGM and RSH contributed to the study design, datagathering, interpretation of results and manuscript draft. KC performed theanalysis. CGA edited the manuscript. All authors revised the manuscriptcritically for important intellectual content; and gave final approval of theversion to be submitted.Competing interestsRS Hogg has held grant funding from the National Institutes of Health,Canadian Institutes of Health Research National Health ResearchDevelopment Program, and Health Canada. He has also received fundingfrom Agouron Pharmaceuticals Inc, Boehringer Ingelheim PharmaceuticalsInc, Bristol-Myers Squibb, GlaxoSmithKline, and Merck Frosst Laboratories forparticipating in continued medical education programmes. JSG Montanerhas received grants from, served as an ad hoc advisor to, or spoken atvarious events sponsored by Abbott, Argos Therapeutics, Bioject Inc,Boehringer Ingelheim, BMS, Gilead Sciences, GlaxoSmithKline, Hoffmann-LaRoche, Janssen-Ortho, Merck Frosst, Pfizer, Schering, Serono Inc,TheraTechnologies, Tibotec, Trimeris. He has also held grant funding fromthe Canadian Institutes of Health Research and National Institutes of Health.He has also received funding for research and continuing medical educationprograms from a number of pharmaceutical companies including Abbott,Boehringer Ingelheim, and GlaxoSmithKline.Received: 28 May 2010 Accepted: 25 October 2010Published: 25 October 2010References1. Ajdacic-Gross V, Zellweger U, Wang J, Fleerackers Y, Somaini B: Howcomplete is AIDS surveillance in Europe? An eagle eye comparison withmortality data. J Epidemiol Community Health 2001, 55:52-56.2. Barchielli A, Buiatti E, Galanti C, Giovannetti L, Acciai S, Lazzeri V:Completeness of AIDS reporting and quality of AIDS death certificationin Tuscany (Italy): a linkage study between surveillance system of casesand death certificates. Eur J Epidemiol 1995, 11:513-517.3. Sackoff JE, Hanna DB, Pfeiffer MR, Torian LV: Causes of death amongpersons with AIDS in the era of highly active antiretroviral therapy: NewYork City. Ann Intern Med 2006, 145:397-406.4. Bernicker EH, Atmar RL, Schaffner DL, Greenberg SB: Unanticipateddiagnoses found at autopsy in an urban public teaching hospital. Am JMed Sci 1996, 311:215-220.5. Wood E, Hogg RS, Yip B, Harrigan PR, O’Shaughnessy MV, Montaner JSG: Isthere a baseline CD4 cell count that precludes a survival response tomodern antiretroviral therapy? AIDS 2003, 17:711-720.6. Miller CL, Kerr T, Strathdee SA, Li K, Wood E: Factors associated withpremature mortality among young injection drug users in Vancouver.Harm Reduction Journal 2007, 4:1.7. Wood E, Hogg RS, Yip B, Moore D, Harrigan PR, Montaner JS: Impact ofbaseline viral load and adherence on survival of HIV-infected adults withbaseline CD4 cell counts > or = 200 cells/microl. AIDS 2006, 20:1117-1123.8. Landon BE, Wilson IB, Wenger NS, Cohn SE, Fichtenbaum CJ, Bozzette SA,Shapiro MF, Cleary PD: Specialty training and specialization amongphysicians who treat HIV/AIDS in the United States. J Gen Intern Med2002, 17:12-22.9. Kitahata MM, Van Rompaey SE, Shields AW: Physician experience in thecare of HIV-infected persons is associated with earlier adoption of newantiretroviral therapy. J Acquir Immune Defic Syndr 2000, 24:106-114.10. Lifson AR, INSIGHT Cause of Death Writing Group, Belloso WH, Carey C,Davey RT, Duprez D, El-Sadr WM, Gatell JM, Gey DC, Hoy JF, Krum EA,Nelson R, Nixon DE, Paton N, Pedersen C, Perez G, Price RW, Prineas RJ,Rhame FS, Sampson J, Worley J: Determination of the underlying cause ofdeath in three multicenter international HIV clinical trials. HIV Clin Trials2008, 9:177-185.11. Delgado J, Heath KV, Yip B, Marion S, Alfonso V, Montaner JS,O’Shaughnessy MV, Hogg RS: Highly active antiretroviral therapy:physician experience and enhanced adherence to prescription refill.Antivir Ther 2003, 8:471-478.12. Kitahata MM, Van Rompaey SE, Dillingham PW, Koepsell TD, Deyo RA,Dodge W, Wagner EH: Primary care delivery is associated with greaterphysician experience and improved survival among persons with AIDS. JGen Intern Med 2003, 18:95-103.13. Landon BE, Wilson IB, Cohn SE, Fichtenbaum CJ, Wong MD, Wenger NS,Bozzette SA, Shapiro MF, Cleary PD: Physician specialization andantiretroviral therapy for HIV. J Gen Intern Med 2003, 18:233-241.14. Wood E, Montaner JS, Tyndall MW, Schechter MT, O’Shaughnessy MV,Hogg RS: Prevalence and correlates of untreated humanimmunodeficiency virus type 1 infection among persons who have diedin the era of modern antiretroviral therapy. J Infect Dis 2003,188:1164-1170.15. Druyts EF, Rachlis BS, Lima VD, Harvard SS, Zhang W, Brandson EK,Strathdee SA, Montaner JSG, Hogg RS: Mortality is influenced by localityin a major HIV/AIDS epidemic. HIV Med 2009, 10:274-281.16. Hessol NA, Buchbinder SP, Colbert D, Scheer S, Underwood R, Barnhart JL,O’Malley PM, Doll LS, Lifson AR: Impact of HIV infection on mortality andaccuracy of AIDS reporting on death certificates. Am J Public Health 1992,82:561-564.17. Centres for Disease Control (CDC): Electronic Record Linkage to IdentifyDeaths Among Persons with AIDS — District of Columbia, 2000–2005.MMWR 2008, 57:631-634.18. Sorlie PD, Rogot E, Johnson NJ: Validity of demographic characteristics onthe death certificate. Epidemiology 1992, 3:181-184.Au-Yeung et al. BMC Public Health 2010, 10:642http://www.biomedcentral.com/1471-2458/10/642Page 4 of 519. Rosenberg HM, Maurer JD, Sorlie PD, Johnson NJ, MacDorman MF,Hoyert DL, Spitler JF, Scott C: Quality of death rates by race and Hispanicorigin: A summary of current research, 1999: National Center for HealthStatistics. Vital Health Stat 2 1999, 129:1-20.20. Arias E, Schauman WS, Eschbach K, Sorlie PD, Backlund E: The validity ofrace and Hispanic origin reporting on death certificates in the UnitedStates. Vital Health Stat 2 2008, 148:1-23.21. Statistics Canada: 2006 Census Dictionary: More information on Urbanarea (UA).[http://www12.statcan.gc.ca/english/census06/reference/dictionary/geo049a.cfm].22. Statistics Canada: 2006 Census Dictionary: More information on Ruralarea (RA).[http://www12.statcan.gc.ca/english/census06/reference/dictionary/geo042a.cfm].23. Wood E, Montaner JS, Yip B, Tyndall MW, Schechter MT, O’Shaughnessy MV,Hogg RS: Adherence and plasma HIV RNA responses to highly activeantiretroviral therapy among HIV-1 infected injection drug users. CMAJ2003, 169:656-661.24. World Health Organization: International Statistical Classification ofDiseases and Related Health Problems: 10th Revision.[http://www.who.int/classifications/apps/icd/icd10online/].Pre-publication historyThe pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2458/10/642/prepubdoi:10.1186/1471-2458-10-642Cite this article as: Au-Yeung et al.: Physician’s manual reportingunderestimates mortality: evidence from a population-based HIV/AIDStreatment program. BMC Public Health 2010 10:642.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitAu-Yeung et al. BMC Public Health 2010, 10:642http://www.biomedcentral.com/1471-2458/10/642Page 5 of 5


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items