UBC Faculty Research and Publications

Extent, trends, and determinants of controller/reliever balance in mild asthma: a 14-year population-based… Khakban, Amir; FitzGerald, J. M; Tavakoli, Hamid; Lynd, Larry; Ehteshami-Afshar, Solmaz; Sadatsafavi, Mohsen Feb 28, 2019

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

Item Metadata


52383-12931_2019_Article_1007.pdf [ 1.01MB ]
JSON: 52383-1.0376562.json
JSON-LD: 52383-1.0376562-ld.json
RDF/XML (Pretty): 52383-1.0376562-rdf.xml
RDF/JSON: 52383-1.0376562-rdf.json
Turtle: 52383-1.0376562-turtle.txt
N-Triples: 52383-1.0376562-rdf-ntriples.txt
Original Record: 52383-1.0376562-source.json
Full Text

Full Text

RESEARCH Open AccessExtent, trends, and determinants ofcontroller/reliever balance in mild asthma:a 14-year population-based studyAmir Khakban1,3, J. Mark FitzGerald2, Hamid Tavakoli3, Larry Lynd1,4, Solmaz Ehteshami-Afshar3 andMohsen Sadatsafavi1,2,3*AbstractBackground: The majority of patients with asthma have the mild form of the disease. Whether mild asthma patientsreceive appropriate asthma medications has not received much attention in the literature. We examined the trends inindicators of controller/reliever balance.Methods: Using administrative health databases of British Columbia, Canada (2000 to 2013), we created a population-based cohort of adolescents/adults with mild asthma using validated case definition algorithms. Each patient-year offollow-up was assessed based on two markers of inappropriate medication prescription: whether the ratio of controllermedications (inhaled corticosteroids [ICS] and leukotriene receptor antagonists [LTRA]) to total asthma-related prescriptionswas low (cut-off 0.5 according to previous validation studies), and whether short-acting beta agonists (SABA)were prescribed inappropriately according to previously published criteria that considers SABA in relation toICS prescriptions. Generalized linear models were used to evaluate trends and to examine the association betweenpatient-, disease-, and healthcare-related factors and medication use.Results: The final cohort consisted of 195,941 mild asthma patients (59.5% female; mean age at entry 29.6 years)contributing 1.83 million patient-years. In 48.8% of patient-years, controller medications were suboptimally prescribed,while in 7.2%, SABAs were inappropriately prescribed. There was a modest year-over-year decline in inappropriate SABAprescription (relative change − 1.3%/year, P < 0.001) and controller-to-total-medications (relative change − 0.5%/year, P < 0.001). Among the studied factors, the indices of type and quality of healthcare (namely respirologistconsultation and receiving pulmonary function test) had the strongest associations with improvement in controller/reliever balance.Conclusions: Large number of mild asthma patients continue to be exposed to suboptimal combinations of asthmamedications, and it appears there are modifiable factors associated with such phenomenon.Keywords: Asthma, Mild asthma, Controller medication, Reliever medication, TrendBackgroundAsthma is one of the most common chronic diseasesworldwide. While it cannot be cured, achieving clinicaland symptomatic control can substantially reduce theburden of asthma. Evidence shows that achieving controlis an attainable goal in the majority of patients, espe-cially those with mild disease [1]. The cornerstone ofpharmacological asthma management is controller medica-tions with anti-inflammatory effects, namely inhaled corti-costeroids (ICS), as well as leukotriene receptor antagonists(LTRA) [2]. On the other hand, many patients perceiveimmediate symptom relief through the use of reliever(or rescue) medications, such as inhaled short-actingbeta-agonists (SABA), which cause rapid resolution ofsymptoms through the temporary relaxation of airwaysmooth muscles. However, reliever medications do not* Correspondence: msafavi@mail.ubc.ca1Collaboration for Outcomes Research and Evaluation, Faculty ofPharmaceutical Sciences, the University of British Columbia, Vancouver, Canada2Division of Respiratory Medicine and Institute for Heart and Lung Health,Vancouver General Hospital, the University of British Columbia, Vancouver,CanadaFull list of author information is available at the end of the article© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Khakban et al. Respiratory Research           (2019) 20:44 https://doi.org/10.1186/s12931-019-1007-0tackle the core underlying inflammatory mechanisms ofasthma. Patients who mainly rely on reliever medica-tions are at increased risk of periods of intensifieddisease activity commonly known as exacerbations orattacks. One postulated mechanism for such an in-creased risk is that through symptomatic relief, relievermedications facilitate inhalation of exacerbation trig-gers which, in the absence of anti-inflammatory ther-apy, make patients particularly vulnerable to severe andpotentially fatal attacks [3–5]. Indeed, landmark studieshave demonstrated that monotherapy with SABAs isassociated with an increased risk of severe exacerba-tions and mortality [6, 7].While adverse outcomes such as severe exacerbationsand mortality mainly affect patients with severe asthma,the population-level consequences of inappropriate useof reliever medications among mild asthma patientsmight be larger given the vastly higher prevalence ofmild asthma. Indeed, recent evidence suggests that control-ler/reliever balance in mild asthma is directly associatedwith risk of poor symptom control and exacerbations.The SYGMA trial convincingly demonstrated thatas-needed budesonide (ICS) plus formoterol (a rapid-onsetlong-acting beta-agonist [LABA]) provided superior asthma-symptom control to as-needed terbutaline (SABA) in termsof weeks with well-controlled asthma, while budesonidemaintenance therapy was superior to both regimens. Exacer-bation rates with the two ICS-containing regimens weresimilar and were significantly lower than the rate withas-needed SABA therapy [8]. In a post hoc analysis of datafrom more than 7000 mild asthma patients from the multi-national Steroid Treatment As Regular Therapy (START)study, Reddel et al. demonstrated that even in the mildestform of asthma (symptoms 0–1 days per week), mainten-ance ICS therapy reduced the combined risk of severeasthma-related outcomes (defined as any of asthma-relatedhospital admission, emergency treatment, or death). Thisbenefit was consistently observed for both persistent andintermittent mild asthma [9]. In a previous work, we havedemonstrated the adverse consequences associated withsuch inappropriate use in a cohort of largely mild asthmapatients [10].Given such recent findings, it is important to understandthe current status of medication use in mild asthma. Suchfindings can enable estimating the population-level burdenof mild asthma that is potentially preventable through thepromotion of evidence-informed treatments. Such informa-tion is crucial to evaluate the return-on-investment of pro-grams and policies towards improving the management ofmild asthma. Accordingly, the purpose of the present studywas to document the extent, trends, and predictors ofthe balance between controller and reliever medicationuse in mild asthma, using a population-based longitu-dinal database.MethodsStudy populationWe used administrative health databases of all legal resi-dents of British Columbia (BC), from 1997 to 2014. BCis a Canadian province with a population of 4.65M (asof 2016, [11]) with a universal healthcare system, whoseadministrative needs have resulted in the creation ofcentralized datasets that capture all resource use recordsof legal residents in the province. We obtained thelinked births and deaths data [12], records of inpatientcare episodes [13], outpatient services use [14], andmedication dispensation [15]. We also had access tocensus data that provided estimates of neighborhood in-come as a proxy for socio-economic status [16]. Thisstudy was approved by the University of British Columbia’sHuman Ethics Board (application H15–00062). All infer-ences, opinions, and conclusions drawn in this research arethose of the authors and do not reflect the opinions or pol-icies of the Data Steward(s).From these datasets, we created a cohort of adolescent/adult asthma patients using a validated case definition ofasthma [17]. According to this definition, an individual isconsidered as having diagnosed asthma if they satisfy oneor more of the following three criteria in any rolling12-month period: 1) filled prescriptions on three differentdates for at least three asthma-related medications, or 2)had two outpatient asthma-related visits on different dates,or 3) had at least one hospitalization with the maindischarge code of asthma diagnosis (however, whilehospital-based component is part of this validatedalgorithm, further restriction of the sample to mildasthma cases removed the subgroup of patients whoentered the main asthma cohort through an instanceof asthma-related hospital admission). We used inter-national classification of disease (ICD), 9th revisioncode of 493.xx and ICD, 10th revision codes of J45-J46for identifying asthma-specific healthcare resourceuse. Asthma-related medications were determinedbased on expert opinion (list provided in Additionalfile 1: Table S1). The above-mentioned criteria wereapplicable for the period in which the individual wasbetween ages of 14 and 45. The upper age limit waschosen to reduce the risk of misclassification due toinclusion of patients with chronic obstructive pulmon-ary disease. However, once patients were identifiedduring this period, they could remain in the cohortregardless of their age (resulting in the upper age limitin the cohort being 59 years). We did not include thefirst three years of data (1997–1999) in the analyses toallow sufficient time for patients with asthma to beidentified. Also, we excluded the last year of data(2014), as the 12-month rolling window for asthmadefinition would result in under-representation ofasthma patients in this year.Khakban et al. Respiratory Research           (2019) 20:44 Page 2 of 9We considered the first date of any asthma-related re-source use after the patient’s 14th birthday as the indexdate, marking the beginning of follow-up. Follow-up wasterminated at the earliest date of last resource use of anytype, death, or the administrative end of study (Dec.31th, 2014). The follow-up time for each patient was di-vided, starting from the index date, into juxtaposing12-month intervals (henceforth referred to as ‘periods’)during which study outcomes were measured.Assessing asthma severityWe used a validated algorithm, developed using Canad-ian administrative databases [18], to classify each periodinto mild or non-mild asthma. This algorithm classifiesa patient-year as mild asthma according to the combin-ation of absence of exacerbation markers (e.g., emer-gency department visits, oral corticosteroid use, orhospital admissions) as well as asthma medication pat-terns, and has been validated, using chart review, againstthe Canadian Asthma Consensus Guidelines [18].OutcomesThe two co-primary outcomes of interest were suboptimalprescription of controller medications and inappropriateprescription of SABA, both defined during each period forevery patient using previously used and validated metrics.Suboptimal controller prescription was defined as a periodin which the ratio of total dose of ICS (beclomethasoneequivalent) and LTRA (all considered having equal po-tency) to total dose of all asthma-related medications wasless than 50% [19]. This is a validated metric of controller/reliever balance with an established association withasthma outcomes [19]. This metric in its original for doesnot consider different potencies of inhaled medications. Aprevious study has demonstrated that both dose-adjustedand unadjusted metrics of ICS use were predictive ofadverse asthma outcomes [20]. However, we decided toincorporate such dose-equivalence information into calcu-lations due to the perceived importance of ICS dosing inmild asthma give the recent evidence [9]. Inappropriateprescription of SABA was defined as a period in which apatient had filled prescriptions corresponding to greaterthan two puffs of SABA per week without any concomi-tant ICS prescription, or nine or more canisters of SABAsper year and no more than an average of 100 μg (beclo-methasone equivalent) per day of ICS [21]. This metrichas been used previously and has been shown to beassociated with adverse outcomes and increased costs.The above-mentioned outcomes were defined only inperiods in which the patient filled a prescription for atleast one asthma-related medication. The asthma wasconsidered dormant in the other periods and such periodswere not included in the subsequent analyses. All in-haled medications were dose-adjusted according to theestablished dose-equivalence information (Additionalfile 1: Table S1).Statistical analysisThe trends of the two outcomes were quantified overthree time axes: calendar year, time since mild asthmadiagnosis, and age. Each period was assigned to eachcalendar year depending on the date at the beginning ofthe period; similarly, the age of the patient was assessedat the beginning of each period. Only periods of mildasthma were included in the analyses. For the trendsover time since the incidence of mild asthma, a sub-group of incident mild asthma was created consisting ofpatients whose first year of asthma was classified as mildand who were present for at least five years in the databefore their asthma diagnosis. We tested the trends inoutcomes using Generalized Linear Models (GLM) withPoisson distribution and logarithmic link function, inwhich the number of periods with suboptimal controllerprescription or inappropriate SABA prescription was thedependent variable, and the logarithm of the total num-ber of periods was the offset variable.For associating factors with suboptimal controller pre-scription and inappropriate SABA prescription, we fittedGLMmodels with binomial distribution and logit link func-tion (equivalent of conventional logistic regression butaccommodating the clustered nature of data [multiple pe-riods for each patient]). The dependent variables for thisanalysis were suboptimal prescription of controllers andinappropriate prescription of SABAs in the current period,and the independent variables were a series of covariates inthe same or in the immediate preceding period. These werevariables pertaining to socio-demographic characteristics,asthma-related treatments, comorbidity burden (includingthe Charlson comorbidity index [22]), and variables repre-senting type (specialist versus primary care) and continuityof care. The latter was measured using the Bice-Boxemanindex for each patient-year [23]. This index varies between0 and 1, with zero meaning that an individual’s physicianvisits were all to different physicians during the year, and 1meaning that the individual only consulted with the samephysician during the year.P-values (P) were considered significant at the 0.05level (two-tailed). SAS Enterprise Guide (version 7.3,SAS Institute, Cary, NC, USA) was used for data linkageand preparation steps and for statistical analyses.ResultsA total of 201,289 patients met the case definition ofasthma. Among these, 195,941 (mean age at entry 29.6years, 59.5% female) had at least one period categorizedas mild asthma and thus were included in the analyticaldataset (Table 1). The incident cohort of asthma in-cluded 88,110 individuals.Khakban et al. Respiratory Research           (2019) 20:44 Page 3 of 9Extent of suboptimal controller and inappropriate SABAprescriptionsThe average ratio of total dose of controllers to totaldose of all asthma-related medications was 0.44 (SD0.4). In 48.8% of periods, controllers were prescribedsuboptimally (ratio < 0.5), and in 7.2% of periods SABAswere prescribed inappropriately.Trends over calendar timeTrend of controller-to-total-asthma medications ratioshowed a slight drop from 52 to 48.2% from 2000 to2013 (Fig. 1- left panel). The annual trend was slightlydownward, with a relative decline of 0.5% per year (P <0.001). Inappropriate prescription of SABAs decreasedfrom 8.5 to 6.8% during the same period. Aside from anupward trend in inappropriate prescription of SABAs inthe 2008–2010 period, the overall trend was decreasing,with an annual relative decline of 1.3% (P < 0.001)(Fig. 1- right panel).Trends over ageTrends are demonstrated in Fig. 2. The proportion ofpatients with suboptimal prescription of controllersgenerally increased over young age groups, from 43% in14 years to 57% in 24 years of age. This proportiondecreased to 44% by age 40, and remained mostly con-stant afterwards. The overall trend was consistent with arelative decrease of 0.8% per year of age (P < 0.001).Inappropriate prescription of SABAs declined from 8.3%at age 14 to 7.0% at age 18. This ratio increased to 9.4%by age 24–29 and then gradually declined, plateauingaround 5.5%. Overall, the trend was consistent with arelative decline of 1.1% per year of age (P < 0.001).Trends over time course of asthmaThe trend in controller-to-total -medication ratio sincediagnosis of asthma was not significant (P = 0.81). Theproportion of individuals with inappropriate SABA pre-scription was 6.2% in the incident year of mild asthma.This ratio dropped sharply in the second year to 4.8%but had a slight increase to 5.4% during the next 10 years(Fig. 3 – right panel). The trend for this metric was con-sistent with 1.4% relative decrease per year, mainly dueto the sharp decline form the first to the second year(P < 0.001).Factors associated with suboptimal controller andinappropriate SABA prescriptionsResults of the regression analyses are provided in Table 2.Individuals with higher socio-economic status as well asfemale patients had a lower likelihood of suboptimal con-troller prescriptions. Variables related to better quality ofTable 1 Follow-up statistics and characteristics of study populationVariable CohortTotal sample 195,941Patient years 1,827,150Female; N (%) 114,673 (59.5%)Age at index date; Mean (SD) 29.6 (9.55)Socio-economic status; N (%)Quintile 1 19,014 (9.7%)Quintile 2 26,332 (13.5%)Quintile 3 35,420 (18.2%)Quintile 4 47,858 (24.6%)Quintile 5 64,326 (33%)Unknown 1824 (0.9%)Follow up years; Mean (SD) 10.7 (4.95)Ratio of ICS to total asthma medications; Mean (SD) 0.44 (0.400)Suboptimal use of ICS (< 50%) 373,911 (48,8%)Inappropriate use of SABAs; N (%) 55,414 (7.2%)ICS inhaled corticosteroids, SABA Short-acting beta agonist, SDStandard deviationFig. 1 Trend of controller prescription to total medications (left panel) and inappropriate use of SABAs (right panel) over calendar yearKhakban et al. Respiratory Research           (2019) 20:44 Page 4 of 9care, including the receipt of a pulmonary function test,non-urgent physician visits related to asthma (to bothgeneral practitioners and respiratory specialists but not tointernal medicine specialists), and higher continuity of carehad an inverse association with suboptimal controller pre-scription. Conversely, higher burden of comorbidity, higherseverity of asthma, and history of asthma-related hospitali-zations in the preceding period had positive associationwith suboptimal controller prescription. We observedgenerally similar patterns of association for inappropriateprescription of SABAs, aside from comorbidity andasthma-related hospitalizations, which were no longerstatistically significant. In addition, asthma related generalpractitioner visits showed a positive association with in-appropriate prescription of SABA.DiscussionWe evaluated the indicators of balance between control-ler and reliever medications and their trends in 195,941mild asthma patients during 14 years of longitudinal,Fig. 2 Trend of controller prescription to total medications (top panel) and inappropriate use of SABAs (bottom panel) over age of mildasthma patientsFig. 3 Trend of controller prescription to total medications (left panel) and inappropriate use of SABAs (right panel) over the time course of asthmaKhakban et al. Respiratory Research           (2019) 20:44 Page 5 of 9population-based data. The controller/reliever balancewas quantified as two metrics: suboptimal (< 0.50) pre-scription ratio of controllers to total asthma medications,and inappropriate prescription of SABAs according topreviously published algorithms. Both metrics showed adecline over the period of 2000–2013. Over the entirefollow up, the suboptimal prescription of controllers rela-tively dropped by almost 8% and inappropriate SABA pre-scription relatively declined by 20%. On the other hand,during the first eleven years of time course of mildasthma, the trends of suboptimal ICS prescriptionremained steady while inappropriate SABA prescriptionfell from the first to the second year but remained rela-tively constant afterwards. Both metrics were had thehighest (worst) values for patients in their twenties. Wealso evaluated the association between such metrics andmultiple factors related to patient or disease characteris-tics and type and quality of care. Among the studied fac-tors, the indices of type and quality of care, namelycontinuity of care and respirologist consultation, had thestrongest associations with improvement in controller/re-liever balance. As well, higher socio-economic status andfemale sex were associated with more appropriate and bal-anced prescriptions of asthma medications.Previous studies have evaluated the trends in asthmamedications [24–27]. Highasi et al. used data from twonational surveys in the United States to document trendsof asthma medications between 1997 and 2008 [24].Their results indicated declines in the prescriptions ofSABAs and increase in ICS prescriptions. They also indi-cated an increase in the ratio of controller to totalasthma medications from 0.5 in 1997 to a peak of 0.7 in2004. Similarly, Johnson et al. used data from a Europeanprospective cohort study with an average follow-up of8.7 years [25]. In the sample with asthma (n = 423), theprescriptions for ICS increased by 12.2% over a 10-yearperiod. Despite this, only 17.2% were using ICS on adaily basis during follow-up.To the best of our knowledge there has not been anystudies on medication balance in mild asthma, whichconstitutes the majority of asthma patients. In additionevaluating trends over age and over the time course ofthe disease are novel features of our design. We showeda decline in the inappropriate prescription of controllersand relievers, especially for the latter, between the firstand second year. This can be due to higher prescriptionof reliever medications and avoidance of ICS in earlystage of asthma [24]. We also showed that controller/Table 2 Factors associated with suboptimal controller use and inappropriate use of relieversController to total medications < 50% Inappropriate SABA useGroup Variable OddsRatio95% CI P value OddsRatio95% CI P value(Lower, Upper) (Lower, Upper)Socio-demographic Sex (female = 1) 0.97 0.95–0.99 <.0003 0.78 0.76–0.81 <.0003Higher SES 0.97 0.97–0.98 <.0001 0.98 0.97–0.98 <.0001Year 0.99 0.98–1.00 <.0001 0.96 0.96–0.97 <.0001Age 0.81 0.81–0.82 <.0001 0.9 0.89–0.91 <.0001Asthma treatment variables Asthma-related hospitalization 1.63 1.51–1.76 <.0001 0.99 0.83–01.18 0.9222Asthma-related outpatient visit 0.95 0.94–0.96 <.0001 0.89 0.88–0.91 <.0001Oral corticosteroid 1.28 1.25–1.30 <.0001 0.4 0.38–0.41 <.0001Severity of Asthma 1.18 1.16–1.20 <.0001 3.6 3.52–3.69 <.0001Comorbidity-related variables Modified Charlson score 1.11 1.09–1.12 <.0001 0.98 0.96–1.00 0.1033None asthma related hospitalization 1.01 1.00–1.02 0.0009 1.01 1.00–1.03 0.0789None asthma related outpatient visit 1.00 1.00–1.00 <.0001 1.00 1.00–1.00 0.0137Type & quality of care Having received pulmonary function test 0.95 0.92–0.97 <.0003 0.87 0.82–0.92 <.0001Respirologist consultation 0.57 0.54–0.59 <.0001 0.54 049–0.60 <.0001Internal medicine consultation 0.98 0.93–1.03 0.5501 0.74 0.67–0.81 0.0014General Practitioner Asthma visit 0.67 0.66–0.68 <.0001 1.16 1.12–1.19 <.0001Continuity of care (COC)COC = 0 – – – – –COC > 0 and COC < 50% 0.84 0.82–0.86 <.0001 0.76 0.74–0.79 <.0001COC > =50% and OC < 100% 0.86 0.84–0.88 <.0001 0.84 0.80–0.88 0.0014COC = 100% 0.86 0.83–0.83 <.0001 0.85 0.80–0.90 <.0001SABA short-acting beta agonist, SES socio-economic status, CI confidence intervalKhakban et al. Respiratory Research           (2019) 20:44 Page 6 of 9reliever balance was worse in the patients in their twen-ties. This might be explained by lower adherence to con-troller therapies in adolescents and young adults, asobserved elsewhere [28]; alternatively it might indicate afailure of proper transfer of care from the pediatrician tothe adult care provider.Several factors can affect the way asthma medicationsare prescribed by care providers and used by patients.Within limitations of administrative health data, we eval-uated multiple factors ranging from socio-demographicvariables to disease characteristics, type and quality ofcare, and comorbidity. Our results are consistent withprevious studies with regard to higher socio-economicstatus and female sex being associated with more balanceduse of asthma medications [29–31]. Other observed asso-ciations appear to be reported for the first time in thisstudy. For instance, continuity of care was associated withbetter asthma medication use. Interestingly, general prac-titioner visits were associated with an increased risk ofinappropriate SABA prescriptions but a decreased risk oflow controller-to-total-asthma-medication ratio. Respirol-ogist consultation was associated with improvement inboth indices. This discrepancy might be due to the differ-ences among the guidelines available to each group, ortheir adherence in following best practice recommenda-tions. Interestingly, asthma-related hospitalization in theprevious year increased the risk of suboptimal prescriptionof controller, which can be a concerning factor.A notable association was observed between comorbidityand suboptimal controller prescription. This association wasour a priori hypothesis, based on the observed associationbetween age and suboptimal asthma medication use in thegeneral population in our previous work [26]. We hypothe-sized that a factor that might increase suboptimal use ofasthma medications is the higher burden of comorbidity inthe older age groups, which might direct the patient’s andcare provider’s attention away from the proper managementof asthma. The association with comorbidity was only statis-tically significant for controller-to-total-asthma medicationratio but not for inappropriate SABA prescription. In an ex-ploratory analysis, when we divided the Charlson comorbid-ity index into its constituents, almost all diseases remainedpositively associated with suboptimal controller prescription(Additional file 2: Table S2).Using a large population-based sample of entire popu-lation of a well-defined geographical region with a longfollow-up period is the key strength of this study. Cover-ing the entire population of a jurisdiction with a publichealthcare system, the study sample is nearly free of theselection bias that affects clinical studies or claims-basedrecords from third-party insurers. This, combined withthe long follow-up time, allowed us to estimate trendsover different time axes with a low level of uncertainty.The large sample size enabled us to robustly evaluatethe association between multiple factors and controller/reliever balance. There are also some limitations in ourstudy. While validated specifically for the Canadian con-text and used in several previous studies [32], the classi-fication of asthma based on severity relies on healthservices use and medication records; as such it can beaffected by many factors such as lack of adherence tomedications. Second, filling prescriptions is not equal tomedication intake. However, we deem it unlikely thatany discrepancy between filled prescriptions and actualintake would substantially change over time; therefore,such discrepancy is unlikely to threaten the validity ofthe analyses of trends. In addition, we did not have in-formation about important factors such as patients’smoking status and degree of airway limitation. Further,classification of patient-years into asthma severity cat-egories was independent of the previous history ofasthma. As such, some patient-years that were classifiedas mild asthma after a severe asthma episode were in-cluded in the analytical dataset. Given the objective ofthis study, the inclusion of such periods is justified, butasthma patients with a past history of severe episodesmight be treated differently. In a secondary analysis (re-sults not shown) when such patient-years were removed,the overall findings stayed the same.ConclusionIn summary, suboptimal prescription of asthma medica-tions are prevalent in mild asthma. Although the decliningtrends over time are encouraging, given the sheer size ofmild asthma population, still unacceptably large number ofpatients are exposed to inappropriate doses of reliever med-ications. For example, our results indicate that in 2013, outof 45,139 non-dormant mild asthma patients in BC, 22,153individuals were exposed to low doses of controller therap-ies in relation with total asthma medications, and 3056 pa-tients had indicators of inappropriate reliever medicationprescription. Recent evidence suggests the benefit of earlieruse of controller therapy in mild asthma patients [8, 9] andour results indicate that there is substantial room for suchevidence to be translated into practice towards improve-ment of patient outcomes in mild asthma.Additional filesAdditional file 1: Table S1. List of asthma related medication. (DOCX49 kb)Additional file 2: Table S2. Odd ratios of association betweencomorbid conditions on ICS to total medications < 50%. (DOCX 15 kb)AbbreviationsBC: British Columbia; GLM: Generalized Liner Models; ICD: InternationalClassification of Disease; ICS: Inhaled Corticosteroids; P: P-values;SABA: Short-Acting Beta-AgonistsKhakban et al. Respiratory Research           (2019) 20:44 Page 7 of 9FundingThis study was funded by an arm’s length research contract from AstraZenecaCanada, mediated through the University-Industry-Liaison Office of the Universityof British Columbia. The funder had no role in study design and its conduct and isnot aware of the content of the submitted manuscript.Availability of data and materialsThe data that support the findings of this study are available fromPopulation Data BC (PopData) but restrictions apply to the availability ofthese data to the general public. BC’s Freedom of Information andProtection of Privacy Act (FIPPA) is committed to upholding the individual’sright to privacy and protection of their personal information; as suchthe data are not publicly available. For more information refer tohttps://www.popdata.bc.ca/privacy/policies/legislativeframeworkAuthors’ contributionsA.K. and M.S. are guarantors of the manuscript. M. S., J. M. F., and L. L. areresponsible for the concept of the study. A.K. and M.S. designed the studyand created the data analysis plan. J. M. F. and L. L. provided feedback onthe design. A.K., with contribution of H.T., performed all the statisticalanalyses. M. S. wrote the first draft of the manuscript. S.E.A. contributed inrevising and preparing the final draft of manuscript. All authors criticallycommented on the manuscript and approved the final version.Ethics approval and consent to participateThis study was approved by the University of British Columbia’s HumanEthics Board (application H15–00062). All inferences, opinions, andconclusions drawn in this research are those of the authors and do notreflect the opinions or policies of the Data Steward(s).Competing interestJMF is a member of the Global Initiative for Asthma (GINA) Executive andScience Committees. He has served on advisory boards for Novartis, Pfizer,AstraZeneca, Boehringer-Ingelheim, and Merck. He has also been a memberof speakers’ bureaus for AstraZeneca, Boehringer-Ingelheim, Novartis, andMerck. He has received research funding paid directly to the University ofBritish Columbia from AstraZeneca, Glaxo-SmithKline, Boehringer-Ingelheim,Merck, Sanofi, and Novartis. MS receives salary support from the CanadianInstitutes of Health Research and Michael Smith Foundation for HealthResearch. He has received research funding paid directly to the Universityof British Columbia from AstraZeneca. Other authors have no conflict ofinterests to declare.Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.Author details1Collaboration for Outcomes Research and Evaluation, Faculty ofPharmaceutical Sciences, the University of British Columbia, Vancouver, Canada.2Division of Respiratory Medicine and Institute for Heart and Lung Health,Vancouver General Hospital, the University of British Columbia, Vancouver,Canada. 3Respiratory Evaluation Sciences Program, Faculty of PharmaceuticalSciences, the University of British Columbia, Vancouver, Canada. 4Center forHealth Evaluation and Outcome Sciences, Vancouver, Canada.Received: 9 May 2018 Accepted: 15 February 2019References1. Bateman ED, Boushey HA, Bousquet J, Busse WW, Clark TJH, Pauwels RA, etal. Can guideline-defined asthma control be achieved? The gaining optimalasthma control study. Am J Respir Crit Care Med. 2004;170(8):836–44.2. Global INitiative for Asthma. GINA Report, Global Strategy for AsthmaManagement and Prevention [Internet]. 2015. Available from: https://ginasthma.org/wp-content/uploads/2016/01/GINA_Report_2015_Aug11-1.pdf. Accessed 12 Jan 2018.3. Cockcroft DW, McParland CP, Britto SA, Swystun VA, Rutherford BC. Regularinhaled salbutamol and airway responsiveness to allergen. Lancet LondEngl. 1993;342(8875):833–7.4. Sears MR. Adverse effects of beta-agonists. J Allergy Clin Immunol. 2002;110(6 Suppl):S322–8.5. Lommatzsch M, Lindner Y, Edner A, Bratke K, Kuepper M, Virchow JC.Adverse effects of salmeterol in asthma: a neuronal perspective. Thorax.2009;64(9):763–9.6. Spitzer WO, Suissa S, Ernst P, Horwitz RI, Habbick B, Cockcroft D, et al. Theuse of beta-agonists and the risk of death and near death from asthma. NEngl J Med. 1992;326(8):501–6.7. Crane J, Pearce N, Flatt A, Burgess C, Jackson R, Kwong T, et al. Prescribedfenoterol and death from asthma in New Zealand, 1981-83: case-controlstudy. Lancet Lond Engl. 1989;1(8644):917–22.8. O’Byrne PM, FitzGerald JM, Bateman ED, Barnes PJ, Zhong N, Keen C, et al.Inhaled Combined Budesonide-Formoterol as Needed in Mild Asthma. NEngl J Med. 2018;378(20):1865–76.9. Reddel HK, Busse WW, Pedersen S, Tan WC, Chen Y-Z, Jorup C, et al. Shouldrecommendations about starting inhaled corticosteroid treatment for mildasthma be based on symptom frequency: a post-hoc efficacy analysis of theSTART study. Lancet Lond Engl. 2017;389(10065):157–66.10. FitzGerald JM, Tavakoli H, Lynd LD, Al Efraij K, Sadatsafavi M. The impact ofinappropriate use of short acting beta agonists in asthma. Respir Med. 2017Oct;131:135–40.11. BC Stats. Central statistical agency of BC. Census Population of BC andCanada 1871 to 2016 [Internet]. BC Stats; 2017. Available from: http://www.bcstats.gov.bc.ca/StatisticsBySubject/Census/2016Census/PopulationHousing/BCCanada.aspx. Accessed 10 Jan 2018.12. BC Vital Statistics Agency. Vital Statistics Deaths. Population Data BC. DataExtract BC Vital Statistics Agency. [Internet]. Population Data BC; 2014.Available from: http://www.popdata.bc.ca/data. Accessed 10 Jan 2018.13. Canadian Institute for Health Information. Discharge Abstract Database(Hospital Separations). Population Data BC [publisher]. Data Extract. MOH.[Internet]. Population Data BC; 2014. Available from: http://www.popdata.bc.ca/data. Accessed 10 Jan 2018.14. British Columbia Ministry of Health. Medical services plan (MSP) paymentinformation file. Population Data BC Data Extract MOH [Internet]. 2014.Available from: http://www.popdata.bc.ca/data. Accessed 10 Jan 2018.15. BC Ministry of Health. PharmaNet. BC Ministry of Health Data Extract DataStewardship Committee [Internet]. 2013. Available from: http://www.popdata.bc.ca/data. Accessed 10 Jan 2018.16. British Columbia Ministry of Health. Consolidation file (MSP Registration &Premium Billing). Population Data BC. Data Extract. MOH. [Internet]. 2014.Available from: http://www.popdata.bc.ca/data. Accessed 10 Jan 2018.17. Prosser RJ, Carleton BC, Smith MA. Identifying persons with treated asthma usingadministrative data via latent class modelling. Health Serv Res. 2008;43(2):733–54.18. Firoozi F, Lemière C, Beauchesne M-F, Forget A, Blais L. Development andvalidation of database indexes of asthma severity and control. Thorax.2007;62(7):581–7.19. Laforest L, Licaj I, Devouassoux G, Eriksson I, Caillet P, Chatte G, et al.Prescribed therapy for asthma: therapeutic ratios and outcomes. BMC FamPract. 2015;16:49.20. Schatz M, Zeiger RS, Yang S-JT, Chen W, Crawford WW, Sajjan SG, et al.Relationship of asthma control to asthma exacerbations using surrogate markerswithin a managed care database. Am J Manag Care. 2010;16(5):327–33.21. Anis AH, Lynd LD, Wang X, King G, Spinelli JJ, Fitzgerald M, et al. Doubletrouble: impact of inappropriate use of asthma medication on the use ofhealth care resources. CMAJ Can Med Assoc J. 2001;164(5):625–31.22. Sangha O, Stucki G, Liang MH, Fossel AH, Katz JN. The self-administeredcomorbidity questionnaire: a new method to assess comorbidity for clinicaland health services research. Arthritis Rheum. 2003;49(2):156–63.23. Bice TW, Boxerman SB. A quantitative measure of continuity of care. MedCare. 1977;15(4):347–9.24. Higashi A, Zhu S, Stafford RS, Alexander GC. National trends in ambulatoryasthma treatment, 1997-2009. J Gen Intern Med. 2011;26(12):1465–70.25. Janson C, de Marco R, Accordini S, Almar E, Bugiani M, Carolei A, et al.Changes in the use of anti-asthmatic medication in an international cohort,vol. 26; 2006. p. 1047.26. Sadatsafavi M, Tavakoli H, Lynd L, FitzGerald JM. Has asthma medication usecaught up with the evidence?: a 12-year population-based study of trends.Chest. 2017;151(3):612–8.27. Haahtela T, Tuomisto LE, Pietinalho A, Klaukka T, Erhola M, Kaila M, et al. A10 year asthma programme in Finland: major change for the better. Thorax.2006;61(8):663–70.Khakban et al. Respiratory Research           (2019) 20:44 Page 8 of 928. Mäkelä MJ, Backer V, Hedegaard M, Larsson K. Adherence to inhaledtherapies, health outcomes and costs in patients with asthma and COPD.Respir Med. 2013;107(10):1481–90.29. Blanchette CM, Culler SD, Ershoff D, Gutierrez B. Association betweenprevious health care use and initiation of inhaled corticosteroid and long-acting beta2-adrenergic agonist combination therapy among US patientswith asthma. Clin Ther. 2009;31(11):2574–83.30. Jobin M-S, Moisan J, Bolduc Y, Dorval E, Boulet L-P, Grégoire J-P. Factorsassociated with the appropriate use of asthma drugs. Can Respir J.2011;18(2):97–104.31. Sundberg R, Torén K, Franklin KA, Gislason T, Omenaas E, Svanes C, et al.Asthma in men and women: treatment adherence, anxiety, and quality ofsleep. Respir Med. 2010;104(3):337–44.32. Chen W, Marra CA, Lynd LD, FitzGerald JM, Zafari Z, Sadatsafavi M. Thenatural history of severe asthma and influences of early risk factors: apopulation-based cohort study. Thorax. 2016;71(3):267–75.Khakban et al. Respiratory Research           (2019) 20:44 Page 9 of 9


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