Measuring Medication Adherence in a Population-Based Asthma Administrative Pharmacy Database: A Systematic Review and Meta-Analysis

Abstract

Background

Limited studies have systematically reviewed the literature to identify and compare the various database methods and optimal thresholds for measuring medication adherence specific to adolescents and adults with asthma. In the present study, we aim to identify the methods and optimal thresholds for measuring medication adherence in population-based pharmacy databases.

Methods

We searched PubMed, Embase, International Pharmaceutical Abstracts (IPA), Web of Science, Google Scholar, and grey literature from January 1, 1998, to March 16, 2021. Two independent reviewers screened the studies, extracted the data, and assessed the quality of the studies. A quantitative knowledge synthesis was employed.

Results

Thirty-eight (38) retrospective cohort studies were eligible. This review identified 20 methods for measuring medication adherence in adolescent and adult asthma administrative health records. Two measures namely the medication possession ratio (MPR) and proportion of days covered (PDC) were commonly reported in 87% of the literature included in this study. From the meta-analysis, asthma patients who achieved adherence threshold of “0.75–1.00” [OR: 0.56, 95% CI: 0.41 to 0.77] and “>0.5” [OR: 0.71, 95% CI: 0.54 to 0.94] were less likely to experience asthma exacerbation.

Conclusion

Despite their limitations, the PDC and the MPR still remain the most common measures for assessing adherence in asthma pharmacy claim databases. The evidence synthesis showed that an adherence threshold of at least 0.75 is optimal for classifying adherent and non-adherent asthma patients.

Keywords:

• medication adherence
• adherence measures
• asthma
• adherence thresholds
• meta-analysis
• administrative health databases
• review

Acknowledgments

We acknowledge the effort of Alison Farrell, a Health Science Librarian at the Memorial University of Newfoundland, for her effort and assistance in performing an exhaustive literature search. M.A.-B. was supported by the Research and Graduate studies (RGS) scholarship at Memorial University of Newfoundland and TPMI/NL SUPPORT Educational scholarship.

Disclosure

The authors report no conflicts of interest in this work.

Additional information

Funding

There is no funding to report.