How temporal patterns of medication adherence to antidepressants, bisphosphonates and statins are associated with healthcare cost

ABSTRACT

We evaluate benefits of measuring temporal patterns of medication use using group-based trajectory modelling (GBTM) by finding associations between medication adherence (MA) and healthcare costs for which inconsistent results have been reported. We conducted a retrospective cohort study of 9,287, 1,660 and 10,242 users of antidepressants, bisphosphonates and statins, respectively, between 2012 and 2016, who participated in the 45 and Up Study in New South Wales, Australia. Associations between MA and subsequent 1-year healthcare cost components from generalized linear models and two-part models were compared across medications and MA measure types including GBTM and proportion of days covered (PDC). Compared to adherers, antidepressant users with declining adherence showed increased inpatient costs ($643 at p < 0.05). Bisphosphonate users with declining and consistently low adherence showed decreased osteoporotic inpatient costs ($387 at p < 0.05 and $419 at p < 0.05, respectively). All types of nonadherence were associated with reduced cost of the respective medications. GBTM revealed relationships between specific nonadherence types and healthcare costs, even in cases where no relationship was identified using the combined nonadherence PDC measure. With additional insights into the dynamics of nonadherence, GBTM could help clinicians and health policymakers design effective MA interventions, leading to improved health outcomes and pharmaceutical use.

KEYWORDS:

• Medication adherence
• healthcare cost
• group-based trajectory model
• data linkage

JEL CLASSIFICATION:

• I10
• H51

Acknowledgments

This research was completed using data collected through the 45 and Up Study (www.saxinstitute.org.au). The 45 and Up Study is managed by the Sax Institute in collaboration with major partner Cancer Council NSW; and partners: the Heart Foundation; NSW Ministry of Health; NSW Department of Communities and Justice; and Australian Red Cross Lifeblood. We thank the many thousands of people participating in the 45 and Up Study. We acknowledge Services Australia for supply of the Medicare Benefits Schedule and Pharmaceutical Benefits Scheme data, the Centre for Health Record Linkage (CHeReL) and Sax Institute for data linkage, and the Secured Unified Research Environment (SURE) for provision of secure data access.

Authors’ contributions

All authors contributed to the development of research questions and methodology, data collection and the interpretation of the results. Kyu Hyung Park conducted the data analysis and drafted the first manuscript. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data used in this study are not available for distribution as required by the ethics approvals.

Code availability

The R and STATA code used in the analysis and original statistical outputs are available from the corresponding author on reasonable request.

Ethical approval

The study was approved by the NSW Population & Health Services Research Ethics Committee (Project no: 2019/ETH12440) and the Macquarie University Human Research Ethics Committee (Reference no: 5201952579218). The conduct of the 45 and Up Study was approved by the University of New South Wales Human Research Ethics Committee.

Informed consent from participants

Consent was provided by 45 and Up Study participants.

Consent to publish

Consent was provided by 45 and Up Study participants.

Notes

¹ While more recent versions are available, the National Efficient Price Determination 2013–2014 is the last version that provides the calculation method not requiring additional components not available in the data sets used in this study..

² Low: 10–15; moderate: 16–21; high: 22–29; and very high: 30–50.

³ The comorbidity index rather than multiple indicator variables for specific comorbidities was used for parsimony.

⁴ The PDC is estimated to provide the proportion of days with medication available and hence is capped at 100%.

⁵ Note that we calculated the average marginal effects, not the marginal effects at the average.

⁶ This occurs when the end time of data for outcomes is earlier than the end of the follow-up period set. Whether a case is excluded or not depends only on the index date and is not related to participant characteristics.

⁷ The exceptions include regressions for primary care cost and medication cost in antidepressants, total cost and medication cost in bisphosphonates, and total cost in statins..

Additional information

Funding

This research is conducted as part of a PhD funded through a programme jointly funded by Macquarie University (Australia) and five pharmaceutical companies (Amgen Australia Pty Ltd, Janssen-Cilag Pty Ltd, Merck Sharp & Dohme Australia Pty Limited, Pfizer Australia Pty Ltd, and ROCHE Products Pty Limited).