Modelling medication adherence patterns for patients with serious mental and physical illnesses

Please check out my recent paper titled “Medication Adherence Patterns Among Patients with Multiple Serious Mental and Physical Illnesses” with co-authors Joanna P. MacEwan, Alison R. Silverstein, Darius N. Lakdawalla, Ainslie Hatch and Felicia M. Forma in Advances in Health.   This study is unique because it looks at medication adherence patterns among patients with serious mental and physical issues.  Rather than focus on a single drug, it looks at adherence trends for medications treating both a patient’s mental and physical illnesses using trajectory modelling.  The abstract is below and full text is here.

Introduction

Patients with mental and physical health conditions are complex to treat and often use multiple medications. It is unclear how adherence to one medication predicts adherence to others. A predictive relationship could permit less expensive adherence monitoring if overall adherence could be predicted through tracking a single medication.

Methods

To test this hypothesis, we examined whether patients with multiple mental and physical illnesses have similar adherence trajectories across medications. Specifically, we conducted a retrospective cohort analysis using health insurance claims data for enrollees who were diagnosed with a serious mental illness, initiated an atypical antipsychotic, as well as an SSRI (to treat serious mental illness), biguanides (to treat type 2 diabetes), or an ACE inhibitor (to treat hypertension). Using group-based trajectory modeling, we estimated adherence patterns based on monthly estimates of the proportion of days covered with each medication. We measured the predictive value of the atypical antipsychotic trajectories to adherence predictions based on patient characteristics and assessed their relative strength with the R-squared goodness of fit metric.

Results

Within our sample of 431,591 patients, four trajectory groups were observed: non-adherent, gradual discontinuation, stop–start, and adherent. The accuracy of atypical antipsychotic adherence for predicting adherence to ACE inhibitors, biguanides, and SSRIs was 44.5, 44.5, and 49.6%, respectively (all p < 0.001 vs. random). We also found that information on patient adherence patterns to atypical antipsychotics was a better predictor of patient adherence to these three medications than would be the case using patient demographic and clinical characteristics alone.

Conclusion

Among patients with multiple chronic mental and physical illnesses, patterns of atypical antipsychotic adherence were useful predictors of adherence patterns to a patient’s adherence to ACE inhibitors, biguanides, and SSRIs.

Modelling medication adherence patterns for patients with serious mental and physical illnesses posted first on http://dentistfortworth.blogspot.com