Non-adherence to Chronic Medications Used in COPD and Its Association with Adverse Events: A Longitudinal, Population Based Study

Background: Medications used treat COPD are effective for symptom management and the prevention of exacerbations, however, adherence tends to be sub-optimal. Previous studies evaluating the association between adherence to COPD medications and COPD related adverse health outcomes have produced conflicting results. Additionally, these studies present several methodological limitations including failure to account for the time-varying nature of adherence or stepwise escalation of COPD therapy over time (based on COPD severity).

Objectives: To determine the association between non-adherence to chronic medications used in COPD and COPD related hospitalizations and emergency department visits in patients with incident COPD.

Methods: We conducted a population based longitudinal cohort study in the province of Ontario, Canada from 2003-2018 using multiple linked population health databases. All individuals aged 66 years or older who met a validated case definition of incident COPD were included. Patients were followed until death, end of insurance coverage or admission to long term care. Adherence was measured as the time varying proportion of days covered (PDC) in the past 90-days. To account for time dependent confounding according to therapeutic step escalations, we applied the parametric G-formula. Other confounders included socio-demographics, chronic conditions, continuity of care use of other medications and home oxygen.

Results: Overall, 78,097 patients were included in the study cohort; 44% were female, mean age was 75 (SD 9) and 20% had five or more chronic conditions. In the 90-days prior to and event or end of follow up, 11,200 (14%) patients were completely non-adherent to their COPD medications (PDC=0%), 10,845(14%) had low adherence (PDC=33%), 13,188 (17%) had medium adherence (PDC=70%) and 42,864 (55%) were completely adherent (PDC=100%). The incidence rate for COPD related ED visits was 1.6 events per 10,000 person days (PD) and for COPD related hospitalization was 3.6 events per 10,000 PD. After adjustment for confounders and time-varying therapeutic step, low, medium and high adherence in 90-days was associated with a decreased risk of both COPD related ED visits and hospitalizations. For example, those who were completely adherent were 16% less likely to experience a COPD related ED visit (adjusted risk ratio [aRR]): 0.61, 95% CI: 0.52-0.69) compared to those who were completely non-adherent while those who had medium adherence were 38% less likely to experience a COPD related hospitalization (aRR: 0.62, 95% CI:0.57-0.62).

Conclusions: After accounting for stepwise escalation of COPD therapy over time, higher adherence to COPD medications was associated with a decreased risk of COPD related emergency department visits and hospitalizations.