Medication non-adherence among heart failure patients: a multi-hospital, multi-dimensional analytical study in Punjab, Pakistan.

BACKGROUND

Heart failure (HF) has been a significant health problem among the population, with a significant impact on morbidity, mortality, and health care expenditure across the world. In Pakistan, HF-specific non-adherence data are limited, which restricts clinicians to come up with interventions that can be used to address the at-risk population. Accordingly, the newly-conducted study intended to determine the incidence of medication non-adherence and predictors at various levels (sociodemographic, clinical, treatment-related, and behavioral) among heart failure patients.

METHODS

This survey was a cross-sectional questionnaire survey among 9,377 eligible patients between January 2024 and April 2025 in 20 public and private healthcare facilities across Punjab, Pakistan. The primary outcome was non-adherence to medications assessed through the Self-Efficacy for Appropriate Medication Scale (SEAMS) and self-reported number of pills. Study variables were summarized using descriptive statistics. Chi-square/Fisher's exact test, independent t-test/ANOVA, and a generalized linear model (GLM) with a logit link function (multivariable logistic regression) were employed to assess associations with demographic characteristics. Seaborn and Matplotlib were used to create a heatmap and an odds ratio plot, respectively.

RESULTS

Non-adherence to medication is alarmingly prevalent among heart failure (Hf) patients in Punjab, Pakistan (50.1%), and the mean SEAMS score was 31.65 ± 7.69, indicating moderate to high levels of self-efficacy in medication-taking behavior. Socio-demographic predictors including female gender (AOR = 1.415, 95% CI: 1.039 - 1.926), primary and secondary education (AOR = 1.285, 95% CI: 1.074-1.481), monthly income 10,000-25000 (AOR = 1.292, 95% CI: 0.616-0.993), no health insurance (AOR = 1.276, 95% CI: 1.138-1.430) had increased odds of non-adherence. Clinical and health-related predictors such as the presence of comorbidities, especially diabetes (AOR = 1.164, 95% CI: 1.0511.428), and non-hospitalized patients (AOR = 1.236, 95% CI: 1.122-1.362).

CONCLUSION

Non-adherence to medication is extremely common in heart failure patients and depends on various sociodemographic, clinical, and behavioral factors. Specific interventions to deal with these determinants can aid in improving adherence and clinical outcomes.