Temporal Trends and Factors Associated with Medication Prescription Patterns in Peritoneal Dialysis Patients.

INTRODUCTION: Patients on peritoneal dialysis (PD) suffer from a high burden of comorbidities, which are managed with multiple medications. Determinants of prescription patterns are largely unknown in this population. This study assesses temporal changes and factors associated with medication prescription in a nationally representative population of patients on PD under the universal coverage healthcare system in Brazil.

METHODS: Incident patients recruited in the Brazilian Peritoneal Dialysis Study (BRAZPD) from December 2004 to January 2011, stratified by prior hemodialysis (HD) treatment, were included in the analysis. Multivariable logistic regression was used to assess the association between medication prescription and socioeconomic factors. Yearly prevalent cross-sections were calculated to estimate prescription over time.

RESULTS: Medication prescription was in general higher among patients who had previously received HD, compared with those who started renal replacement therapy (RRT) directly on PD. Prescription increased from baseline to 6 months of PD therapy, particularly in those who did not previously receive HD. After accounting for patient characteristics, significant associations were found between socioeconomic factors, geographic region, and medication prescription patterns. Finally, the prescription of all cardioprotective and anemia medications and phosphate binders increased significantly over time.

CONCLUSION: In a PD population under universal coverage in a developing country, there was an increase in drug prescription during the first 6 months on PD, and a trend toward more liberal prescription of medications in later years. Independent from patient characteristics and comorbidities, socioeconomic factors influenced drug prescriptions that likely impact patient outcome, calling for public health action to decrease potential inequities in management of comorbidities in PD patients.