Prescribing patterns and medication costs in patients on maintenance haemodialysis and peritoneal dialysis

Abstract

ABSTRACT Background Polypharmacy is a significant clinical issue for patients on dialysis but has been incompletely studied. We investigated the prevalence and costs of polypharmacy in a population-based cohort of participants treated with haemodialysis (HD) or peritoneal dialysis (PD). Methods We studied adults ≥20 years of age in Alberta, Canada receiving maintenance HD or PD as of 31 March 2019. We characterized participants as users of 0–29 drug categories of interest and those ≥65 years of age as users/non-users of potentially inappropriate medications (PIMs). We calculated the number of drug categories, daily pill burden, total annual cost and annual cost per participant and compared this to an age- and sex-matched cohort from the general Alberta population. Results Among 2248 participants (mean age 63 years; 39% female) on HD (n = 1781) or PD (n = 467), the median number of prescribed drug categories was 6 [interquartile range (IQR) 4–8] and the median daily pill burden was 8.0 (IQR 4.6–12.6), with 5% prescribed ≥21.7 pills/day and 16.5% prescribed ≥15 pills/day. Twelve percent were prescribed at least one drug that is contraindicated in kidney failure. The median annual per-participant cost was ${\$}$3831, totalling ≈${\$}$11.6 million annually for all participants. When restricting to the 1063 participants ≥65 years of age, the median number of PIM categories was 2 (IQR 1–2), with a median PIM pill burden of 1.2 pills/day (IQR 0.5–2.4). Compared with PD participants, HD participants had a similar daily pill burden, higher use of PIMs and higher annual per-participant cost. Pill burden and associated costs for participants on dialysis were >3-fold and 10-fold higher, respectively, compared with the matched participants from the general population. Conclusion Participants on dialysis have markedly higher use of prescription medications and associated costs than the general population. Effective methods to de-prescribe in the dialysis population are needed.