Deprescribing Tool to Reduce Polypharmacy in Hemodialysis Patients

Polypharmacy is the “use of multiple medications or the use of more medications than are medically necessary.” There are associations between polypharmacy and decreased adherence to medication regimens as well as increased risk for adverse effects, falls, hospital admissions, and mortality. Among patients on maintenance hemodialysis, the estimated average number of medications is 12 per day, the highest pill burden of all chronically ill populations, putting hemodialysis patients at increased risk for polypharmacy. Further, because patients on hemodialysis are rarely involved in clinical trials, there are few data available regarding the efficacy and safety of multiple drug therapies in that patient population.

Deprescribing is one strategy used to reduce polypharmacy and suboptimal medication use. Deprescribing is defined as “the process of tapering, stopping, discontinuing, or withdrawing drugs, with the goal of managing polypharmacy and improving outcomes.” In studies of deprescribing among elderly patients, results suggest that decreasing the number of medications taken each day is associated with decreased morality, fewer referrals to nursing homes, and lower drug costs; there were no long-term adverse outcomes associated with deprescribing.

Caitlin McIntyre, PharmD, and colleagues, designed a quality improvement study to develop a deprescribing tool for use among hemodialysis patients and to examine its ability to reduce polypharmacy while maintaining patient safety and satisfaction. Results were reported in the American Journal of Kidney Diseases [2017;70(5):611-618].

The study included 240 patients in a tertiary-care outpatient hemodialysis unit. The primary outcome of interest was the proportion of target medications completely deprescribed after 4 weeks. Secondary outcomes included the proportion of target medications completely deprescribed after 6 months, the average number of medications per patient before and after deprescription, and the proportion of successful deprescriptions for each target medication. Patient safety and satisfaction were monitored with drug-specific monitoring parameters.

Phase 1 involved the development of deprescribing algorithms. From a list of eight potential medications, the nephrology team selected five medications to be identified for deprescription: quinine, diuretics, alpha-1 (a-1) blockers, proton pump inhibitors (PPIs), and 3-hydroxy-3-methylglutaryl-coenzyme (HMG-CoA) reductase inhibitors (statins). Diuretics and statins were selected due to a documented lack of efficacy in hemodialysis patients, and quinine, a-1 blockers, and PPIs were included because of safety concerns when used in patients on hemodialysis.

Phase 2 refined the deprescribing algorithms. The researchers met with three groups of five different nephrology team members who were testing the tool and, based on feedback and comments, revised the tool’s flow and content. Following three cycles of testing and feedback, the algorithms in the deprescribing tool were finalized.

Phase 3 included implementation of the deprescribing tool and assessment of its impact. During the period between May 2014 and March 2015, 71% of participants (n=171/240) were identified as taking at least one of the five target medications. Using the deprescribing tool, the researchers determined that 100 of the 171 patients were taking a target medication for a valid indication, defined by the deprescribing algorithm. It was recommended that the remaining 71 patients undergo deprescription. Twelve of the 71 were excluded from the study due to their absence during deprescribing rounds.

The medical team accepted 40 of the 68 algorithm recommendations; 35 patients were included in the study. Average age was 65 years, 87% had been undergoing hemodialysis for >1 year, and 71% were taking >10 medications. Reasons for not accepting the algorithm’s recommendation to deprescribe included those regarding diuretics: 68% of diuretics were not deprescribed due to the disagreement of the nephrology team with the recommendations when patients were still producing urine.

After 4 weeks, 78% (n=31/40) of target medications were deprescribed in 35 patients. Six months following deprescription, only five of the 31 discontinued medications were represcribed. The average number of medications per patient decreased from 13.4 to 12.7 in 35 patients after 4 weeks, and to 12.8 at 6 months. No patient increased their medication total. At the end of the study, 57% of patients were taking fewer medications compared with baseline. There were no adverse events reported.

The least likely algorithm recommendation to deprescribe followed was diuretics: of the 31 patients identified by the algorithm taking a diuretic, only 10 were included in the study. During the study period, nine of the 10 patients had the medication successfully deprescribed. At the end of the study, only one diuretic was reinitiated.

Study limitations cited by the authors included the single-center study design that relied on self-reporting of medication use and adherence to deprescribing recommendations. Further, there was no formal comparison of patients’ clinical outcomes or whether adherence to remaining medications increased, and no measure of patient satisfaction. Finally, there was potential for bias in patient selection because the researchers excluded patients for whom medications were indicated, as defined by the deprescribing algorithm.

In conclusion, the researchers said, “Our findings have broad applicability because we identified commonly prescribed medications in the hemodialysis population. To our knowledge, this is the first report of a deprescribing tool that has been developed and implemented in an outpatient hemodialysis unit. Polypharmacy is a significant problem in hemodialysis patients as a consequence of their high pill burden and the lack of safety and efficacy data for many medications that are used. This study showed that deprescribing algorithms for quinine, PPIs, a1-blockers, statins, and diuretics could effectively and safely be used in this patient population to help reduce polypharmacy. We hope our experience will encourage others to incorporate similar tools into their practice to help decrease polypharmacy and the adverse outcomes associated with polypharmacy in the hemodialysis population.”

Takeaway Points

  1. Polypharmacy in patients receiving maintenance hemodialysis is associated with adverse patient outcomes and decreased medication adherence. Researchers developed and tested an algorithm to aid in deprescription of medications in that patient population.
  2. The deprescribing tool was used for medications in five classes: quinine, diuretics, alpha-1-blockers, proton pump inhibitors, and HMG-CoA reductase inhibitors (statins).
  3. After application of the tool, 31 of 40 medications were completely deprescribed. At the end of the study, 57% of participants were taking fewer medications than at baseline.