Are Proton Pump Inhibitors Overutilized in Hospitalized Patients Thereby Increasing Their Risk of Adverse Effects? A Community Hospital Experience

Background:

Proton pump inhibitors (PPIs) are one of the most prescribed medications and are frequently used in hospitalized patients. Although PPIs have well‐defined indications for both therapy and prophylaxis, they are often overutilized. Recently, there has been increasing evidence of potential adverse effects of PPIs such as decrease absorption of thyroxine, vitamin B12, calcium, and iron supplements; decreased effectiveness of clopidogrel; and increased risk of community‐acquired pneumonia (CAP), Clostridium difficile infection (CDI), and hip fractures. Based on these concerns, we questioned whether our patient population was being prescribed PPIs appropriately and if that group had increased incidence of side effects.

Methods:

We performed a retrospective chart review of adult patients during 1 month and then followed all subsequent admissions for 1 year. We reviewed: frequency of prescribing, indications based on well‐defined criteria, if PPIs were continued at discharge, if they were prescribed to patients with the potential for interaction, and whether their use was associated with CAP, CDI, or hip fractures.

Results:

We evaluated 150 patients of whom 106 (70%) were on PPIs. Of these 106 patients, 32 were on PPIs prior to admission, with only 16 (50%) of those having standard indications. The remaining 74 patients were started on PPIs during hospitalization; of those, 24 patients (22.6%) were started for stress ulcer prophylaxis, with 21 (37.5%) having nonstandard indications. The remaining 50 were started for other complaints, with 34 (68%) having nonstandard indications. Of 106 patients on PPIs, 35 (33%) were identified at risk for adverse interactions, 25 (71%) related to decreased absorption and 10 (28%) related to decreased effectiveness of clopidogrel. Sixty‐eight of 150 (4‐5%) were readmitted, of whom 49 (72%) were using PPIs, 28 (41%) being started during previous admission. We did not lind any adverse event documented, as the records did not reveal any effect of PPIs on absorption or drug‐drug interaction.

Conclusions:

PPIs were widely used in our hospitalized patients and often for ill‐defined indications, supporting our hypothesis that they are being overutilized. Patients are frequently placed on PPIs for “prophylaxis,” and surprisingly, even if not indicated, these medications were continued on discharge. Many of the patients that were placed on PPIs were at potential risk of adverse effects. A concerning feature was that there were no indications in the records suggesting that the treating physicians were aware of these potential interactions. We did not find any of the expected complications of PPIs. In summary, we found an extraordinarily high utilization of PPIs (70%), often with little indication, that were continued in patients even in the face of apparently underrecognized potential adverse effects.

Author Disclosure:

B. Rizvi, none; J. Hanley, none.