Use of proton pump inhibitors with lack of diagnostic indications in 22 Midwestern US skilled nursing facilities.

OBJECTIVES: The primary objective of this study was to identify proton pump inhibitor (PPI) prescribing patterns in a population of older adults admitted to 22 Midwestern skilled long term care facilities (LTCF) with medical coverage provided by the US Medicare Part A program. The relationship between PPI prescribing patterns and specific ICD-9 diagnostic codes and symptoms management was examined. The long-term objective is appropriate PPI prescription guidance through the development of evidence- and regulation-based pharmacy formulary and policy practices, as well as practical prescribing guidance for practitioners who are supported by this pharmacy.

DESIGN: An observational cohort study was conducted, using prospectively collected and de-identified prescribing and diagnostic data from a convenience sample of all Medicare A skilled nursing patients admitted between January 1, 2010, and May 31, 2011, to 22 urban, suburban, and rural Midwestern US LTCFs.

SETTING AND PARTICIPANTS: A common pharmacy service de-identified and aggregated PPI prescribing data and patient diagnostic information. These secondary data were analyzed for trends and patterns related to PPI use for all Medicare A patients admitted to these 22 facilities during a 17-month period in 2010 and 2011.

MEASUREMENT AND RESULTS: Rates of PPI use were determined and were compared with diagnostic codes. Of 1381 total admissions, 1100 patients (79.7%) were prescribed PPI. There was no appropriate diagnosis for PPI use in 718 patients (65.3%). Gastroesophageal reflux disease (GERD) tended to be the blanket diagnosis that was used most frequently for PPIs, but there was usually no follow-up or symptomatic evidence documented of active GERD. When long-term (current) use of nonsteroidal anti-inflammatory drugs (NSAIDs) (including aspirin) and/or anticoagulant therapy (warfarin) was considered as appropriate indications for 382 patients, 336 (24%) of all Medicare patients were still receiving PPIs with no relevant gastrointestinal ICD-9 diagnostic code. Total cost of PPIs prescribed from January 2010 to June 2011 was $348,414.

CONCLUSIONS: The examined PPI prescribing patterns show discordance between ICD-9 diagnostic code and prescribed use of PPIs in the study population. More than half (52%) of the total number of Medicare A patients were taking the medication without an indicated diagnosis. Even when NSAIDs and anticoagulant therapy were taken into consideration as valid reasons for PPI use, 24% of all patients admitted were still prescribed PPIs without a diagnosis that indicated the need for a PPI. Considering the economic cost, potential side effects, and CMS F329 regulations, which require that an LTCF resident's drug regimen be free from unnecessary medication, it is important that prescribers in LTCFs carefully consider use of PPIs in older adults in LTCFs and monitor the continued use of PPIs to prevent both the personal cost of physical side effects and drug-drug interactions, as well as the economic cost of unnecessary medication use.