Background The use of proton pump inhibitors (PPIs) has increased significantly since they were introduced in therapeutics. However, concerns have been raised regarding the appropriateness of these prescriptions.

Purpose To assess the prevalence and appropriateness of PPIs prescribing in patients before admission and at discharge in hospital’s level-two units, with electronic clinical records and access to information of treatment in primary care.

Material and methods Retrospective observational study. Data were obtained during March 2017, by reviewing patients’ discharge reports, electronic clinical records (Jimena) and the pharmacological prescription programme in primary care (Medora) of all hospitalised patients. We analysed the use of PPIs before admission and at discharge, the type of PPI, the reason for its prescription, and whether it was correct or not based on the label indications and the uses recommended in clinical practice guidelines. Epidemiological data and concomitant treatments were also collected. All statistical analysis was performed in an Excel database.

Results A total of 634 patients belonging to both medical and surgical services were included. The mean age was 71.8 (SD=15. 2) years: 58.4% were men. At admission, the patients were taking a mean of 6.4 (SD=3.9) drugs chronically. 61.2% of the patients took PPIs prior to their admission (63.7% omeprazole, 28.4% pantoprazole, 3.9% lansoprazole, 3.0% esomeprazole and 1.0% rabeprazole). In 29.4% of these patients PPIs were not indicated. Of these, 82.2% maintained treatment with PPIs at hospital discharge. On the other hand, at hospital discharge, 65 patients (84.6% omeprazole and 15.4% pantoprazole) initiated a new treatment. In 26.2% of these patients, PPIs were not indicated.

Conclusion PPIs indications should be reviewed before being prescribed because an inappropriate use has shown no benefit and they are not without adverse effects in their long-term use. In addition, their overuse contributes to increasing polypharmacy, drug interactions and health expenditure.

No conflict of interest