Application of STOPP/START criteria in older patients in primary care using RStudio®

To estimate the frequency of potentially inappropriate prescribing (PIP) in outpatients according to STOPP/START criteria, and to identify risk factors. For this purpose, an algorithm was developed and validated in RStudio® based on the information collected in the electronic prescription.

The data corresponds to dispensations from two pharmacies in Spain made to patients over 18 years, over 4 years. For the analysis, only patients aged ≥ 65 years who are targeted by the STOPP/START criteria are included. The statistical programming language RStudio® was used to develop the algorithm. The STOPP criteria used as models for the implementation of the method were L2 and B12. A logistic regression analysis was performed.

A total of 15,601 treatment plans were obtained from 2312 patients ≥ 65 years (56% women), of whom 46.6% had polypharmacy (≥ 5 drugs) and 9.3% had excessive polypharmacy (≥ 10 drugs). In this group, PIPs were detected in 57% of patients and in 38% of their treatment plans; of these PIPs the most common were those related to the use of benzodiazepines for more than 28 days (D5 criterion) in 25.9% of patients, followed by the use of opioids prescribed without an associated laxative (L2 criterion) in 13.8% and finally, drugs duplication (A3 criterion) in 5.7%. The most numerous duplications related to criterion A3 were benzodiazepines (39%) and non-steroidal anti-inflammatory drugs (37.4%).

The method developed and validated in RStudio® with different STOPP criteria allows us to analyse the pharmacological treatment of many patients using different databases and to identify those at risk of suffering a PIP according to the STOPP criteria. Our results indicate a high prevalence of PIPs in patients ≥ 65 years, with polypharmacy being the most common risk factor affecting PIP.