Background At Lamentin Hospital Centre in Martinique, prescriptions are analysed by pharmacists in two clinical units, twice a week. This routine analysis leads to pharmaceutical interventions (PIs) defined by the French Society of Clinical Pharmacy (SFPC). Purpose To describe PIs over 6 months.

Materials and methods A prospective study that included all patients hospitalised in two units, general medical and geriatric, was conducted from 12 November 2012 to 24 April 2013. The pharmaceutical analysis was made following official guidelines with laboratory and clinical data. The detection of interactions was made in a qualified Access database, importing the Thesaurus of the National Agency for the Safety of Medicines and Health Products (MSNA). The PIs were recorded in an Excel file and classified according to the SFPC classification. Doctors were either notified of PIs by phone or in writing. Medical acceptance was defined as changing the prescription.

Results In 19 weeks, pharmacists analysed 940 prescriptions for 658 patients (mean age 70 ± 14 years, 49% male), which led to 58 PIs (6%). The mean number of medicines per patient was 8.5 with more in geriatrics (9). The main problems were drugs interactions (38%), then contra-indication (31%), overdose (14%), indication not treated (10%) and inappropriate administration (5%). Four levels of constraint are defined by MSNA: Contraindication, Disadvised association, Precaution for use and To take into account. The majority of interactions we found fell under Take into account (57%), then Precaution for use (30%) and Disadvised association (11%). Among the severe potential interactions we highlighted the following risks: 35% risk of hyperkalaemia, 17% risk of central nervous system depression, 16% risk of haemorrhage and 14% risk of hypotension. The most frequently drugs involved were: antihypertensives (31%), antithrombotics (16%), analgesics (12%), antibiotics (6%) and electrolytes (6%). The solutions most often suggested by pharmacists were drug switches (31%), dose adjustment (29%), therapeutic drug monitoring (14%), optimisation of administration (14%) and drug discontinuation (12%). In 19% (11/58) of PIs hyperkalaemia had a clinical relevance and was managed. A high rate of PIs (72%) was accepted by the prescribers.

Conclusions This study shows the importance of the pharmacist’s role in detecting potential drug interactions and drug-related problems for adjusting patient treatment. The pharmaceutical validation of the prescriptions and close collaboration with physician should improve the quality, safety and efficacy of patient care. Another study should evaluate the impact of PIs on medical prescriptions to reduce inappropriate prescribing of medicines.

No conflict of interest.