Outcomes of off-label drug uses in hospitals : a multicentric prospective study

Altres ajuts: We would to thank the Spanish Ministry of Health, Social Affairs and Equality for their financial support given though a grant from the scholarship EC-206 in the public call for the promotion of independent clinical research (SAS/2370/2010 order of September 27) The study aims to assess the clinical evidence, outcome and cost of off-label use of medicines in the hospital setting. A multicentric prospective cohort study of patients treated with off-label medicines was carried out in five tertiary hospitals from May 2011 to May 2012. Information on clinical characteristics of patients, drugs, outcomes and costs was collected. Patients were followed up to 6 months, and information was assessed by reviewing clinical records and interviewing physicians. A total of 226 patients were included. The median (interquartile range (IQR)) age of patients was 46 (33-62) years; 59 % were women. Patients had received a median of three previous treatments, and a lack of response (or suboptimal) was the main reason for off-label use (72.1 %). A total of 232 off-label medicines were administered for 102 different indications. The most frequent medicines were rituximab (49; 21.1 %), botulinum toxin (25; 10.7 %) and omalizumab (14; 6.0 %). In 117 (51.8 %) cases, the level of clinical evidence for their use was low. A partial clinical response was observed in 82 patients (36.3 %), complete response in 71 (31.4 %) and stabilization in 11 (4.9 %). A total of 58 (26.5 %) patients had adverse effects, which in 11 (4.9 %) were severe. The median (IQR) cost per patient was €2,943.07 (541.9-5,872.54). There was a high variability of off-label medicines and indications. Although the clinical evidence of off-label medicines was often low, clinical response was observed in many patients with previous multiple treatment failure, but at the expense of some adverse effects and a high cost. Registers of patients would be helpful for clinical decisions, although clinical trials are needed. The online version of this article (doi:10.1007/s00228-014-1746-2) contains supplementary material, which is available to authorized users.