Safety profile of raloxifene as used in general practice in England: results of a prescription-event monitoring study.

Raloxifene, a selective estrogen receptor modulator (SERM) licensed for the prevention of non-traumatic vertebral fractures in postmenopausal women at increased risk of osteoporosis, was launched in the UK in August 1998. The aim of the study was to monitor the safety of raloxifene prescribed in the primary care setting in England using prescription-event monitoring (PEM). Patients were identified by means of prescription data supplied by the Prescription Pricing Authority between September 1998 and November 2000. Demographic and clinical event data were collected from questionnaires posted to primary care physicians (GPs) at least 6 months after the date of the first prescription for each patient. Information on medical events, suspected adverse drug reactions (ADRs), reasons for stopping treatment, pregnancies, and causes of death was requested. Event rates [Incidence Densities (IDs): no. first reports /1000 patient-months of treatment] were calculated. Differences between IDs for events reported in month one (ID(1)) and months 2-6 (ID(2-6)) of treatment were examined. The cohort comprised 13,987 patients [median age 62 years (IQR 55,69); 99.8% female]. The major indication was osteoporosis (40.9%, n=5725). Flushing was the event with the highest ID in month 1 (22.8), reported most frequently by GPs as an ADR to raloxifene (67/461 reports) and as the reason for stopping (700/4592 reports). Events associated with starting treatment included flushing, malaise/lassitude, headache/migraine, nausea/vomiting, sweating, cramp, pain abdomen, dizziness, diarrhea, mastalgia and vaginal hemorrhage. Less common events reported during treatment included deep vein thrombosis (n=13), pulmonary embolism (n=13), thrombophlebitis (n=31) and visual disturbance (n=29). In this study, there were 122 (0.9%) confirmed deaths, of which 32 causes of death were unknown. This study shows that raloxifene is generally well tolerated when used in general practice in England. Potential signals of unrecognised ADRs requiring further evaluation included gastrointestinal adverse symptoms and vaginal hemorrhage. There were also a small number of reports of events associated with venous thromboembolism and visual disorders that require further investigation.