Maurice Würsten, Carla Meyer-Massetti, Aljoscha N Goetschi
Switzerland's hospitals rarely implement the seven types of OSPs identified, suggesting that more coordinated approaches to improving opioid safety are needed. Our interpretations should be treated cautiously, nevertheless, due to potential non-response bias.
OBJECTIVES: Opioid prescription counts are increasing across Europe and in Switzerland, yet inappropriate co-prescribing of opioids is common, with serious unintended clinical effects. There is limited information about the opioid stewardship programmes (OSPs) implemented in Europe and Switzerland. The present study surveyed OSPs in Switzerland's hospitals to characterize implemented OSPs and assessed their association with staffing.
METHODS: A 30-item questionnaire was sent to the director of every pharmacy in Switzerland's national hospital pharmacy association (n = 69). It addressed the pharmacy-reported use of seven types of OSP previously identified in the literature: automated decision support, opioid use guidelines, staff training, discussions of opioid issues/questions in committees, the availability of naloxone, interventions for appropriate opioid use and the monitoring of consumption.
KEY FINDINGS: Thirty-one of the 69 institutions completed the survey (45%) and a further nine responded partially (13%). Hospitals had implemented a median of three of the seven OSPs covered (interquartile range: 2-5). Infrequently implemented OSPs included alert systems to detect high-risk situations involving opioids (15%), opioid use guidelines (41%), education on safe opioid use (32%) and monitoring of opioid or naloxone use (42%). Committees on safe opioid use (59%) and clinical pharmacy interventions addressing safe opioid use (62%) were more frequently implemented. Naloxone was rapidly available in all institutions.
CONCLUSIONS: Switzerland's hospitals rarely implement the seven types of OSPs identified, suggesting that more coordinated approaches to improving opioid safety are needed. Our interpretations should be treated cautiously, nevertheless, due to potential non-response bias.