The opioid-prescribing practices of Australian general practice registrars: an interview study

Background: Approximately half of the opioids prescribed by Australian GP and GP registrars are for chronic non-cancer pain-despite limited therapeutic benefit, and serious risks of harm. Understanding the factors driving non-evidence-based opioid prescribing may improve GP training and education. Objective: To explore attitudes, beliefs, knowledge and self-reported factors influencing the opioid-prescribing decisions of Australian GP registrars. Methods: Telephone interviews were undertaken with 20 GP registrars in 2018-19. Interviews were 30-60 minutes in duration, audio-recorded and de-identified. Braun and Clarke's 6-phase framework was adopted for reflexive thematic analysis of data and managed using QSR NVivo software. Results: Twenty registrars were recruited; 8 men and 12 women.Three themes were identified. Support and supervision: Difficult chronic pain consultations negatively affected the registrar well-being. Registrars role modelled their supervisors' opioid-prescribing practices, even if they perceived it to be unsafe. Confidence: Registrars lacked confidence in initiating, prescribing and weaning opioids, recognizing drug-seeking behaviours and declining to prescribe-but felt confident in their knowledge of opioid pharmacology. Safety: Registrars were aware of evidence-based prescribing recommendations and risk reduction strategies but struggled to translate this into practice. Conclusions: Non-evidence-based opioid prescribing by Australian GP registrars is multifactorial. Emotionally difficult pain consultations, poor supervision and low prescriber confidence may contribute to unsafe prescribing. Improving registrar prescribing may require interventions to improve risk reduction, training in communication and role modelling by supervisors.