Prescription drug monitoring programs (PDMP) identify medication-related risks to support clinical decision-making. This study aims to report prescription medication refusal rates based on PDMP alerts and determine how PDMP alerts and other characteristics influence pharmacists' decisions to supply opioids. Pharmacists completed an online survey and randomised controlled factorial experiment. Generalised linear models explored differences in the proportion of refusals by state. Pharmacists indicated their likelihood to supply opioids based on six clinical vignettes. Mixed-effect linear regression models explored the association between vignette and pharmacy-related characteristics and the likelihood to supply opioids. Data related to 598 pharmacists (n = 3370 vignettes). Jurisdiction was associated with refusals; Western Australian pharmacists had significantly lower odds of refusing supply compared with Victorian pharmacists (odds ratio = 0.186, 95% confidence interval 0.104-0.331). The factorial experiment revealed the strongest predictors of reduced likelihood to supply were the PDMP high dose (β = -2.76, p < 0.001) and multiple prescriber (β = -3.79, p < 0.001) alerts. Unemployment (β = -0.0421, p < 0.001), hepatitis C (β = -0.260, p = 0.009), depression (β = -0.301, p = 0.003), high opioid dose (β = -0.259, p = 0.002) and co-prescribed opioids with benzodiazepines (β = -0.478, p = 0.001) resulted in smaller reductions in the likelihood to supply. Older patient age, patient familiarity and rural/remote pharmacies were associated with significant, albeit small increases in the likelihood to supply with respective 0.293-, 0.250- and 0.339-unit increases. Jurisdictional differences in refusal to supply were observed, while PDMP alerts were the strongest predictor of reduced likelihood to supply opioids in the factorial experiment. Unintended consequences of PDMPs including abrupt opioid discontinuation observed elsewhere, should be avoided following the implementation of this supply-side policy in Australia.
Read full abstract