Impact Evaluation of the Therapeutic Initiative's Prescribing Portrait and Therapeutics Letter on Opioid Use by Dentists

Audit and feedback is a well-established strategy for improving the quality of care in primary healthcare settings. It involves the systematic collection and analysis of healthcare data, followed by the provision of information to healthcare providers to improve the quality of care.

Medical overprescribing of opioids significantly contributed to overdose deaths from 1995-2015. Although illicit synthetic opioids now dominate the overdose crisis, prescription opioids remain implicated in 22% of overdoses and 2% of fatal overdoses in British Columbia (BC). Dentists play an important role in opioid exposure: opioids may be prescribed in up to half of dental procedures, and general dentists account for 13% of all opioid prescriptions in BC.

Based in British Columbia, Canada, the Therapeutics Initiative's Portrait program provides clinicians with individualized prescribing data compared with their peers, paired with evidence-based messaging, and a concise statement summarizing the desired change in prescribing behavior. The Portrait program implements a series of pragmatic trials to improve prescribing where clinicians are randomized into an early intervention and a delayed control group.

This study examines the impact of prescribing Portraits on opioid prescribing by dentists in British Columbia. The Portrait title is "Do you manage dental pain with opioids?" and is accompanied by a Therapeutics Letter entitled "Pharmacological management of dental pain and inflammation."

3,498 active dentists in BC will be randomized into 3 Groups/Arms (n=1166 in each Group) for this intention-to-treat analysis.

On Jan 6, 2026 dentists in Groups 1 and 2 will be mailed their intervention bundles (Group 1 the regular intervention and Group 2 the enhanced intervention). Six months later clinicians in Group 2 will receive a repeat of the enhanced intervention, and 6 months after that (12 months after the first intervention), Group 3 (delayed control group) will receive the regular intervention bundle. The analysis will assess whether the materials led dentists in Groups 1 and 2 to change how they prescribe.

Analysis will be conducted using secondary data from administrative databases from the BC Ministry of Health (under an Information Sharing Agreement). The data are patient level and include anonymized information about demographics, health costs, physician billing codes, prescriptions, outpatient records, and hospitalizations. The data do not include personally identifiable patient information.

The trial will include two primary comparisons: (1) dentists randomized to either intervention arm (regular or enhanced intervention) versus dentists randomized to the control arm, and (2) dentists randomized to the regular intervention versus the enhanced intervention. All analyses will be conducted according to the intention-to-treat principle, with participants analyzed in the groups to which they were originally randomized, regardless of intervention exposure or adherence. In addition to primary outcome measures, the analyses will identify baseline characteristics of dentists and patients by study group (age, age groups, sex, urban vs rural) and monthly opioid prescribing trends for all BC dentists from 12 months prior to release of the materials. This will illustrate background trends in prescribing patterns.

Prescribing outcomes will be analyzed using generalized linear mixed models with a negative binomial distribution to account for overdispersion in count data. Models will estimate rate differences and rate ratios between trial arms over time. Random effects will be included to account for repeated observations within included dentists, and 95% confidence intervals will be adjusted for clustering of patients within included dentists. Time will be modeled explicitly to assess changes in prescribing rates before and after intervention implementation.

Prespecified subgroup analyses will include incident opioid prescribing among opioid-naïve patients. Opioid naïve will be defined using a one-year look-back period with no opioid prescriptions prior to the index dental visit (January 5, 2025 to January 5, 2026).

In addition, based on prior evidence from a large Veterans Affairs study demonstrating an association between high antibiotic prescribing and high opioid prescribing among dentists (adjusted odds ratio 8.40, 95% CI 6.00-11.76), the investigators will conduct a prespecified subgroup analysis examining the association between changes in opioid prescribing and changes in antibiotic prescribing following the research team's prior antibiotic prescribing intervention.