Clinical practice guidelines are continually evolving with the introduction of new information and new therapies, interventions, and medications for dealing with complex medical conditions or diseases. With so much information to absorb, current efforts to continually update physicians may not always be enough to prevent potentially inappropriate prescribing behaviors. This study tests an audit-and-feedback program for physicians prescribing mental health medications to persons in Manitoba, Canada.
Physicians keeping up to date with new guidelines can be fundamental to good practice. In order to stay informed, the numerous professional colleges that oversee medical practitioners require licensed physicians to engage in continuing medical education (CME) activities to ensure continued development and familiarity with current practice. However, with so much information to absorb, CME may not always be enough to prevent potentially inappropriate prescribing behaviors.
In April 2011, the Manitoba Government introduced the Manitoba IMP℞OVE Program, administered by Comprehensive NeuroScience of Canada (CNSC), a subsidiary of Care Management Technologies (CMT). Fifteen Quality Indicators (QIs) for potentially inappropriate prescribing were included in the program. Drug-dispensation data from community pharmacies were analyzed, and when a QI was triggered, an educational package was mailed to the prescribing physician.
Implementation of the program proceeded in two waves. Half of eligible physicians were assigned randomly to an intervention group who began to receive educational packages about their prescribing behaviors immediately. The other half of eligible physicians were scheduled for delayed implementation; they did not begin to receive educational packages until over one year later. This half functioned as a control group in the evaluation. A primary set of QIs concentrated on prescription behaviors for benzodiazepines (six QIs) and anti-insomnia agents (two QIs). A group of secondary QIs initiated six months after the first group concentrated on other types of drugs (opioids, selective serotonin reuptake inhibitors, antidepressants, and antipsychotics).
Table 1: Select Quality Indicators Included in the Manitoba IMP℞OVE Program Evaluation, 2011–2013
Three out of four of the highest-frequency QIs (105: Benzodiazepines for adults, 138: Long-acting benzodiazepines for older adults, 211: Anti-insomnia agents for adults) showed a significant effect of IMP℞OVE. Quality indicators with intermediate trigger frequency had mixed results; two showed a significant improvement and three did not show an improvement. The remaining QI trigger rates that did not show a statistically significant effect of the program were triggered at a low rate. Many of the high-frequency primary QIs targeted by the program showed a clear significant reduction in trigger rates as a result of the educational mailing packages.
Table 2: Summary of Intervention Effect and Number of Quality Indicator Triggers in the Manitoba IMP℞OVE Program, 2011-2013
This study demonstrated that the Manitoba IMP℞OVE Program has been effective at reducing many of the potentially inappropriate prescribing behaviors it targeted. For policy, the current study provides excellent information for the expansion or alteration of this audit-and-feedback program, or the implementation of other audit-and- feedback programs by Manitoba Health, Healthy Living and Seniors , perhaps targeting behaviors other than prescriptions. The result of this analysis clearly points to the type of QIs that are likely to be successful—high- or moderate-frequency QIs in the sole control of the prescribing physician—where success is gauged as a measurable change in prescribing behaviors.