JAMA Intern Med. Published online July 17, 2017. doi:10.1001/jamainternmed.2017.2468
Prescription opioid misuse is a leading public health problem and likely contributes to overdoses. The evidence base supporting interventions to change clinical practice at the level of primary care physicians with respect to opioid prescribing is limited. For example, the effect of non-controlled interventions may be difficult to interpret in the setting of local, state and national level initiatives to reduce inappropriate opioid prescribing.
Lead investigator Jane Liebschutz and others from the Section of General Internal Medicine at Boston Medical Center conducted a cluster-randomized trial of 53 primary care providers and 985 of their patients categorized as long-term opioid users within 4 safety net clinics. Patients had a mean age of 55 and 62% had a history of a mental health diagnosis. 45% were insured by Medicaid. Over sixty percent of patients were using between 0 to 50mg of morphine equivalent doses prior to the start of the intervention. This dose is generally considered to be associated with a lower risk of drug overdose.
Opioid Intervention Methods
The trial was designed to test the efficacy of a multicomponent intervention called TOPCARE that included 1-on-1 academic detailing, nurse care management, an electronic registry and electronic decision tools. Control PCPs received the electronic decision tool only. The detailing component was a single session between one of the study coauthors (considered opioid prescribing experts) and the PCP. Topics discussed included: principles of safe opioid prescribing, monitoring, registry reports, details on patient monitoring and risk and challenging patient cases. The nurse care manager was a credentialed nurse who performed risk assessments, prepared prescriptions for PCPs, collected urine drug tests, pill counts, and checked the prescription drug monitoring program on behalf of PCPs. The web-based registry allowed data to be imported from the electronic health record and produced reports used to direct work flow. Furthermore, the registry provided feedback to the program, for example, by allowing study staff to see a PCP’s percentage of patients with an opioid use agreement. Finally, the decision tool included tools for assessment of opioid risk and interactive tools to assist with interpretation of urine drug testing results (a challenging clinical topic).
Study investigators found that the intervention improved the rate of adherence to recent opioid treatment guidelines. Specifically, at 12 months, patients randomized to the intervention were more likely to have a opioid-use agreement (OR 11.9) and have undergone at least 1 urine drug screen (OR 3.0). Unfortunately, the intervention did not reduce the probability of a proxy for opioid misuse: early refills, defined as a prescription refill for an opioid given more than 3 days before the next expected refill date. In exploratory analysis, they found that intervention patients had lower total morphine equivalent doses (6.8mg lower than controls, p<.001).
In conclusion, this study found that a multicomponent intervention including academic detailing improved prescription opioid monitoring compared with electronic decision tools alone in a randomized trial of 4 urban safety-net primary care practices.