Validation of prescriber risk indicators obtained from prescription drug monitoring program data

Background: Prescription opioids are commonly overprescribed. However, validated measures of inappropriate controlled substance prescribing are lacking. This study examined associations between prescriber risk indicators developed as part of a public health surveillance project and medical board disciplinary actions against prescribers.

Methods: We compiled 12 prescriber risk indicators using data from the Maine prescription drug monitoring program (PDMP) for 2010. We used logistic regression models to assess the relative likelihood of the top 1%, 2%, 5%, and 10% of prescribers on each risk indicator having been subject to medical board disciplinary actions, those citing inappropriate prescribing, or those involving license suspension or revocation, during 2010-2014, controlling for prescriber medical specialty and gender.

Results: The top 1% of prescribers for number of patients, opioid prescriptions per day, and opioid dosage prescribed per day had a greater likelihood of medical board disciplinary actions citing inappropriate prescribing, relative to a matched sample of other (non-top 1%) prescribers. Of the 56 prescribers in the top 1% for opioid prescriptions per day, nine (16.1%) were sanctioned for inappropriate prescribing, compared with 11 of 224 (0.5%) in the comparison group. The top 2% of prescribers for opioid dosage per day, and average distance patients travel to prescriber, had a greater likelihood of actions involving license suspension, revocation, or denial for renewal.

Conclusions: Measures derived from PDMP data may be useful in assessing levels of inappropriate prescribing of controlled substances in a population of prescribers, and in evaluating changes associated with efforts to influence prescriber behavior.