Study data suggest that the majority of opioid prescribing is attributable to a small proportion of practitioners, according to results published in the BMJ. In 2017, the top 1% of prescribers accounted for more than a quarter of all opioid prescriptions. However, among the bottom 99% of prescribers, greater compliance with Centers for Disease Control and Prevention opioid prescribing guidelines was observed. As such, investigators recommended that intervention efforts to reduce opioid overprescribing target the highest-dispensing subgroup of clinicians.
Investigators abstracted opioid prescribing data from the Optum Clinformatics Data Mart Database, which comprises deidentified claims from a private insurer for the years 2003 to 2017. Prescribing data included medication type, quantity, length of supply, and date of transaction. Patient- and provider-level data were also extracted, including diagnosis and physician specialty. Two quantities were calculated for each drug and year by provider: dose and number of prescriptions. Daily dose, dose per patient, and dose for each prescription were also calculated. Opioid doses were standardized to morphine milligram equivalents (MMEs), using National Drug Code data. To estimate distributions of overprescribing, providers, patients, and provider-patient pairs were siloed into 100 groups of equal size. Groups were then stratified by centiles of the volume of opioid prescriptions; the top centile (1%) was compared with the median centile (50%). Sensitivity analyses were performed to test the robustness of results against a high prevalence of low-prescribing clinicians, miscoding or extreme overprescribing outliers, and certain overrepresented opioid subsets. In all cases, outcomes remained the same.
Between 2003 and 2017, 8.2 billion MMEs were prescribed to 3.9 million patients in the Optum database. Prescriptions were written by 669,495 prescribers, corresponding to >700 MMEs per day for each provider. Prescribing distributions were greatly skewed, however: the top 1% of clinicians prescribed 748,000 MMEs in 2017, 1000-fold that of the median centile. This subgroup also accounted for 27% of opioid prescriptions and 19% of benzodiazepine prescriptions in the same year. These trends remained significant after adjustments for prescription length, number of patients per provider, and number of prescriptions per provider. Opioid prescriptions have been disproportionately attributable to the top centile since 2003: During the studied period, 1% of practitioners consistently accounted for ≥20% of all opioid prescriptions. Approximately half of all providers who were in the top centile one year were in the top centile of adjacent years. Centers for Disease Control and Prevention guidelines recommend that opioid prescriptions comprised a dose <50 MMEs and a duration <7 days. Of prescriptions written by the top 1%, 42% to 49% were for more than 50 MMEs and 81% to 98% were for longer than 7 days. Among the bottom 99% of prescribers, 86% were for fewer than 50 daily MMEs and 71% were for fewer than 7 days. Among the top centile of opioid prescribers, the most common specialties were family medicine (24%), physical or pain medicine and rehabilitation (14%), anesthesiology (14%), and internal medicine (13%).
Investigators noted that many prescriber-patient pairs in the top centile overlapped in adjacent years, suggesting established relationships. As such, longer durations of opioid prescriptions may have fallen within the natural course of care for long-term patients. In addition, family medicine and anesthesiology, which are specialties overrepresented in the top centile, require additional training on opioid prescribing. Even so, opioid overprescribing trends were overwhelmingly driven by a small subgroup of clinicians for whom targeted interventions may be appropriate.
Kiang MV, Humphreys K, Cullen MR, Basu S. Opioid prescribing patterns among medical providers in the United States, 2003-17: retrospective, observational study. BMJ. 2020;368:l6968.
This article originally appeared on Medical Bag