No Easy Way to Tell Who`s at Low Risk for Prescription-Opioid Addiction

By Will Boggs MD

NEW YORK (Reuters Health) - Several individual features can identify patients at high risk of prescription opioid addiction, but identifying patients at low risk remains problematic, according to a systematic review.

"Although much research has been conducted to develop strategies to identify patients who can be safely prescribed opioid analgesics, the risk-screening tools that are in widespread use among clinicians are based on low-quality studies and/or are not helpful," Dr. Jan Klimas from British Columbia Center on Substance Use, University of British Columbia, in Vancouver, Canada, told Reuters Health by email.

Based on a 2015 national survey, more than a third of U.S. adults (91.8 million, 37.8%) used prescription opioids; 12.5% of these reported misuse, and 16.7% of those reporting misuse reported a prescription-opioid-use disorder (OUD).

The U.S. Centers for Disease Control and Prevention (CDC) pain guidelines highlight the importance of screening patients to identify those at high risk of OUD, but how to do that accurately remains unclear.

Dr. Klimas and colleagues assessed the diagnostic accuracy of strategies used for identifying opioid-naive patients at high versus low risk of prescription OUD who were being prescribed opioids for pain.

The six high-quality studies included in the qualitative synthesis and four high-quality studies included in the quantitative synthesis identified a history of any pain disorder, personality disorder, somatoform disorder, psychotic disorder and non-opioid substance-use disorder as being associated with increased risk of prescription-opioid addiction.

In contrast, only the absence of a mood disorder appeared to meaningfully reduce the likelihood of prescription opioid addiction, they report in JAMA Network Open, online May 3.

Opioid prescription characteristics associated with an increased risk of developing OUD included a new prescription for any opioid for 30 days or more (versus a prescription for less than a 30-day supply) and opioid doses greater than 120 morphine milligram equivalents per day.

The Pain Medication Questionnaire (using a cutoff score of 30 or higher) appeared most promising for assessing the risk of prescription OUD, but it poorly differentiated patients at high risk from those at low risk.

No scale proved useful for assessing patients newly presenting with chronic or acute pain, and none of the other risk assessment tools (Opioid Risk Tool, Brief Risk Questionnaire, Brief Risk Interview, and Screener and Opioid Assessment for Patients with Pain) were useful for discerning patients at high versus low risk.

"There are few valid tools to identify pain patients who can be safely prescribed opioids," Dr. Klimas concluded.

He cautioned that these findings apply only to patients with pain who have never been prescribed opioids. The review did not include individuals who already use opioids.

Dr. Weihsuan Jenny Lo-Ciganic of the University of Florida College of Pharmacy, in Gainesville, who recently evaluated a machine-learning algorithm for predicting prescription-opioid overdose risk among Medicare beneficiaries, told Reuters Health by email, "Opioid overdose or opioid-use disorder is a complicated problem involving multiple factors. Relying on simple criteria (e.g., high-dose use) may not be efficient and accurate to identify those truly at high risk for intervention. Built upon our current work, our team is developing an intelligent clinical-decision-support tool to help clinicians to identify patients at risk of opioid overdose or opioid-use disorder."

"We need to simultaneously take multiple relevant predictors including patient sociodemographics, health status factors, provider-level and regional-level factors into account when predicting overdose," said Dr. Lo-Ciganic, who was not involved in the review.

SOURCE: https://bit.ly/2LKJZ2b

JAMA Netw Open 2019.

(c) Copyright Thomson Reuters 2019. Click For Restrictions - https://agency.reuters.com/en/copyright.html
Source: