A retrospective, matched-tool comparison of the opioid risk tool and revised opioid risk tool in community pharmacy screenings.

BACKGROUND

The Opioid Risk Tool (ORT) was compared with the Revised Opioid Risk Tool (ORT-R) in a large community pharmacy population. The ORT-R removes the preadolescent sexual abuse (PSA) item and sex-based scoring weights used in the original ORT.

OBJECTIVE

To compare opioid risk classification between the ORT and ORT-R in community pharmacy screenings, with a focus on reclassification patterns, PSA-related sex differences, and agreement between tools.

METHODS

A retrospective, matched-tool comparison was conducted among 21,295 patients screened with the ORT in community pharmacy settings. Identical patient data were rescored using the ORT-R. Descriptive analyses summarized overall distributions, reclassification patterns, and PSA-related sex differences. Because no external clinical outcomes were available, measures of predictive performance such as sensitivity and specificity could not be assessed.

RESULTS

A total of 90.7% and 93.4% of participants were classified as low risk on the ORT and ORT-R, respectively. The ORT-R produced modest shifts in classification. Approximately 6% of patients classified as moderate on the ORT were reclassified as either low or high risk on the ORT-R. Nearly 80% of women with PSA histories were reassigned under the ORT-R, reflecting the structural removal of the PSA item from scoring, compared with fewer than half of men. The factors most strongly associated with reclassification were family (odds ratio [OR] 16.21, confidence interval [CI] 13.82-19.02) or personal alcohol abuse (OR 6.04, CI (4.85-7.52), depression (OR 4.64, CI 3.99-5.39), and psychiatric disease (OR 4.41, CI 3.63-5.36).

CONCLUSION

These findings suggest potential sex-based differences associated with PSA weighting in the ORT. In the ORT, PSA contributes to scoring for women but not for men, and the item may also trigger survivors during pharmacy-based screening. The ORT-R may represent a simpler scoring approach with similar overall classification patterns, though outcome-based validation is needed.