Search

Although ketorolac is an effective adjunct for managing pain in the perioperative period, it is associated with a risk of postoperative bleeding. This study retrospectively investigated the association between ketorolac use and both reoperation and postoperative opioid use among mastectomy patients. The study identified all women undergoing mastectomy (unilaterally or bilaterally) at our ambulatory surgery cancer center from January 2016 to June 2019. The primary outcome was reoperation for bleeding on postoperative day 0 or 1, and the secondary outcome was postoperative opioid use. The association between ketorolac and outcomes was assessed using multivariable regression models. The covariates were age, body mass index, breast reconstruction, bilateral surgery, peripheral nerve block, and preoperative antiplatelet and/or anticoagulation medication. A cohort of 3469 women were identified. Ketorolac was given to 1549 (45%) of the women, with 922 women (60%) receiving 30 mg and 627 women (40%) receiving 15 mg. The overall reoperation rate for bleeding was 3.1% (1.8% without ketorolac vs 4.8% with ketorolac). In the multivariable analysis, ketorolac was associated with a higher risk of reoperation [odds ratio (OR) 2.43; 95% confidence interval (CI) 1.60–3.70; P < 0.0001]. Ketorolac also was associated with a lower proportion of patients receiving any postoperative narcotic within 24 h (15 mg: OR 0.73; 95% CI 0.57–0.94; P = 0.014 vs 30 mg: OR 0.52; 95% CI 0.42–0.66; P < 0.0001). Ketorolac use decreased postoperative opioid use, but this benefit was outweighed by the increased risk of bleeding requiring reoperation. This finding led to a change in practice at the authors’ center, with ketorolac no longer administered in the perioperative care of the mastectomy patient.

Background A variety of statistics have been proposed as tools to help investigators assess the value of diagnostic tests or prediction models. The Brier score has been recommended on the grounds that it is a proper scoring rule that is affected by both discrimination and calibration. However, the Brier score is prevalence dependent in such a way that the rank ordering of tests or models may inappropriately vary by prevalence. Methods We explored four common clinical scenarios: comparison of a highly accurate binary test with a continuous prediction model of moderate predictiveness; comparison of two binary tests where the importance of sensitivity versus specificity is inversely associated with prevalence; comparison of models and tests to default strategies of assuming that all or no patients are positive; and comparison of two models with miscalibration in opposite directions. Results In each case, we found that the Brier score gave an inappropriate rank ordering of the tests and models. Conversely, net benefit, a decision-analytic measure, gave results that always favored the preferable test or model. Conclusions Brier score does not evaluate clinical value of diagnostic tests or prediction models. We advocate, as an alternative, the use of decision-analytic measures such as net benefit. Trial registration Not applicable.