Papers of the Week

Acute postoperative pain (APOP) is often evaluated through granular parameters, though monitoring postoperative pain using trends may better describe pain state. We investigated acute postoperative pain trajectories in cardiac surgical patients to identify subpopulations of pain resolution and elucidate predictors of problematic pain courses. We examined retrospective data from 2810 cardiac surgical patients at a single center. The k-means algorithm for longitudinal data was used to generate clusters of pain trajectories over the first five postoperative days. Patient characteristics were examined for association with cluster membership using ordinal and multinomial logistic regression. We identified three subgroups of pain resolution after cardiac surgery: 37.7% with good resolution, 44.2% with moderate resolution, and 18.2% exhibiting poor resolution. Type I diabetes (2.04 [1.00-4.16], P = 0.05), preoperative opioid use (1.65 [1.23-2.22], P = 0.001), and illicit drug use (1.89 [1.26-2.83], P = 0.002) elevated risk of membership into worse pain trajectory clusters. Female gender (1.72 [1.30-2.27], P <0.001), depression (1.60 [1.03-2.50], P = 0.04) and chronic pain (3.28 [1.79-5.99], P < 0.001) increased risk of membership in the worst pain resolution cluster. This study defined three APOP resolution subgroups based on pain score trend after cardiac surgery and identified factors that predisposed patients to worse resolution. Patients with moderate or poor pain trajectory consumed more opioids and received them for longer before discharge. Future studies are warranted to determine if altering postoperative pain monitoring and management improve postoperative course of patients at risk of moderate or poor pain resolution.