AI-assisted pain monitoring: evaluation of pain, analgesics use, and opioid dose trends in combat-related trauma patients

Background. Combat-related injuries are frequently associated with intense acute pain, complex analgesic regimens, and an elevated risk of prolonged opioid consumption. Artificial intelligence (AI)-driven tools such as the Rapid Pain Assessment Tool (R-PAT) offer novel opportunities for real-time monitoring and early prediction of long-term pain outcomes, including sustained opioid use. Materials and methods. A prospective cohort of 33 combat-injured patients (mean age 32.4 ± 7.8 years; 91 % male) with blast (52 %), gunshot (33 %), or drone-related (15 %) trauma were monitored using R-PAT across eight timepoints from baseline to six months post-injury. Three indices were evaluated: pain index (PI), pain medication index (PMI), and oral morphine equivalent daily dose (oMEDD). Longitudinal trends were assessed using one-way ANOVA. Predictive accuracy of baseline metrics for high oMEDD at 6 months (above-median threshold) was evaluated using receiver operating characteristic (ROC) analysis. Results. All three indices demonstrated statistically significant reductions over time, reflec­ting effective pain control and tapering of analgesic interventions (PI: p < 2.0 10–45; PMI: p < 5.4 10–80; oMEDD: p < 3.5 10–71). ROC analysis of AI-enhanced baseline data revealed robust predictive performance for identifying patients with high opioid requirements at six months: PI_simulated — moderate predictive value (AUC ≈ 0.66–0.72), PMI_simulated — strong predictive value (AUC ≈ 0.75–0.80), oMEDD_simulated — very strong predictive value (AUC ≈ 0.80–0.85). This indicates that baseline opioid dosing patterns, when processed through AI-enhanced models, carry substantial prognostic value. oMEDD_simulated was the most accurate early predictor of sustained opioid use, outperforming both subjective pain scores and analgesic complexity. Conclusions. The R-PAT system proved effective not only in tracking pain and opioid usage over time but also in forecasting long-term opioid requirements. AI-augmented baseline metrics demonstrated strong predictive potential, supporting their use in the early identification of high-risk patients. These findings advocate for the integration of intelligent monitoring systems into trauma care to enable personalized, proactive pain management strategies.