Rethinking Thoracotomy Analgesia: Paravertebral Block Versus Thoracic Epidural Analgesia in Real-World Practice

Background Open thoracotomy causes severe dynamic pain that compromises ventilation and cough, with downstream risk of pulmonary complications and delayed recovery. Enhanced Recovery After Surgery (ERAS) pathways emphasize multimodal, opioid-sparing analgesia, but the relative real-world performance of thoracic paravertebral block (PVB) versus thoracic epidural analgesia (TEA) in open thoracotomy remains uncertain and requires appraisal of both effectiveness and safety within ERAS priorities. Objective This study aims to compare the analgesic effectiveness and safety of PVB versus TEA after open thoracotomy, with a focus on dynamic pain during the first 48 hours and a composite safety endpoint of hypotension or urinary retention within 72 hours to clarify the clinical benefit-risk profile. Methods This retrospective cohort included consecutive adults undergoing open thoracotomy at a tertiary cardiothoracic center between January 1 and December 31, 2024. Both surgical techniques were placed before the induction phase according to program standards: TEA involves a titratable infusion of 0.125% bupivacaine, while PVB consists of a single-shot bupivacaine injection (not repeated), which typically provides six to eight hours of postoperative analgesia. The primary endpoint was time-weighted mean dynamic pain on the numerical rating scale (0-10) over 48 hours. The prespecified safety composite comprised hypotension (≥20% reduction in mean arterial pressure requiring vasopressor therapy) or urinary retention within 72 hours. Secondary outcomes included rest-pain trajectories, opioid consumption (oral morphine equivalents, 0-48 hours), postoperative nausea and vomiting, pulmonary complications, intensive care unit admission within 48 hours, hospital length of stay, and 30-day readmission. Analyses employed propensity-score overlap weighting with robust models and sensitivity checks. Results Of 458 screened patients, 380 were analyzed (TEA: 196 (51.6%); PVB: 184 (48.4%)). Pain-score completeness was 97.4% (TEA) and 97.8% (PVB). Dynamic pain declined in both groups (post-anesthesia care unit: TEA 6.8 ± 2.0 vs. PVB 6.2 ± 2.1; 24 h: 5.8 ± 1.9 vs. 5.1 ± 1.8; 48 h: 4.8 ± 1.7 vs. 4.2 ± 1.6). Adjusted mean differences favored PVB at 24 h (-0.6; 95% CI, -0.9 to -0.3) and 48 h (-0.5; 95% CI, -0.8 to -0.2), remaining below the 1-point minimal clinically important difference (consistent with non-inferior dynamic analgesia). The safety composite occurred in 54/196 (27.6%) with TEA versus 26/184 (14.1%) with PVB (adjusted odds ratio 0.44; 95% CI, 0.26-0.72; absolute risk difference -13.5 percentage points; 95% CI, -20.7 to -6.3), with concordant reductions in component outcomes. Opioid use was similar (36 vs. 38 mg; adjusted difference +2 mg; 95% CI, -2 to +6). Other postoperative outcomes were comparable. Conclusions In a contemporary ERAS thoracotomy cohort, PVB provided non-inferior dynamic analgesia with superior safety versus TEA, meaningfully reducing hypotension and urinary retention. These findings support PVB as a pragmatic default when hemodynamic stability is prioritized and align with ERAS-aligned benefit-risk considerations.