Association of preoperative hemoglobin and postoperative outcome in non-cardiac surgery patients with COPD: a retrospective cohort study

Chronic obstructive pulmonary disease (COPD) patients face elevated perioperative risks, yet evidence-based guidelines for preoperative hemoglobin optimization in non-cardiac surgery remain undefined. This study investigated the association between preoperative hemoglobin levels and postoperative outcomes in the high-risk population, with emphasis on nonlinear dynamics and clinically actionable thresholds. In this retrospective cohort study, 346 COPD patients undergoing non-cardiac surgery were analyzed. Multivariable regression models (adjusted for demographics, comorbidities, and surgical factors) evaluated the association of hemoglobin and length of stay (LOS), mortality, and other complications. Restricted cubic splines and threshold analysis characterized nonlinear relationships, while subgroup analysis and sensitivity analysis assessed effect heterogeneity. For every 1 g/L increase in Hemoglobin, LOS was shortened by 0.09 days (95% CI −0.15 ~ −0.04; P = 0.001) and mortality was decreased by 9% (OR = 0.92; 95% CI 0.86 ~ 0.98; P = 0.013). Moderate/Severe anemia (Hb < 90 g/L) increased mortality 23-fold (OR = 23.42; 95% CI 1.8 ~ 304.23; P = 0.013) and postoperative delirium (POD) risk ninefold (OR = 9.27; 95% CI 1.77 ~ 48.53; P = 0.02). There was a non-linear relationship between Hemoglobin and LOS/POD. Subgroup analysis and forest plots indicated heterogeneity in the inverse hemoglobin-LOS association. Sensitivity analysis showed the effect of anemia on LOS was significant in the elderly (OR = 5.19, 95% CI −1.1 ~ 11.49; P = 0.007) and pulmonary hypertension subgroup (OR = 6.88, 95% CI 2.33 ~ 11.43; P < 0.001). Preoperative hemoglobin optimization may be particularly beneficial for COPD patients undergoing non-cardiac surgery, with effects potentially more pronounced in elderly patients or those with pulmonary hypertension. However, our study was underpowered to confirm subgroup-specific effects due to sample size limitations. Future research should prioritize these high-risk subgroups for targeted preoperative optimization trials.