Differences in the association between epidural analgesia and length of stay by surgery type: an observational study.

Introduction: Despite a decline in the use of thoracic epidural analgesia related in part to concerns for delayed discharge, it is unknown whether changes in length of stay (LOS) associated with epidural analgesia vary by surgery type. Therefore, we determined the degree to which the association between epidural analgesia (vs no epidural) and LOS differed by surgery type.

Methods: We conducted an observational study using data from 1747 patients who had either non-emergent open abdominal, thoracic, or vascular surgery at a single tertiary academic hospital. The primary outcome was hospital LOS and the incidence of a prolonged hospital LOS defined as 21 days or longer. Secondary endpoints included escalation of care, 30-day all-cause readmission, and reason for epidural not being placed. The association between epidural status and dichotomous endpoints was examined using logistic regression.

Results: Among the 1747 patients, 85.7% (1499) received epidural analgesia. 78% (1364) underwent abdominal, 11.5% (200) thoracic, and 10.5% (183) vascular surgeries. After adjustment for differences, receiving epidural analgesia (vs no epidural) was associated with a 45% reduction in the likelihood of a prolonged LOS (p<0.05). This relationship varied by surgery type: abdominal (OR 0.42, 95% CI 0.23 to 0.79, p<0.001), vascular (OR 1.66, 95% CI 0.17 to 16.1, p=0.14), and thoracic (OR 1.07, 95% CI 0.20 to 5.70, p=0.93). Among abdominal surgical patients, epidural analgesia was associated with a median decrease in LOS by 1.4 days and a 37% reduction in the likelihood of 30-day readmission (adjusted OR 0.63, 0.41 to 0.97, p<0.05). Among thoracic surgical patients, epidural analgesia was associated with a median increase in LOS by 3.2 days.

Conclusions: The relationship between epidural analgesia and LOS appears to be different among different surgical populations.