Factors affecting length of stay after elective posterior lumbar spine surgery: a multivariate analysis

Abstract

Background contextElective posterior lumbar fusion is a common surgical procedure, but reported length of hospital stay is variable (usually 3–7 days). The effect of a limited number of factors on length of stay (LOS) has previously been evaluated. However, multivariate analysis using LOS as a dependent variable to separate potentially confounding variables has not been performed. PurposeTo facilitate setting of realistic expectations and considering the significant costs of hospitalization, it would be ideal to have a clear understanding of the variables affecting LOS for this surgery. Study design/settingThis is a retrospective case series at a tertiary care center. Patient sampleOne hundred three patients undergoing elective, open, one- to three-level posterior lumbar instrumented fusion (with or without decompression) by the orthopedic spine service at our institution between January 2010 and June 2012 were included in the study. Outcome measuresLOS was determined from the date of surgery to the date of discharge. MethodsPreoperative factors (patient demographics, previous surgery, levels instrumented, American Society of Anesthesiologists [ASA] score, and major medical comorbidities including diabetes, hypertension, malignancy, pulmonary disease, or heart disease), intraoperative factors (complications, drain placement, estimated blood loss, blood transfusion, fluids administered, operating room time, and surgery time), and postoperative factors (drain removal, blood transfusion, complications, and discharge destination) were collected and analyzed with multivariate stepwise regression to determine predictors of LOS. “Postoperative complications” were excluded as an independent variable from the regression analysis because of its close relationship with LOS. No funding was received for the completion of this study, and there are no potential conflicts of interests. ResultsOur sample included 70 one-level, 26 two-level, and 7 three-level operations. Average LOS was 3.6±1.8 days (mean±SD) with the range 0 to 12 days. Of this cohort, 79% (81 of 103) had a stay of 4 days or less. The only preoperative variables associated with LOS in the multivariate model were age (p=.038) and ASA score (p=.001). History of heart disease (p=.005) was significantly associated with a decreased hospital stay. Intraoperative complications included six dural tears and one pedicle fracture. No intraoperative factors were found to be associated with a longer LOS. Postoperative complications occurred in 32% of patients (33 of 103). Common complications included anemia requiring transfusion (11), altered mental status (8), pneumonia (4), hardware complications requiring reoperation (3). Only one serious complication, renal failure, occurred. Average LOS for patients with a postoperative complication was 5.1±2.3 vs. 2.9±0.9 days for patients with no complication (p<.001). Discharge to a subacute or nursing facility (p<.001) was significantly associated with increased LOS. Levels fused were not predictive of LOS, possibly due to the skew toward one-level cases in our sample. ConclusionPatients who are older and have widespread systemic disease tend to stay in the hospital longer after surgery. Contrary to our expectations, no single comorbidity was predictive of longer hospital stays. Heart disease was associated with a shorter LOS, but this may have been due to a more extensive preoperative workup and closer medical management. Intraoperative events did not affect LOS; however, postoperative events did. These data should prove useful for counseling patients and setting expectations of patients and the health care team.