Study design: A retrospective cohort study.
Objective: The aim of this study was to determine the difference in 30-day readmission, reoperation, and morbidity for patients undergoing either posterior or anterior lumbar interbody fusion.
Summary of background data: Despite increasing utilization of lumbar interbody fusion to treat spinal pathology, few studies compare outcomes by surgical approach, particularly using large nationally represented cohorts.
Methods: Patients who underwent lumbar interbody fusion were identified using the NSQIP database. Rates of readmission, reoperation, morbidity, and associated predictors were compared between posterior/transforaminal (PLIF/TLIF) and anterior/lateral (ALIF/LLIF) lumbar interbody fusion using multivariate regression. Bonferroni-adjusted alpha-levels were utilized whereby variables were significant if their P values were less than the alpha-level or trending if their P values were between 0.05 and the alpha-level.
Results: We identified 26,336 patients. PLIF/TLIF had greater operative time (P = 0.015), transfusion (P < 0.001), UTI (P = 0.008), and stroke/CVA (P = 0.026), but lower prolonged ventilation (P < 0.001) and DVT (P = 0.002) rates than ALIF/LLIF. PLIF/TLIF independently predicted greater morbidity on multivariate analysis (odds ratio: 1.155, P = 0.0019).In both groups, experiencing a complication and, in PLIF/TLIF, ASA-class ≥3 predicted readmission (P < 0.001). Increased age trended toward readmission in ALIF/LLIF (P = 0.003); increased white cell count (P = 0.003), dyspnea (P = 0.030), and COPD (P = 0.005) trended in PLIF/TLIF. In both groups, increased hospital stay and wound/site-related complication predicted reoperation (P < 0.001). Adjunctive posterolateral fusion predicted reduced reoperation in ALIF/LLIF (P = 0.0018). ASA-class ≥3 (P = 0.016) and age (P = 0.021) trended toward reoperation in PLIF/TLIF and ALIF/LLIF, respectively. In both groups, age, hospital stay, reduced hematocrit, dyspnea, ASA-class ≥3, posterolateral fusion, and revision surgery and, in PLIF/TLIF, bleeding disorder predicted morbidity (P < 0.001). Female sex (P = 0.010), diabetes (P = 0.042), COPD (P = 0.011), and disseminated cancer (P = 0.032) trended toward morbidity in PLIF/TLIF; obesity trended in PLIF/TLIF (P = 0.0022) and ALIF/LLIF (P = 0.020).
Conclusion: PLIF/TLIF was associated with a 15.5% increased odds of morbidity; readmission and reoperation were similar between approaches. Older age, higher ASA-class, and specific comorbidities predicted poorer 30-day outcomes, while procedural-related factors predicted only morbidity. These findings can guide surgical approach given specific factors.
Level of evidence: 3.