#151 The Impact of Diabetes Mellitus in Surgical Outcomes after Lumbar Spine Surgery
Poster Presented by: K. Singh
A.J. Marquez-Lara (1)
S.V. Nandyala (1)
S.J. Fineberg (1)
M. Noureldin (1)
K. Singh (1)
(1) Rush University Medical Center, Orthopaedic Surgery, Chicago, IL, United States
Introduction: Diabetes Mellitus (DM) is a prevalent comorbidity among patients undergoing spine surgery. The purpose of this study was to utilize a population-based database to determine the impact of DM on the outcomes, costs, and mortality after lumbar spine surgery (LSS).
Methods: The Nationwide Inpatient Sample (NIS) database was queried from 2002-2011. Patients undergoing an elective lumbar decompression (LD) or a lumbar fusion (LF) for degenerative diagnoses were identified. The selected cohorts were further divided into three groups: 1) Non diabetics, 2) Diabetics without end-organ damage (Uncomplicated), and 3) Diabetics with end-organ damage (Complicated). Patient demographics, comorbidities (CCI), length of stay (LOS) and costs were compared between these three groups. SPSS v.20 was utilized for statistical analysis with one-way ANOVA for discrete variables and χ2-tests for categorical data. Multinomial logistic regression identified the independent predictors of postoperative complications in diabetics after controlling for age, other comorbidities, and hospital variables. A p-value <0.001 denoted statistical significance.
Results: A total of 292,833 LD procedures and 263,263 LF procedures were identified in the NIS database. In the LD cohort, 44,357 (15.1%) patients were diabetic of which 3,388 (7.6%) demonstrated complicated diabetes. Similarly, in the LF cohort, 38,067 (14.5%) patients were diabetic, of which 3,022 (7.9%) demonstrated complicated diabetes. In both cohorts, the diabetic patients were significantly older and demonstrated a greater comorbidity burden (p< 0.001). Patients with DM incurred a significantly greater LOS, total hospital costs and postoperative complications than non-diabetics (p< 0.001). In addition, diabetics demonstrated a greater mortality rate than non-diabetics in both the LF (p< 0.001) and LD (p=0.03) cohorts. Regression analysis demonstrated that uncomplicated DM was a predictor of postoperative neurologic complications, urinary tract infections (UTI), ileus, hemorrhagic anemia, and gastrointestinal (GI) bleeds. Similarly, complicated DM was a predictor of pulmonary emboli (PE), infections, UTI, and hemorrhagic anemia.
Conclusion: Diabetic patients who underwent a lumbar spinal procedure incurred a greater LOS, total hospital costs, postoperative complications and mortality than non-diabetics. In addition, diabetes was an independent predictor of postoperative infections, hemorrhagic anemia, PE, and UTI. Further research is warranted to demonstrate if preoperative blood glucose control will help mitigate the risk of postoperative complications and mortality that is associated with this prevalent comorbidity.