Predicting 30-Day Perioperative Outcomes in Adult Spinal Deformity Patients With Baseline Paralysis or Functional Dependence.

Abstract

Patients undergoing surgical treatment of adult spinal deformity (ASD) are often preoperatively risk stratified using standardized instruments to assess for perioperative complications. Many ASD instruments account for medical comorbidity and radiographic parameters, but few consider a patient's ability to independently accomplish necessary activities of daily living (ADLs). Patients ≥18 years undergoing ASD corrective surgery were identified in National Surgical Quality Improvement Program. Patients were grouped by (1) plegic status and (2) dependence in completing ADLs ("totally dependent" = requires total assistance in ADLs, "partially dependent" = uses prosthetics/devices but still requires help, "independent" = requires no help). Quadriplegics and totally dependent patients comprised "severe functional dependence," paraplegics/hemiplegics who are "partially dependent" comprised "moderate functional dependence," and "independent" nonplegics comprised "independent." Analysis of variance with post hoc testing and Kruskal-Wallis tests compared demographics and perioperative outcomes across groups. Logistic regression found predictors of inferior outcomes, controlling for age, sex, body mass index (BMI), and invasiveness. Subanalysis correlated functional dependence with other established metrics such as the modified Frailty Index (mFI) and Charlson Comorbidity Index (CCI). A total of 40,990 ASD patients (mean age 57.1 years, 53% women, mean BMI 29.8 kg/m2) were included. Mean invasiveness score was 6.9 ± 4.0; 95.2% were independent (Indep), 4.3% moderate (Mod), and 0.5% severe (Sev). Sev had higher baseline invasiveness than Mod or Indep groups (9.0, 8.3, and 6.8, respectively, P < 0.001). Compared with the Indep patients, Sev and Mod had significantly longer inpatient length of stay (LOS; 10.9, 8.4, 3.8 days, P < 0.001), higher rates of surgical site infection (2.2%, 2.9%, 1.5%, P < 0.001), and more never events (17.7%, 9.9%, 4.0%, P < 0.001). Mod had higher readmission rates than either the Sev or Indep groups (30.2%, 2.7%, 10.3%, P < 0.001). No differences in implant failure were observed (P > 0.05). Controlling for age, sex, BMI, CCI, invasiveness, and frailty, regression equations showed increasing functional dependence significantly increased odds of never events (OR, 1.82 [95% CI 1.57-2.10], P < 0.001), specifically urinary tract infection (OR, 2.03 [95% CI 1.66-2.50], P < 0.001) and deep venous thrombosis (OR, 2.04 [95% CI 1.61-2.57], P < 0.001). Increasing functional dependence also predicted longer LOS (OR, 3.16 [95% CI 2.85-3.46], P < 0.001) and readmission (OR, 2.73 [95% CI 2.47-3.02], P < 0.001). Subanalysis showed functional dependence correlated more strongly with mFI (r = 0.270, P < 0.001) than modified CCI (mCCI; r = 0.108, P < 0.001), while mFI and mCCI correlated most with one another (r = 0.346, P < 0.001). Severe functional dependence had significantly longer LOS and more never-event complications than moderate or independent groups. Overall, functional dependence may show superiority to traditional metrics in predicting poor perioperative outcomes, such as increased LOS, readmission rate, and risk of surgical site infection and never events.