OBJECTIVE: The aim of this study was to develop a perioperative metastatic spinal tumor frailty index (MSTFI) that could predict morbidity, mortality, and length of stay. METHODS: A large inpatient hospitalization database was searched from 2002 to 2011 to identify 4583 patients with spinal metastasis from breast (21.1%), lung (34.1%), thyroid (3.8%), renal (19.9%), and prostate (21.1%) cancer who underwent surgery. A multiple logistic regression model identified 9 independent parameters that were used to construct the MSTFI: anemia, chronic lung disease, coagulopathy, electrolyte abnormalities, pulmonary circulation disorders, renal failure, malnutrition, emergent/urgent admission, and anterior/combined surgical approach. Patients with 0 points were categorized as "not frail,"1 as "mildly frail," 2 as "moderately frail," and >= 3 as "severely frail." RESULTS: The overall perioperative complication rate was 19.3% and in-patient mortality was 3.0%. Compared with patients with no frailty, patients with moderate frailty (odds ratio [OR] 5.15; 95% confidence interval [95% CI] 2.44-10.86), and severe frailty (OR 5.74; 95% CI 2.69-12.24) had significantly increased odds of inpatient mortality (all P < 0.001). Similarly, patients with mild frailty (OR 1.88; 95% CI 1.33-2.66), moderate frailty (OR 3.83; 95% CI 2.71-5.41), and severe frailty (OR 6.97; 95% CI 4.98-9.74) had significantly increased odds of developing a major in-hospital complication (all P < 0.001). Length of stay also increased significantly by MSTFI (P < 0.001). CONCLUSIONS: In surgically treated patients with spinal metastasis, certain perioperative parameters may significantly predict the risk of major in-hospital complications and mortality.
