External Validation of the European Spine Study Group-International Spine Study Group Calculator Utilizing a Single Institutional Experience for Adult Spinal Deformity Corrective Surgery.

ABSTRACT

BACKGROUND: The International Spine Study Group (ISSG) and the European Spine Study Group (ESSG) developed an adult spinal deformity (ASD) risk calculator based on one of the most granular, prospective ASD databases. The calculator utilizes preoperative radiographic, surgical, and patient-specific variables to predict patient-reported outcomes and complication rates at 2 years. Our aim was to assess the ISSG-ESSG risk calculator's usability in a single-institution ASD population. METHODS: Frail ([F], 0.3 > 0.5) ASD patients were isolated in a single-center ASD database. Basic demographics were assessed via χ 2 and t tests. Each F patient was inputted into the ESSG risk calculator to identify individual predictive rates for postoperative 2-year health-related quality of life questions (HRQL) outcomes and major complications. These calculated predicted outcomes were analyzed against those identified from the ASD database in order to validate the calculator's predictability via Brier scores. A score closer to 1 meant the ISSG-ESSG calculator was not predictive of that specific outcome. A score closer to 0 meant the ISSG-ESSG calculator was a predictive tool for that factor. RESULTS: A total of 631 ASD patients were isolated (55.8 ± 16.8 years, 26.68 kg/m2, 0.95 ± 1.3 Charlson Comorbidity Index). Of those patients, 7.8% were frail. Fifty percent of frail patients received an interbody fusion, 58.3% received a decompression, and 79.2% underwent osteotomy. Surgical details were as follows mean operative time was 342.9 ± 94.3 minutes, mean estimated blood loss was 2131.82 ± 1011 mL, and average length of stay was 7.12 ± 2.5 days. The ISSG-ESSG calculator predicted the likelihood of improvement for the following HRQL's Oswestry Disability Index (ODI) (86%), Scoliosis Research Society (SRS)-22 mental health (71.1%), SRS-22 total (87.6%), and major complication (53.4%). The single institution had lower percentages of improvement in ODI (24.6%), SRS-22 mental health (21.3%), SRS-22 total (25.1%), and lower presence of major complication (34.8%). The calculated Brier scores identified the calculator's predictability for each factor was as follows ODI (0.24), SRS-22 mental health (0.21), SRS-22 total (0.25), and major complication (0.28). CONCLUSIONS: All of the variables had low Brier scores, indicating that the ISSG-ESSG calculator can be used as a predictive tool for ASD frail patients.