TY - JOUR
T1 - Comparative Analysis of Peri-operative Outcomes Using Nationally Derived Hospital Discharge Data Relative to a Prospective Multi-center Surgical Database of Adult Spinal Deformity Surgery
AU - International Spine Study Group (ISSG)
AU - Poorman, Gregory W.
AU - Passias, Peter G.
AU - Buckland, Aaron J.
AU - Jalai, Cyrus M.
AU - Kelly, Michael
AU - Sciubba, Daniel M.
AU - Neuman, Brian J.
AU - Hamilton, Kojo K.
AU - Jain, Amit
AU - Diebo, Bassel
AU - Lafage, Virginie
AU - Bess, Shay
AU - Klineberg, Eric Otto
PY - 2016/12/2
Y1 - 2016/12/2
N2 - STUDY DESIGN.: Retrospective analysis of three prospectively collected databases. OBJECTIVE.: compare peri-operative outcomes in Adult Spinal Deformity (ASD) surgeries in a surgeon-run (SR-ASD) and two national databases: the Nationwide Inpatient Sample (NIS) and the National Surgical Quality Improvement Program (NSQIP). SUMMARY OF BACKGROUND DATA.: has been learned on the treatment of ASD in the last decade with prospective multicenter collaborative research focusing on this specific condition. Non-disease specific national databases are being used for hypothesis and quality control testing on a large numbers of ASD patients. Their accuracy and applicability remains unevaluated. METHODS.: Patients were identified on each respective database undergoing lumbar spine fusion for ASD. Propensity score matching established cohorts of patients on each database with similar procedures being performed. Complication prevalence and relative risk was compared on the NIS and NSQIP against SR-ASD. Secondary outcome measures included hospital-stay characteristics, surgical invasiveness, patient demographics, and patient comorbidities. RESULTS.: 255 patients were identified on each database 1:1:1 with similar overall surgical intensity. Querying the databases using ICD-9 codes, CPT codes, and surgeon-reports resulted in different complication incidences: overall complication rates were 17.65% on NIS, 24.31% on NSQIP, and 68.24% on SR-ASD. The relative risk of a medical complication in SR-ASD was 1.87 (1.42–2.48) relative to NIS and 1.91 (1.44–2.54) relative to NSQIP. The relative risk of a surgical complication was 5.45 (2.69–11.05) compared to NIS and 12.05 (3.98–36.49) compared to NSQIP. CONCLUSIONS.: After selecting patients using the same criteria and diagnosis, NIS, NSQIP, and SR-ASD databases captured different patient populations and different complication incidences. There were total absences of certain complications contrary to usual literature rates in all three databases. Faithful reporting necessitates understanding database limitations, and careful evaluation of database strengths and weaknesses is paramount to accurate reports.Level of Evidence: 3
AB - STUDY DESIGN.: Retrospective analysis of three prospectively collected databases. OBJECTIVE.: compare peri-operative outcomes in Adult Spinal Deformity (ASD) surgeries in a surgeon-run (SR-ASD) and two national databases: the Nationwide Inpatient Sample (NIS) and the National Surgical Quality Improvement Program (NSQIP). SUMMARY OF BACKGROUND DATA.: has been learned on the treatment of ASD in the last decade with prospective multicenter collaborative research focusing on this specific condition. Non-disease specific national databases are being used for hypothesis and quality control testing on a large numbers of ASD patients. Their accuracy and applicability remains unevaluated. METHODS.: Patients were identified on each respective database undergoing lumbar spine fusion for ASD. Propensity score matching established cohorts of patients on each database with similar procedures being performed. Complication prevalence and relative risk was compared on the NIS and NSQIP against SR-ASD. Secondary outcome measures included hospital-stay characteristics, surgical invasiveness, patient demographics, and patient comorbidities. RESULTS.: 255 patients were identified on each database 1:1:1 with similar overall surgical intensity. Querying the databases using ICD-9 codes, CPT codes, and surgeon-reports resulted in different complication incidences: overall complication rates were 17.65% on NIS, 24.31% on NSQIP, and 68.24% on SR-ASD. The relative risk of a medical complication in SR-ASD was 1.87 (1.42–2.48) relative to NIS and 1.91 (1.44–2.54) relative to NSQIP. The relative risk of a surgical complication was 5.45 (2.69–11.05) compared to NIS and 12.05 (3.98–36.49) compared to NSQIP. CONCLUSIONS.: After selecting patients using the same criteria and diagnosis, NIS, NSQIP, and SR-ASD databases captured different patient populations and different complication incidences. There were total absences of certain complications contrary to usual literature rates in all three databases. Faithful reporting necessitates understanding database limitations, and careful evaluation of database strengths and weaknesses is paramount to accurate reports.Level of Evidence: 3
UR - http://www.scopus.com/inward/record.url?scp=85002131851&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85002131851&partnerID=8YFLogxK
U2 - 10.1097/BRS.0000000000002002
DO - 10.1097/BRS.0000000000002002
M3 - Article
C2 - 28742736
AN - SCOPUS:85002131851
JO - Spine
JF - Spine
SN - 0362-2436
ER -