A Comparative Analysis Among the SRS M&M, NIS, and KID Databases for the Adolescent Idiopathic Scoliosis.


Journal Article

STUDY DESIGN: Retrospective cohort analysis. OBJECTIVES: A growing number of publications have utilized the Scoliosis Research Society (SRS) Morbidity and Mortality (M&M) database, but none have compared it to other large databases. The objective of this study was to compare SRS complications with those in administrative databases. SUMMARY OF BACKGROUND DATA: The Nationwide Inpatient Sample (NIS) and Kid's Inpatient Database (KID) captured a greater number of overall complications while the SRS M&M data provided a greater incidence of spine-related complications following adolescent idiopathic scoliosis (AIS) surgery. Chi-square was used to obtain statistical significance, with p < .05 considered significant. METHODS: The SRS 2004-2007 (9,904 patients), NIS 2004-2007 (20,441 patients) and KID 2003-2006 (10,184 patients) databases were analyzed for AIS patients who underwent fusion. Comparable variables were queried in all three databases, including patient demographics, surgical variables, and complications. RESULTS: Patients undergoing AIS in the SRS database were slightly older (SRS 14.4 years vs. NIS 13.8 years, p < .0001; KID 13.9 years, p < .0001) and less likely to be male (SRS 18.5% vs. NIS 26.3%, p < .0001; KID 24.8%, p < .0001). Revision surgery (SRS 3.3% vs. NIS 2.4%, p < .0001; KID 0.9%, p < .0001) and osteotomy (SRS 8% vs. NIS 2.3%, p < .0001; KID 2.4%, p < .0001) were more commonly reported in the SRS database. The SRS database reported fewer overall complications (SRS 3.9% vs. NIS 7.3%, p < .0001; KID 6.6%, p < .0001). However, when respiratory complications (SRS 0.5% vs. NIS 3.7%, p < .0001; KID 4.4%, p < .0001) were excluded, medical complication rates were similar across databases. In contrast, SRS reported higher spine-specific complication rates. Mortality rates were similar between SRS versus NIS (p = .280) and SRS versus KID (p = .08) databases. CONCLUSIONS: There are similarities and differences between the three databases. These discrepancies are likely due to the varying data-gathering methods each organization uses to collect their morbidity data. LEVEL OF EVIDENCE: Level IV.

Full Text

Duke Authors

Cited Authors

  • Lee, NJ; Guzman, JZ; Kim, J; Skovrlj, B; Martin, CT; Pugely, AJ; Gao, Y; Caridi, JM; Mendoza-Lattes, S; Cho, SK

Published Date

  • November 2016

Published In

Volume / Issue

  • 4 / 6

Start / End Page

  • 420 - 424

PubMed ID

  • 27927571

Pubmed Central ID

  • 27927571

Electronic International Standard Serial Number (EISSN)

  • 2212-1358

Digital Object Identifier (DOI)

  • 10.1016/j.jspd.2016.05.005


  • eng

Conference Location

  • England