Research Fellow Rothman Orthopaedic Institute Rothman Orthopaedic Institute
Disclosure(s):
Jarod Olson, BS: No financial relationships to disclose
Introduction: Between 2013 and 2015, the number or publications incorporating large databases increased by 175%. While these databases increase statistical power, some studies have highlighted variability in data coding and a lack of granularity as weaknesses of this methodology. The aim of this study is to identify trends in large database utilization across prominent spine surgery journals to assess their impact and growth over time.
Methods: Articles published from 2012-2022 across six spine journals were analyzed: Global Spine Journal(GSJ), Spine, The Spine Journal(TSJ), European Spine Journal(ESJ), International Journal of Spine Surgery(IJSS), and Journal of Neurosurgery-Spine(JNS). Three independent researchers manually reviewed each volume and issue of the above journals over the ten-year period for inclusion of the following databases as a source: National Inpatient Sample(NIS), Centers for Medicare & Medicaid Service Data(CMS), National Surgical Quality Improvement Program(NSQIP), National Trauma Data Bank(NTDB), PearlDiver(PD), and IBM Market Scan(MS). Studies were excluded if not reporting on patient data. Bivariate analyses were performed using independent t-tests with significance set to p< 0.05.
Results: A total of 409 spine studies using large databases were included. The number of spine surgery studies increased over the ten-year period with a max of 157 studies in 2017-2018(p < 0.001). From 2012-2022, the journal-specific distribution showed a decrease in Spine(68.7%-29.5%), increase in GSJ(0%-29.5%), and consistent trend in TSJ(25.4%-26.3%, p< 0.001). Database use varied but PearlDiver was the most used in every time period(p < 0.001). There was an increase in first authors with BS/BA degrees(22.4%-40.0%, p< 0.001) and number of affiliated institutions(2.30 to 3.44, p< 0.001) from 2012 to 2022. Study topics saw a decrease in surgical complications(40.3%-25.3%) and increase in pre-operative risk factors(7.46%-16.8%, p< 0.001).
Conclusion : This study demonstrates that the use of large databases has expanded significantly in spine literature. The utilization of these databases has increased the capacity for inter-institutional collaboration and inclusion of medical students in authorship. Additionally, this practice has increased the statistical power of these analyses. However, further investigation is needed to evaluate and address concerns over data heterogeneity, potentially low clinical significance of statistically significant results, and the ability to draw conclusions from agranular large datasets.