Does State Malpractice Environment Affect Outcomes Following Spinal Fusions? A Machine Learning Analysis. (16th November 2020)

• Record Type: Journal Article
• Authors: Chan, Andrew K.H; Santacatterina, Michele; Pennicooke, Brenton H; Shahrestani, Shane; Ballatori, Alexander; Burke, John F; Manley, Geoffrey T; Tarapore, Phiroz E; Huang, Michael C; Dhall, Sanjay S; Chou, Dean; Mummaneni, Praveen V; DiGiorgio, Anthony M
• Is Part Of: Neurosurgery. Volume 67(2010)Supplement 1
• Publication Date: 2020-11-16
• Subjects: Nervous system -- Surgery -- Periodicals; 617.48005
• DOI: 10.1093/neuros/nyaa447_117
• Languages: English
• ISSNs: 0148-396X

Abstract

Abstract: INTRODUCTION: Spine surgery is especially susceptible to malpractice claims. Critics of the United States (US) medical liability system argue that it drives up costs while proponents argue it deters negligence. METHODS: From 2005–2010, the (1) National Practitioner Data Bank was used to determine the number of malpractice claims per 100 physicians in each state, (2) National Inpatient Sample was queried for spinal fusion patients using ICD-9 procedure codes, and (3) Area Resource File was queried to determine the density of neurosurgeons and orthopaedic surgeons, by state. States were categorized into quartiles with regards to frequency of malpractice claims per 100 physicians. To evaluate the relationship of malpractice claims on outcomes, an inverse-probability-weighted regression-adjustment estimator was used. We controlled for patient and hospital characteristics-including comorbidity using the Charlson Comorbidity Index and complication occurrence-socioeconomic variables, and neurosurgeon and orthopaedic surgeon density. To predict death, discharge, length of stay (LOS) and total costs we used a gradient-boosting classifier and regressor. RESULTS: Overall, 462, 444 discharges were included. Of these, 62, 922 (13.6%), 262, 055 (56.7%), 88, 611 (19.2%), and 48, 856 (10.6%) were from the lowest, 2 nd lowest, 2 nd highest, and highest quartile states for malpractice claims per 100 physicians. The states with more claims showed a statistically significant higher odds of a non-home discharge (OR = 1.07 95%CI[1.05-1.10], 1.18 95%CI[1.15-1.22], 1.12 95%CI[1.06-1.19] in the 2nd, 3rd and highest quartiles respectively; lowest quartile as reference), longer LOS (Mean difference: 0.16 95%CI[0.11-0.21], 0.21 95%CI[0.14-0.28], 0.25 95%CI[0.17-0.33]) and higher total charges (Mean difference (log scale): 0.13 95%CI[0.12-0.14], 0.31 95%CI[0.30-0.32], 0.30 95%CI[0.28-0.31]) without any difference in mortality. For home discharge, the prediction model has a PPV = 0.59, NPV = 0.91, sensitivity = 0.77 and specificity = 0.79. The model also predicts LOS and total charges with an R2 = 0.55 and 0.65, respectively. CONCLUSION: Spinal fusion procedures from states with a higher frequency of malpractice claims were associated with increased odds of non-home discharge, longer LOS, and higher total charges. The frequency of malpractice claims, serving as a proxy for medicolegal climate, may alter practice patterns for a given spine surgeon. This may have important implications for healthcare and tort reform.