Clinicosocial determinants of hospital stay following cervical decompression: A public healthcare perspective and machine learning model

• Published in: Journal of clinical neuroscience 2024-08, Vol.126, p.1-11
• Main Authors: Biswas, Sayan, Aizan, Luqman Naim Bin, Mathieson, Katie, Neupane, Prashant, Snowdon, Ella, MacArthur, Joshua, Sarkar, Ved, Tetlow, Callum, Joshi George, K.
• Format: Article
• Language: English
• Subjects: ACDF; Index of multiple deprivation; Machine learning
• ISSN: 0967-5868; 1532-2653; 1532-2653
• DOI: 10.1016/j.jocn.2024.05.032

•Impact of socio-economic deprivation on post-ACDF length of stay is unknown.•A total of 2033 patients who had undergone ACDF at our institution were analysed.•Index of multiple deprivation decile significantly predicted post-op length of stay.•The XGBoost model had 80.95 % accuracy, 71.52 % sensitivity and 85.76 % specificity.•Non-clinical pre-operative comorbidities and patient factors can predict length of stay. Post-operative length of hospital stay (LOS) is a valuable measure for monitoring quality of care provision, patient recovery, and guiding hospital resource management. But the impact of patient ethnicity, socio-economic deprivation as measured by the indices of multiple deprivation (IMD), and pre-existing health conditions on LOS post-anterior cervical decompression and fusion (ACDF) is under-researched in public healthcare settings. From 2013 to 2023, a retrospective study at a single center reviewed all ACDF procedures. We analyzed 14 non-clinical predictors—including demographics, comorbidities, and socio-economic status—to forecast a categorized LOS: short (≤2 days), medium (2–3 days), or long (>3 days). Three machine learning (ML) models were developed and assessed for their prediction reliability. 2033 ACDF patients were analyzed; 79.44 % had a LOS ≤ 2 days. Significant predictors of LOS included patient sex (HR:0.81[0.74–0.88], p