Study objective: The objective of this study was to evaluate the accuracy of a new elective surgery clinical decision support system, the 'Patient Tacking List' (PTL) tool (C2-Ai(c)) through receiver operating characteristic (ROC) analysis.
Methods: We constructed ROC curves based on risk predictions produced by the tool and compared these with actual patient outcomes on a retrospective cohort of patients awaiting elective surgery.
Results: A total of 11 837 patients were included across three National Health Service (NHS) hospitals in England. ROC analysis revealed an area under the curve of 0.95 (95% CI 0.92 to 0.98) for mortality and 0.8 (95% CI 0.78 to 0.82) for complications.
Discussion: The PTL tool was successfully integrated into existing data infrastructures, allowing real-time clinical decision support and a low barrier to implementation. ROC analysis demonstrated a high level of accuracy to predict the risk of mortality and complications after elective surgery. As such, it may be a valuable adjunct in prioritising patients on surgical waiting lists.Health systems, such as the NHS in England, must look at innovative methods to prioritise patients awaiting surgery in order to best use limited resources. Clinical decision support tools, such as the PTL tool, can improve prioritisation and thus positively impact clinical care and patient outcomes.
Conclusions: The high level of accuracy for predicating mortality and complications after elective surgery using the PTL tool indicates the potential for clinical decision support tools to help tackle rising waiting lists and improve surgical planning.
Keywords: artificial intelligence; data interpretation, statistical; general surgery; medical Informatics.
© Author(s) (or their employer(s)) 2023. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.