Editorial: Patient validation of retrospective data

Accuracy of operating neurosurgeons' prediction of functional levels after intracranial tumor surgery

OBJECTIVE In the absence of practical and reliable prognostic tools in intracranial tumor surgery, decisions regarding patient selection, patient information, and surgical management are usually based on neurosurgeons' clinical judgment, which may be influenced by personal experience and knowledge. The objective of this study was to assess the accuracy of the operating neurosurgeons' predictions about patients' functional levels after intracranial tumor surgery. METHODS In a prospective single-center study, the authors included 299 patients who underwent intracranial tumor surgery between 2011 and 2015. The operating neurosurgeons scored their patients' expected functional level at 30 days after surgery using the Karnofsky Performance Scale (KPS). The expected KPS score was compared with the observed KPS score at 30 days. RESULTS The operating neurosurgeons underestimated their patients' future functional level in 15% of the cases, accurately estimated their functional levels in 23%, and overestimated their functional levels in 62%. When dichotomizing functional levels at 30 days into dependent or independent functional level categories (i.e., KPS score < 70 or ≥ 70), the predictive accuracy was 80%, and the surgeons underestimated and overestimated in 5% and 15% of the cases, respectively. In a dichotomization based on the patients' ability to perform normal activities (i.e., KPS score < 80 or ≥ 80), the predictive accuracy was 57%, and the surgeons underestimated and overestimated in 3% and 40% of cases, respectively. In a binary regression model, the authors found no predictors of underestimation, whereas postoperative complications were an independent predictor of overestimation (p = 0.01). CONCLUSIONS Operating neurosurgeons often overestimate their patients' postoperative functional level, especially when it comes to the ability to perform normal activities at 30 days. This tendency to overestimate surgical outcomes may have implications for clinical decision making and for the accuracy of patient information.