Statistical modeling can determine what factors are predictive of appropriate follow-up in patients presenting with incidental pulmonary nodules on CT.

摘要: Abstract Purpose To assess the performance of statistical modeling in predicting follow-up adherence incidentally detected pulmonary nodules (IPN) on CT, based patient variables (PV), radiology report related (RRRV) and physician-patient communication (PPCV). Methods 200 patients with IPN CT were retrospectively identified randomly selected. PV (age, gender, smoking status, ethnicity), RRRV (nodule size, context, whether recommendations provided) PPCV (whether referring physician documented ordered electronic medical record) recorded. Primary outcome was received appropriate within +/- 1 month recommended time frame. Statistical methods included logistic regression machine learning (K-nearest neighbors support vector machine). Results Adherence low, or without provided (23.4 %–27.4 %). Whether dominant predictor all models. The following statistically significant predictors follow-up: report, context nodule size (FDR logworth respectively 21.18, 11.66, 2.35, 1.63, p  Conclusion are most important adherence. Amongst variables, radiologist adherence, supporting utility for analytics, quality assurance optimization outcomes to IPN.