Improving the Prediction of Total Surgical Procedure Time Using Linear Regression Modeling

Abstract

For efficient utilization of operating rooms (ORs), accurate schedules of assigned block time and sequences of patient cases need to be made. The quality of these planning tools is dependent on the accurate prediction of total procedure time (TPT) per case. In this paper, we attempt to improve the accuracy of TPT predictions by using linear regression models based on estimated surgeon-controlled time (eSCT) and other variables relevant to TPT. We extracted data from a Dutch benchmarking database of all surgeries performed in six academic hospitals in The Netherlands from 2012 till 2016. The final dataset consisted of 79,983 records, describing 199,772 h of total OR time. Potential predictors of TPT that were included in the subsequent analysis were eSCT, patient age, type of operation, American Society of Anesthesiologists (ASA) physical status classification, and type of anesthesia used. First, we computed the predicted TPT based on a previously described fixed ratio model for each record, multiplying eSCT by 1.33. This number is based on the research performed by van Veen-Berkx et al., which showed that 33% of SCT is generally a good approximation of anesthesia-controlled time (ACT). We then systematically tested all possible linear regression models to predict TPT using eSCT in combination with the other available independent variables. In addition, all regression models were again tested without eSCT as a predictor to predict ACT separately (which leads to TPT by adding SCT). TPT was most accurately predicted using a linear regression model based on the independent variables eSCT, type of operation, ASA classification, and type of anesthesia. This model performed significantly better than the fixed ratio model and the method of predicting ACT separately. Making use of these more accurate predictions in planning and sequencing algorithms may enable an increase in utilization of ORs, leading to significant financial and productivity related benefits.

Improving Operating Room Efficiency: Machine Learning Approach to Predict Case-Time Duration

“…Master et al compared multiple machine learning (ML) techniques including decision tree regression, random forest regression, and gradient boosted regression trees as well as hybrid combinations to predict case durations. (10) However, the models were trained on only 10 operations within a single specialty, thus limiting their generalizability (22) The majority of studies aimed at improving surgical case-time estimates have focused on a single subspecialty, which provides limited utility for a clinical administrator managing the entire set of operating room suites. Moreover, many of the models did not restrict the model inputs to only pre-operatively available information, potentially leading to lower accuracy in a prospective implementation.…”

Case study of the prediction of elective surgery durations in a New Zealand teaching hospital

“…For example, Devi et al 3 considered only three types of elective surgeries in the ophthalmology department of a hospital, namely, cataract surgery, corneal transplant surgery and oculoplastic surgery. However, Edelman et al 5 considered all elective and emergency surgeries occurring across six hospitals.…”

Application of Predictive Modelling to Improve the Discharge Process in Hospitals

“…There are various examples of how linear regression has been used in the healthcare setting. A study titled "Improving the prediction of total surgical procedure time using linear regression modeling, " [16] also used linear regression to create a predictive model to optimize service TAT in a clinical setting. In that study the linear regression model gave the most accurate value for the predicted procedure time.…”

“…With a 77% accuracy for the predicted time, the result of that study is similar to the accuracy of the predicted TAT in this study. This can be used as an effective reference to use similar linear regression models for the prediction of TAT of similar administrative and clinical processes and procedures [16].…”