Predicting discharge placement after elective surgery for lumbar spinal stenosis using machine learning methods

Abstract

Purpose An excessive amount of total hospitalization is caused by delays due to patients waiting to be placed in a rehabilitation facility or skilled nursing facility (RF/SNF). An accurate preoperative prediction of who would need a RF/SNF place after surgery could reduce costs and allow more efficient organizational planning. We aimed to develop a machine learning algorithm that predicts non-home discharge after elective surgery for lumbar spinal stenosis. Methods We used the American College of Surgeons National Surgical Quality Improvement Program to select patient that underwent elective surgery for lumbar spinal stenosis between 2009 and 2016. The primary outcome measure for the algorithm was non-home discharge. Four machine learning algorithms were developed to predict non-home discharge. Performance of the algorithms was measured with discrimination, calibration, and an overall performance score. Results We included 28,600 patients with a median age of 67 (interquartile range 58-74). The non-home discharge rate was 18.2%. Our final model consisted of the following variables: age, sex, body mass index, diabetes, functional status, ASA class, level, fusion, preoperative hematocrit, and preoperative serum creatinine. The neural network was the best model based on discrimination (c-statistic = 0.751), calibration (slope = 0.933; intercept = 0.037), and overall performance (Brier score = 0.131). Conclusions A machine learning algorithm is able to predict discharge placement after surgery for lumbar spinal stenosis with both good discrimination and calibration. Implementing this type of algorithm in clinical practice could avert risks associated with delayed discharge and lower costs.

The Role of Machine Learning in Spine Surgery: The Future Is Now

“…Performance is graded according to its level of discrimination (probability of predicting outcomes accurately) and calibration (degree of over-or underestimating the predicted vs. true outcome) (17). Examples of ML applications encountered by spine surgeons include image classification [i.e., automated detection of vertebral body compression fractures on CT or MRI (18)(19)(20)], preoperative risk stratification models, clinical decision support tools (21)(22)(23)(24)(25), among others. The purpose of this review is to define basic ML terminology, discuss the difference between ML and classical statistics, detail common ML models, and introduce examples in spine research.…”

“…ML enables the development of tools that allow surgeons to plug-in variables and generate probabilities of a nonroutine discharge. Ogink et al recently developed learners to predict discharge to a rehabilitation or skilled nursing facility after surgery for lumbar stenosis using the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database (22). They built multiple models in parallel and ultimately arrived at a neural network that achieved high levels of discrimination and calibration with an Area Under the Curve (AUC) of 0.74 from a Receiver Operating Characteristics curve (22).…”

“…Ogink et al recently developed learners to predict discharge to a rehabilitation or skilled nursing facility after surgery for lumbar stenosis using the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database (22). They built multiple models in parallel and ultimately arrived at a neural network that achieved high levels of discrimination and calibration with an Area Under the Curve (AUC) of 0.74 from a Receiver Operating Characteristics curve (22). This tool has since been externally validated in a smaller cohort, where 97% of patients were accurately predicted to return to home after elective lumbar surgery (21).…”

“…While supervised learners, including CARTs and SVMs, have been used to predict postoperative outcomes, there is evidence to suggest that ANNs may be the preferred method for such tasks going forward (22,23,27,28,68). Kim et al utilized an ANN to predict cardiac and wound complications, venous thromboembolism (VTE), and mortality rates after posterior lumbar fusion from an ACS-NSQIP cohort (68).…”

The Impact of Artificial Intelligence on Quality and Safety

“…For example, as Medicare payments are standardized by procedures performed regardless of hospital LOS, ML systems have been designed with the ability to accurately predict spine surgeryrelated LOS, discharge to nonhome facility, and early unplanned readmissions using only presurgical or predischarge variables. [11][12][13][14] These models can help identify/target certain high-risk patients and the variables that contribute to that risk status, allowing hospitals to allocate specific clinical and social resources to reduce costly LOS and readmissions. This can help to maximize efficiency of care delivered, while also keeping constant or even increasing the quality of care delivered.…”