These cases were analyzed retrospectively. The goal of this project was to use machine learning models to predict how much money will be needed when an anterior cervical discectomy and fusion (ACDF) is performed within 90 days. There has been a disproportional decline in reimbursement despite an increased incidence of ACDF. As bundled payment models gain popularity, it is crucial to isolate the variables contributing to the final bill.
All main ACDFs performed in 2018 were looked up using the International Classification of Diseases 10th Revision (ICD-10) procedure codes in the National Readmissions Database (NRD). The overall hospital charge was used along with the cost-per-charge ratio for each facility, were used to determine final costs. Researchers also asked questions about the hospital itself, such as the number of surgeries they conduct and their pay scale. Costs associated with readmissions within 90 days of admission were discovered and added to the overall admission cost to give a total healthcare cost for the first 90 days. Patients with 90-day admission costs that were more than 1 standard deviation (SD) above the mean were identified with the help of machine learning algorithms.
The study covered 42,485 procedures on a population with a mean age of 57.7±12.3 years and a gender split of 50.6% men. The average surgical admission cost is $24,874±25,610, the average readmission cost is $25,371±11,476, and the average total cost $26,977±28,947. A total of 10,624 patients met the criteria for “high cost.” The overall cost of treatment was most strongly related to the wage index, the amount of hospitalizations, the patient’s age, and the severity of the patient’s underlying medical diagnosis. The area under the curve for predictions made by the gradient boosting trees technique was 0.86, making it the most accurate model for estimating the overall cost of care.
ACDF reimbursement is based on the salary index and diagnosis-related groups in bundled payment models. Other factors, such as hospital volume, readmission, and patient age, were found to be significant in influencing healthcare costs using machine learning algorithms. Machine learning can increase efficiency and decrease financial strain on doctors and hospitals by applying patient-specific reimbursement.