Improving Medical Efficiency with Machine Learning

Medical coding is a crucial process that converts clinical documentation into standardized medical codes for various purposes, such as insurance reimbursement and performance analysis. With the increasing demand for accurate and efficient medical coding, several solutions based on artificial intelligence have been proposed to assist in the process. However, their effectiveness is limited, and there is a need for more innovative approaches.

To address this issue, a study was conducted to develop a multimodal machine learning-based solution that detects the degree of coding complexity before coding is performed. The notion of coding complexity was used to better distribute work among medical coders to minimize errors and improve throughput.

To train and evaluate the approach, the researchers collected 2060 cases rated by coders in terms of coding complexity from 1 (simplest) to 4 (most complex). Two expert coders rated 3.01% (62/2060) of the cases as the gold standard. The agreements between experts were used as benchmarks for model evaluation. A case contains both clinical text and patient metadata from the hospital electronic health record. Text and metadata features were extracted, concatenated, and fed into several machine learning models. Two models were selected for evaluation.

The first model achieved a macro-F1-score of 0.51 and an accuracy of 0.59 on classifying the 4-scale complexity. The model distinguished well between the simple (combined complexity 1-2) and complex (combined complexity 3-4) cases with a macro-F1-score of 0.65 and an accuracy of 0.71. The second model achieved 61% agreement with experts’ ratings and a macro-F1-score of 0.62 on the gold standard, whereas the 2 experts had a 66% agreement ratio with a macro-F1-score of 0.67.

In conclusion, the proposed multimodal machine learning approach leverages information from both clinical text and patient metadata to predict the complexity of coding a case in the precoding phase. By integrating this model into the hospital coding system, distribution of cases among coders can be done automatically with performance comparable with that of human expert coders, thus improving coding efficiency and accuracy at scale. This approach has the potential to significantly improve the efficiency of medical coding processes, reducing errors, and ultimately improving patient care.

References:

Xu H, Maccari B, Guillain H, Herzen J, Agri F, Raisaro JAn End-to-End Natural Language Processing Application for Prediction of Medical Case Coding Complexity: Algorithm Development and ValidationJMIR Med Inform 2023;11:e38150URL: https://medinform.jmir.org/2023/1/e38150DOI: 10.2196/38150

‍