Author
Listed:
- Hayat Ouadi
- Ilhame El Farissi
- Ilham Slimani
Abstract
The use of healthcare cost prediction models to assist with resource allocation and risk management is on the rise, but their use is limited by the requirement for reliable, easily interpretable explanations. Although there are multiple methods of providing these explanations, their operational use in healthcare settings remains inadequately evaluated. This research analyses four of the available explainable models (i.e., Permutation Feature Importance, Tree Importance, Local Interpretable Model Explanations, and Shapley-based explanations) using a Random Forest healthcare cost prediction model developed on a real-world dataset (comprising 2,302 aggregated patient segments and 54 features) with a coefficient of determination of 0.9957. The analysis used three criteria that are critical to the deployment of these models: steadiness of explanation under data perturbation, correspondence of feature importance across methods, and computational latency. Results show that each of the global explainable models has relatively high steadiness and relatively low latency, making them applicable for real time and regulatory use; however, the local explainable models provide more intuitive instance-level explanations at the expense of lower steadiness and greater computational demands. A tiered deployment framework for method selection is proposed based on these results as guidance for selecting methods based on clinical, regulatory, and operational needs. This research provides insight into the practical use of explainable healthcare cost prediction systems within real-world environments.
Suggested Citation
Handle:
RePEc:gvu:tcsst0:v:8:y:2026:i:2:id:2196
DOI: 10.36548/jtcsst.2026.2.008
Download full text from publisher
Corrections
All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:gvu:tcsst0:v:8:y:2026:i:2:id:2196. See general information about how to correct material in RePEc.
If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.
We have no bibliographic references for this item. You can help adding them by using this form .
If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: Journal Manager (email available below). General contact details of provider: https://irojournals.com/tcsst .
Please note that corrections may take a couple of weeks to filter through
the various RePEc services.