Abstract
Purpose
Many countries have developed clinical decision-making support tools, such as the smart work injury management (SWIM) system in Hong Kong, to predict rehabilitation paths and address global issues related to work injury disability. This study aims to evaluate the accuracy of SWIM by comparing its predictions on real work injury cases to those made by human case managers, specifically with regard to the duration of sick leave and the percentage of permanent disability.
Methods
The study analyzed a total of 442 work injury cases covering the period from 2012 to 2020, dividing them into non-litigated and litigated cases. The Kruskal–Wallis post hoc test with Bonferroni adjustment was used to evaluate the differences between the actual data, the SWIM predictions, and the estimations made by three case managers. The intra-class correlation coefficient was used to assess the inter-rater reliability of the case managers.
Results
The study discovered that the predictions made by the SWIM model and a case manager possessing approximately 4 years of experience in case management exhibited moderate reliability in non-litigated cases. Nevertheless, there was no resemblance between SWIM’s predictions regarding the percentage of permanent disability and those made by case managers.
Conclusion
The findings indicate that SWIM is capable of replicating the sick leave estimations made by a case manager with an estimated 4 years of case management experience, albeit with limitations in generalizability owing to the small sample size of case managers involved in the study.
Implications
These findings represent a significant advancement in enhancing the accuracy of CDMS for work injury cases in Hong Kong, signaling progress in the field.
Many countries have developed clinical decision-making support tools, such as the smart work injury management (SWIM) system in Hong Kong, to predict rehabilitation paths and address global issues related to work injury disability. This study aims to evaluate the accuracy of SWIM by comparing its predictions on real work injury cases to those made by human case managers, specifically with regard to the duration of sick leave and the percentage of permanent disability.
Methods
The study analyzed a total of 442 work injury cases covering the period from 2012 to 2020, dividing them into non-litigated and litigated cases. The Kruskal–Wallis post hoc test with Bonferroni adjustment was used to evaluate the differences between the actual data, the SWIM predictions, and the estimations made by three case managers. The intra-class correlation coefficient was used to assess the inter-rater reliability of the case managers.
Results
The study discovered that the predictions made by the SWIM model and a case manager possessing approximately 4 years of experience in case management exhibited moderate reliability in non-litigated cases. Nevertheless, there was no resemblance between SWIM’s predictions regarding the percentage of permanent disability and those made by case managers.
Conclusion
The findings indicate that SWIM is capable of replicating the sick leave estimations made by a case manager with an estimated 4 years of case management experience, albeit with limitations in generalizability owing to the small sample size of case managers involved in the study.
Implications
These findings represent a significant advancement in enhancing the accuracy of CDMS for work injury cases in Hong Kong, signaling progress in the field.
| Original language | English |
|---|---|
| Pages (from-to) | 1-13 |
| Journal | Journal of Occupational Rehabilitation |
| DOIs | |
| Publication status | Published - 14 Jun 2024 |
Keywords
- Clinical decision-making support
- Prediction
- Artificial intelligence
- Worker’s compensation
- Work disability management
