Insight Diagnosis Related Grouper: Our Biggest Reconstruction Yet

The Insight DRG was introduced in 2013 and quickly went on to be the most widely used DRG in the South African market. A key part of ensuring the DRG remains relevant is ensuring clinical coding is up to date and that groupings appropriately reflect clinical practice. We are pleased to say the sixth iteration of the grouper was launched in November 2025 and is the most significant update to the algorithm to date.

A DRG is a mechanism to assign hospital admissions to clinically intuitive and statistically homogeneous [1] categories. The Insight DRG has been used to facilitate risk adjustment in the context of cost and quality benchmarking exercises. These exercises are central to network selection processes and cost and quality improvement initiatives. The Insight DRG has been used to empower hospitals and funders to develop more meaningful reporting frameworks and to underpin alternative reimbursement models encompassing billions of rands.

The latest iteration of the DRG marks a significant departure from previous versions. Significant effort was put int making groupings are more clinically intuitive and more statistically homogeneous. This allows more meaningful reporting, a deeper understanding of variations in the cost and quality of care and the development of more robust alternative reimbursement models.

More than 100 changes were made to the DRG logic. Additional groups were created to ensure that disparate admission types are no longer grouped together. Some existing groups were merged to ensure that the total number of categories remains manageable. Where necessary, diagnosis and procedure codes were reassigned from one DRG to another to ensure clinical consistency and interpretability. These changes were the culmination of extensive data analysis and clinical reviews. Artificial intelligence models were used to enhance the fit.

Now with Benchmarks

The DRG now outputs utilisation benchmarks based on industry statistics. This includes median and third quartile lengths of stay by ward type, as well as theatre time benchmarks. These benchmarks can be used to enhance preauthorisations and case management protocols and to facilitate more robust retrospective case reviews. They can also be used by facilities working to improve their own utilisation metrics.

Effectiveness

A more effective DRG better explains variations in the cost per admission. A statistical measure known as the R-squared can be used to measure the proportion of variability explained by the DRG [2].
The latest iteration of the DRG has a substantially higher R-squared than previous versions which bears testament to the progress made. When coupled with the Insight hospital efficiency assessment methodology, the DRG yields an R-squared of 82%.

Adoption

Outside of Discovery Health which employs its own proprietary software, medical schemes representing more than ninety percent of beneficiaries rely on the Insight DRG directly or through their managed care organisations. Six of the seven largest hospital groups use the Insight DRG.

Broad support from healthcare funders and providers is critical to ensuring that the DRG allows for a common understanding of clinical and cost experience. This helps ensure that discussions on variations in the cost and quality of care and alternative reimbursement models are efficient and fruitful.

Looking forward

Insight looks forward to engaging with stakeholders across the healthcare sector on the review and application of our DRG. Feedback from clients and our relentless focus on continuous improvement is what keeps our DRG grouper the market standard.

[1] All the admissions within a DRG are expected to be associated with similar levels of resource utilisation and similar costs.

[2] R-squared ranges between 0 and 1. A R-squared of 0 indicates that the model explains none of the variation. A R-squared of 1 indicates that the model explains all the variation. The higher the R-squared, the better the explanatory power of the model.