NEWS

On openIMIS and actuarial analysis

The Bhela meetings in May and June explored how actuarial analysis strengthens health insurance design and how openIMIS and the ILO/Health actuarial tool can contribute to this.

What is actuarial analysis?

In his opening presentation at the Bhela meeting in May, Jens Geissler, Professor for Health and Social Sector Management at FOM University in Hamburg, Germany, differentiated between actuarial analysis for private versus public or social health insurances. For the latter, actuarial analysis applies mathematical and statistical methods to assess the financial implications of different health insurance design options, for example in relation to the benefit package or to contributions. Data processed in health insurance IT systems is one extremely important source of this analysis. However, the primary function of these IT systems and the data are the day-to-day operations of the health insurance like collection of premiums, and verification and payment of claims. Actuarial analysis enables the use of this data for policy making, enabling health policy makers to ask the right questions and answer them with real-life data. Health insurance IT systems like openIMIS can generate data for actuarial analysis in addition to supporting insurance operations.

How does openIMIS support actuarial analysis in Nepal?

Purushottam Sapkota, Advisor on Integrated Health Information System at GIZ Nepal, provided insights into the openIMIS implementation status in Nepal. He explained that, with more than 30.000 claims processed daily and almost 500 empaneled service providers serving clients in the voluntary and the mandatory schemes, openIMIS in Nepal manages a wealth of data. This ranges from members’ demographic and contribution information to data generated in the claims process, including health facility, medical procedures and amounts claimed. The Health Insurance Board of Nepal, which is responsible for the voluntary scheme, already has access to both beneficiary and claims data for actuarial analysis.   

Pratima Rai, Jr. Advisor – Social Health Protection at GIZ Nepal, gave specific examples of the ways actuarial analysis is applied in Nepal. Since 2024, the Health Insurance Scheme is implemented in all 753 health facilities of the country. According to Pratima, actuarial analysis is applied to calculate how the benefit package can be expanded and how complete health insurance coverage of Nepal’s population could be reached. In the discussion that followed Pratima and Jens talked about the specific challenges in Nepal resulting from the expansion of coverage from less than 25.000 insurees in 2016 to almost 5 million insurees in 2024. Ultimately, insuring the entire population of more than 30 million will require evidence-based financing and investment decisions by the Health Insurance Fund to ensure the long-term financial viability of the scheme. Puru highlighted that openIMIS is already contributing to this decision making by providing the government with extensive data on on-going health insurance operations and expenditures. 

What needs to be considered when openIMIS data are used for actuarial analysis?

In his introductory presentation at the Bhela meeting in June, Prof. Geissler pointed at the challenges in using business process data for actuarial analysis. One aspect is possible bias in real-world data: The data may only include data on some population groups, types of health care services or regions, depending on the specific health insurance legislation of the country or scheme. Using this data for actuarial projections poses the risk of drawing wrong conclusions. It is the role of actuarial experts to ensure that existing biases are considered when making actuarial projections.

What is the ILO/Health actuarial model?

Andrés Acuña-Ulate, Social Protection Actuary in the Social Protection Department of the ILO, introduced the ILO/Health actuarial model. He highlighted that actuarial analysis could help, both in understanding historical data, and in making projections for future scenarios. In both cases, the specific health insurance systems of the country must be considered. Andrés illustrated two ways of making projections on the financial stability of health insurance schemes: by calculating from the demand side, i.e. multiplying a specific covered population with the costs resulting from a specific treatment protocol; and by analyzing the supply side of a health insurance scheme e.g. in terms of numbers of consultations per insuree per year or percentage of emergency visits and multiplying these with the costs per procedure. The result in both cases is a projection on the financial sustainability of the specific scheme. This projection can inform policy making, e.g. by showing the need for changes in contribution rates or subsidies. Actuarial analysis can support this decision making but cannot replace it. 

Could and should there be automated data transfer between openIMIS and ILO/Health?

The discussion focused on the relation between openIMIS and ILO/Health, specifically whether there could or should be an automatic transfer of data between the two systems. One participant pointed out the risk of mistakes in a manual data transfer process. Andrés highlighted the need for quality assurance, mentioning examples where operational data from an insurance IT system needed to be adjusted to reflect the actual dates of birth of insurees. It could be feasible to develop data sets in openIMIS which match the data requirements of ILO/Health, but even in this case the actuarial analysis should never be an automated black box which produces results without the scrutiny of actuarial experts. Andrés concluded by saying that the standardisation of data in openIMIS could facilitate the process of actuarial analysis: an actuarial expert who knew the openIMIS data from one country could more quickly do the quality assurance for the data from another country. 

 

For further information and recordings to the presentations, please use the following links: