Towards the Inclusion of a Health Status Parameter in the Israeli Capitation Formula

In accordance with the National Health Insurance Law, each year the government of Israel distributes over NIS 20 billion among the nation’s four competing health plans on the basis of a capitation formula. The formula has two main objectives:

  • To reduce the incentives for the health plans to favor low-cost members over high-cost members
  • To distribute resources among the health plans in a way commensurate with the needs of the members

The current formula takes into account only the number of members in each health plan and their age distribution. There is a broad consensus among disinterested parties that in distributing monies among the health plans it is desirable to also take into account the health status of the members. However, there is considerable debate in Israel on how to do so and, even more fundamentally, whether Israelhas the data sources which would be needed to incorporate health status into the capitation formula.

This study explores how data from the Health Survey fielded periodically by the Central Bureau of Statistics and the Ministry of Health might be used as the basis for adding health status to the capitation formula. The 1993 and 1997 rounds of that survey included questions on health status, in the form of questions on the presence or absence of various chronic diseases.

A key concern about the possible use of survey data on health status for the purposes of constructing a capitation formula is that self-reports of health status may not be sufficiently reliable. This paper explores whether and how self-reported data on health status could be used to improve upon the existing Israeli capitation formula, if such data are reliable. It leaves to other studies and researchers the empirical assessment of the degree of reliability of self-reported health status, particularly in the Israeli context.

The study explores the pros and cons of various model specifications for analyzing the relationship between age, health status and health service utilization. It also demonstrates how data from national health surveys can be used to develop capitation formulas that 1) have good construct validity; 2) do a significantly better job of explaining variation in health service utilization than do formulas based on age alone; and 3) are relatively stable, both with respect to method of construction, and over time.

However, the study also found that formulas based on the 1993 and 1997 surveys were limited somewhat by the fact that those surveys did not include sufficient detail on key components of health service utilization or questions on functional and activity limitations and self-assessed general health status. Analysis of the 1997 Elderly Survey demonstrated that those variables can contribute significantly to the explanatory power of a capitation formula. As a result of the early dissemination of these findings, the year 2000 round of the Health Survey was refined and expanded to include more detailed questions on health service utilization as well as questions on functional and activity limitations.

Still, important questions remain about the advisability and feasibility of using survey data to determine the allocation of funds among health plans, and these, too, are spelled out in the report.

This paper is based on a cooperative project involving researchers from the JDC-Brookdale Institute and the Hadassah-Hebrew University’s Braun School of Public Health. The project was funded, in part, by the Israel National Institute for Health Policy and Health Services Research.

Citations in the professional and academic literature

Rosen, B., Goldwag, R., Thomson, S., Mossialos, E., & World Health Organization. (2003). Health care systems in transition: Israel-2003 (No. EUR/02/5037322 (ISR)). Copenhagen: WHO Regional Office for Europe.

Rosen, B., Goldwag, R., Thomson, S., & Mossialos, E. (2003). in Transition.

Gross, R., Bennun, G., & Rosen, B. Change in allocation formula to health plans.