Background
The voluntary health insurance marketed by the health plans (HP-VHI) covers about 77% of the population and fulfills an important role in the financing and provision of health services in Israel. Reports published in recent years by the Ministry of Health’s Branch for Supervision of the Health Plans and HP-VHI point to disparities between regions in the utilization of the services covered by the HP-VHI.
Objectives
The Ministry of Health asked the Myers-JDC-Brookdale Institute to assess the disparities in HP-VHI utilization across geographic regions, while controlling for health status and socioeconomic characteristics.
The research questions were as follows:
- In which HP-VHI services are there disparities in utilization by place of residence and what is the magnitude of those disparities?
- Have these disparities changed in magnitude during the decade between 2012 and 2021?
- Are there disparities in the probability of using HP-VHI services by place of residence even after controlling for differences in needs (as measured by health status indicators) and what is their magnitude?
- Which background and socioeconomic characteristics (age, gender, population group, household size, education and income) have an effect on the probability of using these services?
Methods
The study is based on data gathered by the “Public Opinion on the Level of Service in the Healthcare System in Israel and its Functioning” survey carried out in 2012 and 2021 among the adult population (aged 22+) in Israel. The following HP-VHI services were examined: choice of surgeon; choice of private hospital; visits to a specialist; discounts on dental treatment; discounts on the purchase of drugs; fertility treatment, pregnancy, and childbirth services; and child development services.
Bivariate analyses (ꭓ2) were carried out to estimate the differences in the rate of utilization for each of the services by region for 2012 and 2021. Subsequently, multivariate models were constructed for 2021 for each of the six services based on three-stage logistic regression. The dependent variable was use of a particular service at least once during the two years prior to the survey. In the first stage, the only explanatory variable was the region of residence. In the other two stages, in which additional regressions were estimated, control variables were added: population group (ultra-Orthodox Jews, Arabs, and non-ultra-Orthodox Jews), education, income, gender, age and health status.
Main Findings
For the population as a whole, the reported rate of utilization of at least one HP-VHI service two years prior to the survey fell from 76% in 2012 to 67% in 2021. The only service in which there was an increase in utilization was visits to a specialist. The multivariate analyses carried out on the data for 2021 showed that the region of residence accounts in part for utilization differences for some of the services, even after controlling for other variables. Other key findings were as follows:
- Health status was found to be associated with HP-VHI utilization. Respondents who reported a mediocre, poor, or very poor health status were more likely to visit a specialist and to use discounts on the purchase of drugs. Chronically ill respondents had a high probability of using discounts on the purchase of drugs.
- Income in the lower three quintiles and a non-academic education lowered the probability of visiting a specialist while a non-academic education of the parent lowered the probability of using the child development service.
- Arabs had a higher probability of using fertility treatment, pregnancy, and childbirth services and of choosing a surgeon and a private hospital than non-ultra-Orthodox Jews. Ultra-Orthodox Jews had a higher probability of using discounts to purchase drugs than non-ultra-Orthodox Jews.
Conclusion and Discussion
This is the first in-depth assessment of the utilization of HP-VHI services and the first attempt to estimate disparities between population groups. The data for 2021 show differences in utilization for some of the services according to region of residence, even after controlling for other variables. However, it is important to emphasize that the study does not include data on the supply of services (such as availability, as measured by per capita physician hours per unit of time) or measures of access (ability and desire of the patient to obtain the available services) and therefore the disparities that were found among the regions cannot be attributed directly to differences in availability of or access to services. Moreover, HP-VHI expands, supplements and duplicates the coverage of national health insurance (NHI) and therefore policy with regard to NHI coverage, as well as the availability of and access to NHI services, affect the probability of HP-VHI utilization.
Furthermore, other factors were found to be associated with HP-VHI utilization, including age, health status, population group (Arabs/ultra-Orthodox) and household size. Income and education explain some of the variation in the use of services and suggest the existence of barriers to access for some services. This should be taken into account when considering interventions to reduce inequality.