Community Engagement’s Impact on Healthcare Utilization and Health Insurance Enrollment in Ghana

Stephen Duku
Christine Fenenga
Wendy Janssens
Edward Nketiah-Amponsah
Fieldwork by:
Greater Accra and the Western Regions
64 healthcare facilities; 1920 households
Target group:
  • Families and households
Outcome of interest:
  • Health outcomes
  • Take-up of program/social service/healthy behavior
Intervention type:
  • Health care delivery
  • Insurance
  • Preventive health
  • Community health workers
  • Pricing and fees

Many Sub-Saharan African countries struggle to get their citizens to take up affordable or free health insurance. This might be due to low quality service, leaving people wondering if health insurance is valuable. Researchers in Ghana evaluated the impact of community-led assessments on health insurance enrollment as well as healthcare and health insurance quality. The community engagement intervention improved the medical quality of healthcare, reduced community members’ chances of getting sick, and increased health insurance enrollment among previously uninsured people. However, households’ views on the quality of healthcare and insurance services did not improve.

Policy issue

Many Sub-Saharan African countries struggle to get their citizens to take up affordable or free health insurance, despite the financial risks of being uninsured. This may be due to low quality service by covered providers, leaving people wondering if health insurance is valuable. When households believe covered providers offer poor services, or if they believe providers offer better service to clients paying out-of-pocket, they may seek services from non-covered providers like private clinics or pay out-of-pocket in an attempt to receive better care. People may also hesitate to sign up for insurance due to difficult registration procedures, long waiting periods, or poor management of complaints by the insurer. However, research is not conclusive when it comes to engaging community members, healthcare providers, and insurers to improve the actual and perceived quality of health and insurance services, which could build trust and encourage people to enroll in health insurance. Can a community engagement program improve the quality of available healthcare and health insurance services, and shift households’ perceptions of these?

Context of the evaluation

The evaluation took place in Ghana’s Western Region and Greater Accra Region. In 2010, the Western Region was rural, with roughly two million people or 9.6 percent of Ghana’s population. Greater Accra’s population was mostly urban in 2010, with four million or 16.3 percent of Ghana’s people.  

Though the government subsidized health insurance, only 38 percent of residents in the country were enrolled in National Health Insurance Scheme (NHIS) insurance in 2013. In the catchment areas of health clinics participating in this study, 42 percent of people surveyed had insurance from the NHIS, and in the preceding six months, a third of survey respondents were sick at least once and 38 percent visited a health clinic for services. Health insurance enrollment was likely low due to low quality healthcare leading citizens to prefer alternative providers like private clinics and traditional practitioners. Unwieldy enrollment processes and delays in getting insurance credentials also served as obstacles. 

In the study catchment areas, survey respondents were, on average, 26 years old from a household with five people. About half of respondents were female or lived in rural villages. Forty-nine percent of individuals had at least finished primary school. Forty-three percent of respondents worked over the last year. Annually, the average household spent GH₵3,972 or US$946 (in 2017 dollars).

Group discussion about health insurance enrollment, healthcare, and health insurance quality with a group men from the village
A group of men discuss health insurance enrollment, healthcare, and health insurance quality in Greater Accra, Ghana.
Fenenga, Christine | Department of Health Science, University Medical Centre Groningen, Groningen, The Netherlands

Details of the intervention

Researchers evaluated the effect of a community engagement program on health insurance enrollment and healthcare and health insurance quality in Ghana. Researchers randomly assigned 64 government-funded healthcare clinics from the Greater Accra and Western Regions to one of the following groups: 

  1. Community engagement group (32 clinics): In this group, each health clinic participated in a community engagement program for just under a year, in partnership with one or two active community organizations in the area surrounding each of the 32 health clinics (53 groups total). The community engagement activities involved facilitated feedback meetings between community group members, medical professionals, and insurance staff. Community group members first completed quality assessments of healthcare and NHIS services and then shared their findings with health clinic leaders and NHIS leadership. Together, they agreed on quality improvement plans and followed up with medical providers and insurance staff three months later to see if the plans were being implemented. After another three months, the best-performing health clinic in each district could receive an award based on a second round of community assessments to encourage competition between clinics to improve their performance.
  2. Comparison group (32 clinics): In the comparison group, health clinics did not participate in any of the above community engagement activities. 

The researchers surveyed a randomly selected set of 1920 households total living across each of the catchment areas of the program and comparison group health clinics in 2012 and 2014, before and three months after the intervention, to assess the quality of health care and insurance services. Respondents in the community engagement group catchment areas did not have to directly participate in any of the community engagement activities to participate in the survey, though they may have been aware of the activities through their community groups.

In addition to receiving ethical review and approvals from an institutional review board (the Ghana Health Service Ethical Review Committee, clearance numbers: GHS-ERC: 18/5/11 and GHS-ERC 08/5/11), researchers made efforts to address and account for ethical questions by obtaining informed consent from individual respondents in the communities for the baseline and follow-up surveys, and also from community group/association leadership for the community engagement assessments. Literate respondents provided written informed consent while illiterate respondents thumb-printed the informed consent form before participating in the study. For more on the researchers’ discussion of ethical considerations, see page number 2132 of the paper.

Results and policy lessons

The community engagement program did not improve patients’ perception of the quality of healthcare or health insurance, despite the technical quality of medical services that patients received improving. However, the intervention group’s health outcomes were better than the comparison group’s health outcomes. Formerly uninsured patients in the intervention group took up health insurance at higher rates than their insured counterparts.

Citizens’ perceptions of healthcare and insurance: The community engagement program did not improve households’ views on government-provided healthcare or health insurance, despite researchers finding notable improvements in the medical-technical quality of care. It may be that citizens value or notice improvements in non-technical aspects of healthcare more, such as provider attitudes and waiting times, while providers focused their efforts on improving technical aspects of healthcare, like following defined protocols of care.

Illness and visits to healthcare clinics: The community engagement program improved health outcomes for households in the program catchment areas. While the incidence of illness was higher among both participants in the program and comparison groups by 2014, households in the program group catchment areas had a 67.2 percent chance, on average, of getting sick. They were 7.6 percentage points less likely to get sick than their counterparts in the comparison group, who had a 74.8 percent chance, on average, of getting sick. The community engagement program decreased households’ chance of visits to healthcare facilities to 31 percent, a reduction of 4.8 percentage points, relative to the comparison group, who had a 35.8 percent average chance of visiting a clinic.

Insurance enrollment: After the intervention, uninsured households in the community engagement catchment areas were 5.9 percentage points more likely to enroll in the NHIS insurance than their counterparts in the comparison group, who had a 38 percent chance of being enrolled at endline. If the comparison group participants were excluded, enrollment among the previously uninsured increased by 10.6 percentage points. Already insured households did not change their enrollment. Researchers suggest that involving patients in assessing and improving health clinics and insurance could encourage them to enroll in health insurance in greater numbers.

The health clinics and NHIS have not adopted the intervention exactly how it was implemented during the evaluation, but they did pay more attention to engagement of clients and input from users through regular meetings at the district-level. They also created a well-announced phone number in case of complaints. 

Duku, Stephen Kwasi Opoku, Edward Nketiah‐Amponsah, Christine J. Fenenga, Wendy
Janssens, and Menno Pradhan. “The Effect of Community Engagement on Healthcare Utilization
and Health Insurance Enrollment in Ghana: Results from a Randomized Experiment.” Health
Economics 31, no. 10 (2022): 2120–41.