Emerging Project Lessons
The selection of an implementing partner and an HMI scheme to pilot-test an intervention such as consumer education should be well-structured and carefully considered.
Freedom from Hunger established and followed several criteria for selecting a test site and partner:
· Scheme had to be operational, stable, and with a membership of at least several thousand; ideally it will be a private-sector product
· Product had to be appropriate for the poor, i.e. the coverage is appropriate for the risks faced by poor consumers, the product is simple to understand and access, and the premium is affordable (or perceived to be)
· Adequate access to acceptable healthcare providers
· Distribution partner capable of reaching MFI clients, self-help groups or other organized groups with education
Based on the overall experience of working with SAT in the field test, including staff training, Freedom from Hunger thinks that the upfront effort and extra time to select a partner was a worthwhile investment, and that even more scrutiny of the organization at the level of local operations would have been well-advised. SAT came to the project with enthusiasm for linking more of its clients to health insurance, and with pre-existing capacity for training and delivery of education. Non-financial services such as education were reported by SAT senior management as being an integral part of its mission and goal to improve the lives of its clients. However, a number of challenges emerged later in the project. Randomized control study design requires extra effort on communication. Furthermore, researchers must be able to identify and access clients for baseline interviews. A capacity to maintain the research protocols throughout the process is also vital to protect the integrity of the design and reliability of the findings.
Freedom from Hunger’s consideration of potential partners would have benefited from closer examination of the operations of the MFI at the level of the operating branch in the area where the study was to be launched. In particular, after the fact, Freedom from Hunger discovered that repayment issues as well as management challenges were occurring in one of the branch offices where the study was conducted. Although it is unknown how discoverable these would have been prior to the study, it is a lesson for future efforts that are intended to demonstrate and evaluate a specific intervention. On the other hand, since interventions such as health insurance education, or even the introduction of new health insurance schemes, will always need to work in the “real” world, the challenges at SAT with ensuring uniform programme roll-out and data-collection are not atypical of those that will be present in further replications.
Plan adequate time to identify pilot-test sites with an existing health microinsurance product and a capable implementing partner. In hindsight, the time frame established by Freedom from Hunger for finding a suitable local partner, product and target market following project kick-off was overly optimistic. In reality, the field of HMI is still nascent, with limited availability of appropriate and affordable commercial products, and while MFIs are perceived to be a viable and promising distribution channel, the real numbers of those who are actively promoting and offering appropriate products is still quite small.
Working with public health insurance programmes can mean unforeseen challenges due to shifting government policy and changes in financing, management or infrastructure issues. A private HMI product would have been ideal, since in theory data would be more readily available for enrolment and claims, and results could be more transferable to other schemes, but this turned out to be an additional constraint that ultimately had to be relaxed. Given that the product being tested is Ghana’s NHIS, enrolment in the scheme heavily depends on the financial subsidy of the premium, how the government manages the scheme, and the quality of the healthcare providers and services. It is also subject to the effects of any changes in government policy with respect to premium payment, eligibility, and covered services. For example, due to political and financial events, midway through the study the Ghanaian government indicated its plan to restructure the annual premium of $8 to $9 to a single lifetime premium, which would likely cost several times more. Although in the end this change was not implemented during the study period, news reports of its imminence could have affected enrolment decisions and fueled ongoing confusion among those eligible regarding the costs and terms of insurance.
Additionally, access to claims and enrolment data was blocked due to government rules, concerns over confidentiality, lack of resources, and administrative bottlenecks to enrol members and process claims. During the time of the study, the NHIA was operating with delays of more than three months to issue identification cards, and had claims backlogs of six months or more.
Finally, the National Health Insurance Fund (NHIF) operates with a complex system of registration (valid for five years) coupled with annual enrolment and premium payment (except for children, pregnant women and people over age 70); there is an initial three-month waiting period, and often identification cards are delivered after the effective date, leading to considerable confusion on the part of clients as to whether, and when, they are actively enrolled.
Findings of lack of impact, while disappointing, can still be helpful towards the overall goal of understanding barriers to insurance take-up and to inform future strategies to improve enrolment of poor families.
While it is undeniable that our findings at midline and endline were disappointing, they should not have been entirely surprising given the unexpected findings of high levels of insurance knowledge, awareness and positive attitudes about the insurance at baseline. There were many contextual factors at work in Ghana and with the implementation that quite likely limited the project’s ability to completely evaluate the value and contributions of the education. These are discussed in great detail in the research report. Further, the education module improved knowledge (even at already high levels). These findings do not lead to the conclusion that the module would not improve enrolment in other contexts.
The findings of high initial levels of knowledge, along with the results from midline and endline, surfaced an important follow-up question: If knowledge was not the barrier to greater enrolment, then what was? As a result, a qualitative study component was incorporated into the endline quantitative study. This involved adding some qualitative questions to the survey about why people were not enrolled, as well as focus groups with clients. It was this process of inquiry that enabled Freedom from Hunger to reach important conclusions about the most important barriers in this setting: the availability of cash on hand to pay premiums when needed and the failure to enrol despite intending to do so.
Date of last Learning Journey update: March 2013
