Brief Report: Use of the Consolidated Framework for Implementation Research (CFIR) to Characterize Health Care Workers' Perspectives on Financial Incentives to Increase Pediatric HIV Testing

Abstract

Background: 

A prior randomized control trial showed financial incentives increase HIV testing rates for children of unknown HIV status. Translating evidence-based interventions such as these to scale requires an implementation science approach.

Methods: 

A qualitative study evaluating health care providers' perceptions of barriers and facilitators of a previously completed financial incentives intervention for pediatric HIV testing was conducted at health care facilities in Kisumu, Kenya. Six focus group discussions with 52 providers explored determinants of acceptability, feasibility, and sustainability of financial incentive scale-up for pediatric HIV testing using the Consolidated Framework for Implementation Research to inform question guides and thematic analysis.

Results: 

Providers found the use of financial incentive interventions for pediatric HIV testing to be highly acceptable. First, providers believed financial incentives had a relative advantage over existing strategies, because they overcame cost barriers and provided additional motivation to test; however, concerns about how financial incentives would be implemented influenced perceptions of feasibility and sustainability. Second, providers expressed concern that already overburdened staff and high costs of financial incentive programs would limit sustainability. Third, providers feared that financial incentives may negatively affect further care because of expectations of repeated financial support and program manipulation.

Conclusions: 

Providers viewed financial incentives as an acceptable intervention to scale programmatically to increase uptake of pediatric testing. To ensure feasibility and sustainability of financial incentives in pediatric HIV testing programs, it will be important to clearly define target populations, manage expectations of continued financial support, and establish systems to track testing.

INTRODUCTION

Children lag behind adults in reaching the UNAIDS 95-95-95 goals for HIV testing, treatment, and suppression.^1,2 Index case testing is an efficient strategy for case detection of HIV positive children, but has historically had suboptimal uptake, limiting coverage.³ Barriers to testing children for HIV include emotional, structural, and logistical barriers, including costs of missing work and travel to clinics.⁴ Financial incentives, a conditional monetary incentive to motivate behavior, are an attractive intervention to address these testing coverage barriers.

Financial incentives have been shown to increase uptake of HIV testing for adults and adolescents in fixed value and lottery based financial incentives studies and programs.^5–10 Financial incentives can directly offset costs associated with child HIV testing and motivate caregivers who are willing to test their children to take action earlier, acting as a behavioral “nudge.”^11,12 A recent randomized controlled trial of financial incentives to increase uptake of pediatric HIV testing through an index case testing model (FIT,¹³ NCT03049917), conducted in Kenya, found that financial incentives were feasible to implement, and significantly increased uptake of testing¹⁴ among caregivers who were offered financial incentives.

Evidence-based interventions often face a delay in broad implementation in practice.¹⁰ This delay negatively affects health systems' ability to promptly address health concerns. To identify factors influencing implementation outcomes of acceptability, feasibility, and sustainability of the FIT intervention, we evaluated providers involved in the FIT trial's perception of the barriers and facilitators of a proposed financial incentive program. Evaluating perceived obstacles and benefits to a large-scale implementation of a financial incentive to motivate pediatric HIV testing could better inform pediatric HIV testing related health care policy decisions in the future.

METHODS

Study Design and Population

This qualitative study was conducted in 6 HIV clinics in Kisumu County, Kenya where the original FIT trial was conducted between January 2017 and October 2018. Kisumu County has a prevalence of HIV infection of almost 20%^15,16; and is primarily an urban environment, with some clinics also attracting patients from more outlying, rural regions. The nonstudy staff facility supervisor at each clinic facilitated recruitment, identifying 7–10 providers who were ≥18 years of age and had experience with pediatric HIV testing to participate in focus group discussions (FGDs).

Data Collection and Analysis

We used the Consolidated Framework for Implementation Research (CFIR) to inform data collection and analysis. CFIR is a meta-theoretical framework used to assess determinants of implementation¹⁷ and provides a menu of 39 constructs within 5 domains that are theoretically associated with effective implementation. Based on the team's experience running the trial, the FGD guide was developed to probe a subset of specific constructs from 3 CFIR domains believed to be most influential in provider perceptions of acceptability, feasibility, and sustainability: inner setting, outer setting, and intervention characteristics.¹⁸ A broad range of provider cadres (ie, nurses, counselors, clinical officers, pharmacists, and physicians) were purposively recruited from each facility to obtain diverse views on the topic. FGDs were conducted in English by a trained Kenyan social scientist (K.O.) and notes were taken by a separate person (D.L.A.). The FGDs were audio recorded and transcribed verbatim (D.L.A., V.O.). FGDs lasted an average of 95 minutes.

ATLAS.ti version 8 was used to support data analysis and management. Transcripts were analyzed using directed content analysis guided by CFIR constructs.¹⁹ Coders (D.L.A., J.Z.) performed multiple rounds of consensus coding, where the same transcript was reviewed and coded by all coders to ensure consistent code application. Following code application agreement, each transcript was independently coded by one member of the coding team and coding application was reviewed by another team member and disagreements in code application were noted. All disagreements were resolved through group discussions with the larger team. Code co-occurrence tables and queries were used to identify the constructs most influential on implementation outcomes (acceptability, feasibility, and sustainability).

Ethics Statement

The University of Washington Institutional Review Board and Kenyatta National Hospital and University of Nairobi Ethics and Research Committee approved the study. All study participants provided written informed consent.

RESULTS

A total of 52 providers participated in 6 FGDs (8–10 providers each). Providers represented a broad range of cadres, age groups, and experience (Table 1). An overview of CFIR constructs described and their influence on provider beliefs about future scalability of financial incentives interventions for pediatric HIV testing are presented in Table 2.

TABLE 1.: Health Care Worker Demographics

TABLE 2.: Illustrative Quotes Related to Acceptability, Feasibility, and Sustainability

Acceptability: Financial Incentives Provided a Relative Advantage Compared With Current Pediatric HIV Testing Strategies

Providers believed that financial incentives provided a relative advantage over current testing strategies for children—including passive index case testing—by providing additional motivation for caregivers who were reluctant to test or for whom finances were a barrier to getting to clinic. Providers suggested that pediatric populations may not be the only population to target with financial incentives, and that financial incentives could meet the barriers experienced by adolescents and adult men, who also have financial challenges accessing testing and merit equal prioritization. In addition, a few providers noted how financial incentives for HIV testing could act as a platform for identification of other non-HIV health issues and strengthen provision of other services. All providers believed that financial incentives had potential to improve pediatric HIV testing uptake and almost all believed financial incentives would be an acceptable intervention to scale-up.

Acceptability: Financial Incentives Would Not Address All Patient Needs and Resources

Although viewed as providing a relative advantage over existing programs, such as encouraging caregivers to test children while participating in partner notification services, providers noted that financial incentives would not motivate all caregivers, especially those where financial burden was not the driver for testing refusal. Financial incentives were not seen to overcome barriers to testing related to challenges with HIV disclosure to children and spouses. Providers also expressed concerns about the negative impact that receiving a financial incentive may have, fearing that caregivers who receive incentives for testing would come to expect incentives for other visits, leading to decreased linkage to care services and increased rates of loss to follow-up. Providers felt that in the long run, establishing a strong sense of agency for health-seeking behaviors in the community—improving intrinsic motivation—could replace the need for financial incentives programs—which rely on extrinsic motivation—while still achieving high rates of pediatric HIV testing.

Feasibility: Available Resources Challenged Perceived Feasibility of Financial Incentives Implementation

Although many providers believed a financial incentives intervention was acceptable, providers noted concerns regarding the resources that would need to be allocated for implementation. Specifically, providers believed financial incentives may divert staff time away from existing responsibilities. In addition, providers already experience space limitations, noting that testing was often conducted in tents outside the facility and in places with limited privacy. Financial incentives were anticipated to increase the demand for pediatric HIV testing, which could strain already overburdened sites.

Participants also identified the absence of a tracking system to identify eligible children for testing as a limitation that would affect feasibility and have potential for manipulation. Providers described hypothetical situations where caregivers may cycle through visiting multiple facilities, testing and retesting their children of known status, to collect the financial incentives. Within individual facilities, providers described how some caregivers may be motivated to bring children other than their own for testing to profit financially. Providers felt that program manipulation barriers could be overcome using electronic databases to avoid repeat testing between clinics or using tracking systems that allow pre-financial incentive intervention verification of caregiver–child relationships.

Sustainability: Perceived High Cost of a Financial Incentives Program Limited Provider Beliefs About Sustainability

Providers were concerned about the costs of sustaining a financial incentives program long-term. Providers were divided on the need for providing a standard compensation rate or allowing rates to vary based on distance from the clinic. Some providers felt that a varied compensation rate could cause tension between neighbors and communities, and therefore believed a fixed value would be better. Others felt that caregivers would be affected differently based on distance travelled or how many children brought, making a varied compensation rate preferable. Regardless of rate, high perceived costs associated with distributing money to patients limited provider perceptions of sustainability.

DISCUSSION

This qualitative study used the CFIR to guide identification of determinants influencing future scalability of a financial incentive intervention for pediatric HIV testing. The study engaged a broad range of providers and noted general enthusiasm and acceptability for translating financial incentives into a programmatic setting, concerns about unintended consequences and feasibility, and concerns about costs limiting sustainability. Providers felt financial incentives offered a relative advantage over existing interventions by motivating individuals for whom cost was a barrier, but not those who faced other personal or psychosocial barriers. Providers raised concerns that clients may become dependent on financial incentives to continue seeking health services. As with many interventions, the additional workload being added to an already overburdened system caused concern for some providers. Despite noted challenges and suggestions for implementation, overall acceptability of a proposed financial incentives program for pediatric HIV testing was high.

Prior studies of financial incentives to promote HIV testing uptake in adults, as well as adolescents and children, have found financial incentives to be effective,^8,20–23 but have not empirically assessed the effect of financial incentives on subsequent health-seeking behavior. However, one ART adherence study assessed the impact of financial incentives on intrinsic motivation and reported no negative impact.²⁴ Despite concerns about the impact of receiving a financial incentive on future health-seeking behavior among children who test positive, caregivers in the FIT trial reported being more likely to seek care in the future for their newly tested positive and negative children because of no longer fearing an HIV test.²⁵ Financial incentives have been scaled up programmatically in non-HIV settings to promote child welfare,²⁶ and support orphans and vulnerable children.²⁷ However, to date most of these financial incentives have been provided unconditionally, or with a condition for the child to remain in school; additional studies are needed to determine the long-term impact of providing financial incentives to caregivers to promote specific health-seeking behaviors for children.

Program cost was perceived as a barrier to programmatic scale-up and sustainability. Analysis of the cost and cost-effectiveness of this intervention is underway. To date, cost-effectiveness analyses of financial incentives interventions and pediatric HIV testing interventions are limited.^3,28 One recent study that assessed the cost-effectiveness of financial incentives found that a conditional fixed incentive of $3 and $10 increased identification of new diagnoses in adults and decreased the cost per person tested and per positive individual identified.^29,30 It is possible that financial incentives for pediatric HIV testing may also lead to cost savings by increasing efficiency of fixed resources—such as provider salaries and facility operating costs—and by identifying HIV-infected children before symptomatic disease. However, as providers in this study noted, it is also possible that incentives may lead to an increased volume of patients seeking services, potentially further burdening fixed numbers of providers or necessitating additional provider hires. Having sufficient human resources is a key building block of health systems, but human resource shortages are a feature of many resource-limited health systems. Motivating fixed numbers of providers to take on additional responsibilities may also incur costs. One strategy would be to consider performance-based financing alongside financial incentives for testing, an effective strategy that has been shown to also improve coverage of HIV testing,³¹ but which may have complex and unintended consequences.

Providers suggested that using existing patient data tracking systems was a strategy that could overcome the potential barrier of manipulating the testing system for the purpose of financial gain. Studies in Kenya and Malawi have used a family identification card to support index case testing and found high testing coverage.^32,33 This card system has since been incorporated into HIV care programs nationally in Kenya, suggesting that using programmatic records to confirm caregiver–child relationships would be a feasible and effective approach if financial incentives were scaled to a programmatic setting.

Expanding financial incentives programmatic coverage to include adolescents and adult men is a reasonable expansion of this strategy. Financial incentives increased uptake of adolescent HIV testing in Zimbabwe,²⁰ and among male partners in Uganda and Malawi^21,30 and were found to be cost-effective among male partners.³⁰ There may be programmatic efficiencies and increased perception of equity in offering financial incentives to several populations within one program.

This study elicited perspectives from a wide range of provider cadres who had experience with pediatric HIV testing and who had been exposed to the FIT trial. It also benefitted from having a strong conceptual model, CFIR, which prospectively informed the creation of FGD question guides and analysis. CFIR was created specifically for implementation science and identifies determinants of successful implementation. It has a strong history of use in implementation science in the US,¹⁷ and emerging use in global health.^34,35 This study was limited in that the providers did not have direct experience delivering financial incentives in a programmatic setting. Therefore, providers' concerns about barriers to successful implementation are hypothesized, rather than experienced. Opinions from caregivers are planned in future analyses.

CONCLUSIONS

Financial incentives were perceived to be acceptable, but with feasibility concerns, and challenges to long-term sustainability, primarily driven by overextended health systems and increased costs. Acceptability of scale up may be enhanced by including men and adolescents, managing expectations for future financial support, and leveraging existing routine data for tracking programmatic eligibility.