Designing feasible anti-corruption strategies in the Nigerian health system: A latent class analysis of a discrete choice experiment. (June 2023)

• Record Type: Journal Article
• Authors: Angell, Blake; Onwujekwe, Obinna; Roy, Pallavi; Nwokolo, Chukwudi; McKee, Martin; Mandeville, Kate; Obodoechi, Divine; Agwu, Prince; Odii, Aloysius; Orjiakor, Charles; Hutchinson, Eleanor; Balabanova, Dina
• Is Part Of: World development. Volume 166(2023)
• Publication Date: 2023-06
• Subjects: Anti-corruption -- Nigeria -- Health system -- Absenteeism -- Primary health care; Economic history -- 1990- -- Periodicals; Economic assistance -- Developing countries -- Periodicals; 330.9
• DOI: 10.1016/j.worlddev.2023.106208
• Languages: English
• ISSNs: 0305-750X

Abstract

Highlights: Interventions targeting health worker absenteeism in Nigeria have ignored incentives facing workers and been unsuccessful. To inform sustainable anti-corruption strategies, we used a discrete choice experiment to elicit data on these incentives. All attributes significantly influenced choices; interventions to target these could affect health worker behaviour. Two distinct groups were found in our sample, who would potentially respond differently to interventions in practice. Addressing the needs of those who want to be at work offers a path to incrementally reduce the level of absenteeism. Abstract: Health worker absenteeism is a major form of corruption in the Nigerian health system, reducing the impact of healthcare investment and disproportionately affecting vulnerable communities. Top-down governance and accountability measures to improve attendance have had limited effect, potentially because they have not addressed underlying incentives driving health worker behaviour. To inform feasible interventions to reduce absenteeism, we sought to investigate how potential interventions targeting key drivers of absenteeism could influence behaviour through a discrete choice experiment with 412 primary health care workers in Enugu State, Nigeria. The experiment included six attributes drawing on a literature review, qualitative research and stakeholder consultation: facility relationship with the community, method of supervision, attendance-related rewards or punishments, flexible working policy, posting location and financial incentives. Respondent choices were analysed with mixed multinomial logistic and latent class models and used to simulate job uptake under different policies. We found all attributes significantly influenced choices: workers strongly preferred jobs that facilitated supportive community relationships (β = 0.37), had a financial incentive worth 10 % of a worker's salary (β = 0.39), rewarded attendance with credit towards promotion (β = 0.25) or placement in training (β = 0.14), and offered flexible work conditions (β = 0.14). Using latent class analysis, we identified two sub-groups of workers who differed in their willingness to accept a job and preferences over specific attributes, suggesting some may be more susceptible to feasible interventions to improve attendance. We argue that focusing on the needs of these groups may offer a path to incrementally, but sustainably, reduce absenteeism across the system. Our findings build on frameworks developed in anti-corruption research in other sectors only recently been applied to the health sector. The work represents some of the only empirical evidence to inform anti-corruption strategies in Nigeria and internationally, with findings relevant to other low- and middle-income countries seeking to combat absenteeism among primary health care workers.