Using nationally representative health data on children under five living in Sierra Leone, we found that the implementation of the FHCI in 2010 resulted in a 20 percentage points increase in the prevalence of care-seeking for febrile children. In addition, we identified several determinants of care-seeking behavior relative to children and caregivers of which, the household head gender, the region of residence, and the age of children were the most significant.
Our study revealed that 7 children under five in 10 sought care during fever after the FHCI when it was only 5 children under five in 10 before. The rise we observed in the prevalence of care-seeking for febrile children following the implementation of the FHCI in Sierra Leone is in keeping with the findings of Garchitorena et al. in Madagascar, where they found similar increase in the care-seeking in children U5 after two years of free health care initiation [20].
In the present study, considering the period following the implementation of the FHCI, the prevalence of care-seeking for febrile children was lower for households in the poorest quintile. The limited ability of the poorest household to bear the indirect costs of health care is a likely contributor to this difference. In addition, recurrent stock-out of drugs and supplies and the resulting out-of-pocket payment incurred by caregivers for health services despite the FHCI [21, 22], contribute to explain this difference between wealth quintiles in terms of care-seeking. A similar finding was reported in a previous study where the increase in the prevalence of care-seeking observed after the abolition of user fees left the most financially vulnerable people behind [13]. Another finding was the higher prevalence of care-seeking when the head of the household was aged over 25 years compared with that of household heads aged less than 25 years. A possible explanation could be a limited experience of young caregivers in the manifestations of childhood diseases as reported in studies conducted in Sierra Leone and Nigeria where the failure to recognize the symptoms of childhood febrile illnesses led to a delay in seeking care or not seeking care at all [23, 24].
We also found that children over one year of age had a low prevalence of care-seeking for fever. The same association was found in other studies in Africa [25, 26]. This could be explained by the fact that the caregivers of older children, having dealt with several episodes of childhood febrile illnesses as the child was aging, may have got experienced in treating fevers without resorting to a public health center.
In the present study, the prevalence of care-seeking for febrile children was lower in male compared to the female-headed households. These results highlight the gender influence at the household decision level in care-seeking decision/behavior for sick children. Care seeking behavior in caregivers of children was reported by Arthur E. in Ghana, Kenya and Zambia to be improved when both parents of the child were involved in the care-seeking decision [27].
In addition, the northern and western regions had low prevalence of care-seeking compared to the eastern region. The population-based surveys in 2013 and 2016 coincided with periods of outbreaks in Sierra Leone. It was the cholera outbreak from 2012 and the Ebola outbreak from 2015. During these outbreaks, the northern and western regions were the most affected [28, 29], and population in these regions could have avoided health centers for fear of contamination, this could explain the low prevalence of care-seeking in children under five in the context of free health care observed in our study. In addition, poor road conditions, availability of affordable and reliable transport means, and physical inaccessibility in the rains may contribute to explain the persisting low prevalence in care seeking in these regions despite the free-of-charge policy [30].
We acknowledge some limitations in our study. The study data collection might have been subject to a social desirability bias. Also, some aspects like social networks of the caregivers, the perception of the necessity of the healthcare services, and the causes of the illnesses of the child contributing to explain the care-seeking behavior could be better addressed by the qualitative research which was not included in the population-based surveys we used data for our analyses.
The strengths of this study include the large size of the sample and its representativeness of the population of children under five living in Sierra Leone.