Disparities of Plasmodium falciparum infection, malaria-related morbidity and access to malaria prevention and treatment among school-aged children: a national cross-sectional survey in Côte d’Ivoire

Plasmodium falciparum malaria remains a key global driver of mortality and morbidity with people in sub-Saharan Africa affected most [1, 2]. In Côte d’Ivoire, malaria is the primary cause of consultation in school health services and might be responsible for up to 40% of school absenteeism [3]. According to the world malaria report, the entire population of Côte d’Ivoire is at risk of malaria [2] and Anopheles gambiae is the primary vector species [4, 5]. However, there is a strong heterogeneity as wealthier people and those living in urban areas are at lower risk of malaria than poorer counterparts in rural settings [6].

Besides its direct impact on health, malaria places a heavy economic and social burden on endemic countries [7, 8]. Key tools and strategies to fight against malaria include, among others, early diagnosis and treatment with artemisinin-based combination therapy (ACT) and distribution of long-lasting insecticidal nets (LLINs) to populations at risk. While great progress has been registered in the control of malaria in many countries, the burden remains intolerably high in other countries [9]. In Côte d’Ivoire, control efforts by the national malaria control programme are facilitated through continued support from the Global Fund to Fight AIDS, Tuberculosis and Malaria. For example, eight million LLINs were distributed in 2011 and further scaling-up of free LLIN distribution (12 million) to the entire population was planned for the last quarter of 2014.

Here, results are presented from the first national, cross-sectional school-based survey pertaining to parasitic diseases in Côte d’Ivoire, placing particular emphasis on Plasmodium infections. The study was carried out in late 2011/early 2012 and involved more than 5,000 children aged five to 16 years [10] and thus provides an up-to-date situation of the extent of Plasmodium infection in the school-aged population, associated morbidity, preventive and curative measures and physical access to health systems. The information will be useful for the design of equity-oriented malaria control interventions in Côte d’Ivoire.

This is the first national, cross-sectional, school-based survey conducted in Côte d’Ivoire that investigated Plasmodium infection patterns, malaria morbidities and people’s preventive and curative measures. The study was carried out over a four-month period in the dry season towards the end of 2011 and early 2012. This timing allowed the minimisation of logistic and operational challenges, such as high rate of school absenteeism and inaccessibility to the most isolated localities during the rainy season. Plasmodium falciparum was the predominant species with a very high overall prevalence (73.9%), despite efforts in place by the national malaria control programme. A decade of socio-political crisis, which further deteriorated an already weak health system, is likely to have played a major role in this finding [22].

In the current study, microscopy revealed a slightly higher prevalence than RDT (69.1 versus 66.9%) suggesting a small proportion of false negative results revealed by RDT. This result is somewhat surprising, as RDT is a device that detects malaria parasite antigen in a small amount of blood with monoclonal antibodies impregnated on a test strip [23]. The RDT used in the present study is based on histidine-rich protein 2 (HRP-2). Although this antigen persists in the patient’s blood for weeks after successful antimalarial treatment, it has been suggested to be more sensitive in detecting low-level, fluctuating parasitaemia in chronic malaria [24]. The sensitivity of the RDT employed here is above 95%; however, prior studies have shown that RDT sensitivity declines at parasitaemia levels below 500 parasites/μl of blood to 83% [25]. In this study, over 60% of participants had parasitaemia <500 parasites/μl of blood, which might explain the slightly lower sensitivity of the RDT compared to microscopy. It should, however, be noted that a good agreement was found between microscopy and RDT.

Significant differences were found in P. falciparum prevalence between males and females, socioeconomic groups (less wealthy versus wealthier households) and place of residence (rural versus urban settings). Hence, findings reported here are in line with previous observations made elsewhere in Africa [17, 26‐29] and confirm that malaria is a poverty-related disease [7] and that urbanisation is negatively associated with malaria transmission, morbidity and mortality [30]. No significant association was found between age and P. falciparum infection status, which is in contrast to other studies [28]. However, children aged 11–16 years showed significantly lower levels of parasitaemia than their younger counterparts. In high endemicity areas, the early exposition to mosquito bites builds up a partial immunity, which in turn results in lower levels of P. falciparum infection and parasitaemia with age [31].

With regard to prevention and treatment against malaria, results reported here confirm previous investigations in Côte d’Ivoire and elsewhere, as the use of preventive measures and availability of anti-malarial drugs was associated with the ability to afford the related costs [32‐34]. The low bed net use among schoolchildren is consistent with previous findings that people avoid to sleep under a net because of perceived discomfort as highlighted in a recent study from central Côte d’Ivoire [34]. It may also be explained by systematic non-compliance by certain household members [35]. In Côte d’Ivoire, such patterns are likely to occur given past and contemporary malaria control strategies that primarily targeted children below the age of five years and women of childbearing age. These results thus confirm the need for scaling up interventions to other population groups, including school-aged children [34]. The National Malaria Control Programme in Côte d’Ivoire currently places high priority on the distribution of LLINs to cover the entire at-risk population, facilitated by additional funds from the Global Fund to Fight AIDS, Tuberculosis and Malaria. These findings emphasise an existing gap between net ownership and usage, and thus call for additional studies to deepen the current understanding of LLINs among the population, so that control programmes can further improve community effectiveness using bed nets as a major malaria control strategy [36].

Regarding self-reported and clinical morbidity, children from less wealthy households reported significantly more often morbidities such as headache and abdominal pain, and were more frequently diagnosed with anaemia. However, the significant difference in self-reported headache and abdominal pain could be due to the large sample size of the study and might be clinically less relevant. As in previous studies [37, 38], positive associations between P. falciparum infection and parasitaemia with self-reported vomiting was found. However, it is important to highlight that all reported morbidities observed in the current study would not necessarily result from P. falciparum infections since the self-reported data used were among all children and not only among children with P. falciparum infection. Furthermore, those who reported to have suffered from malaria two weeks before the current survey were less likely to have a P. falciparum infection or high parasitaemia suggesting that children likely followed efficacious antimalarial treatment [39, 40]. Related to clinical morbidity, children with anaemia or fever were at higher odds of P. falciparum infection or high levels of parasitaemia than non-anaemic or non-feverish children. These findings are again consistent with previous studies that established a link between malaria parasite density and fever, and hence, parasite density might confirm a malaria case in the face of fever [41, 42]. Nonetheless, in patients with a negative test result who present with fever, malaria cannot automatically be ruled out, since parasitaemia can fluctuate and remain undetectable for a certain time [42]. With regard to anaemia, it is well established that the Plasmodium erythrocyte stages lead to erythrocyte death, which can result in anaemia [43, 44]. Nonetheless, the aetiology of anaemia is multifactorial and other factors including bioavailability of iron in food and other nutritional deficiencies, other parasitic infections (e.g. hookworm), chronic inflammation and genetic conditions must be considered [45‐48].

Another interesting finding is that two out of three children surveyed went to a school where the closest health facility was within a 1-km radius. This means that once at school, most of these children can have access to health care. Surprisingly, children attending schools with health facilities in close proximity were more likely to be infected with P. falciparum and had higher levels of parasitaemia than those going to schools where health facilities were further away. The exact reasons for these observations remain to be investigated. It might be speculated that the presence of health facilities in close proximity to schools influences health-seeking behaviour; children who go to schools that are far away from health facilities might stay at home when they are sick, while children living in villages with health facilities might seek care when sick. Geostatistical modelling for prediction of malaria risk within a Bayesian framework [49‐51] that looks at which risk factors contribute to the spatial distribution of malaria risk might elucidate key reasons behind this observation.

The study has several limitations. First, the results reported here were obtained during a larger investigation focusing on co-infection patterns of Plasmodium and intestinal helminths [10]. Second, considering logistical and financial constraints, the diagnosis of Plasmodium infection was based on a single finger-prick blood sample per child. Multiple blood samples might have revealed higher prevalence rates. Nevertheless, a combination of microscopy and RDT was used to enhance diagnostic sensitivity. It would have been interesting to employ additional diagnostic assays, particularly polymerase chain reaction (PCR) methods that are highly sensitive [52, 53] and thus would allow to clarify the false-negative diagnoses obtained with RDT compared to microscopy. Third, the survey was carried out during the dry season when malaria transmission is low [4, 54, 55], which might have resulted in an underestimation of the overall P. falciparum prevalence. It is conceivable that the overall P. falciparum prevalence among school-aged children is somewhat higher during the rainy season. Fourth, although a recall period of only two weeks was considered in the questionnaire interview, in accordance to previous studies [16], there might still be a recall bias.

This first national school-based cross-sectional survey confirmed that P. falciparum endemicity is high in Côte d’Ivoire. Hence, continued and stronger efforts are still necessary to reduce the intolerable burden of malaria in this West African country. Significant disparities in the prevention and treatment of malaria according to socioeconomic groups are apparent, calling for adapting current control strategies to further enhance equity. Although progress has been registered to increase net coverage in high-risk groups, only a relatively small proportion of children at school age reported to have access to preventive measures, including LLINs, and the actual number of children making regular use of nets is quite low. This calls for concerted efforts to increase access to information and preventive measures in the entire population. Furthermore, improved knowledge on the effect of a list of determinants, including climatic, environmental, socioeconomic and control interventions, on the distribution of P. falciparum infection in schools needs to be generated through rigorous monitoring platforms. Finally, geostatistical modelling and prediction of malaria risk as done in previous studies [17, 56] is needed to spatially target control needs.