The present work is a follow-up longitudinal study spanning over six years on the evolution of malaria prevalence among febrile patients consulting spontaneously at two sentinel sites in Nouakchott. The majority of patients with PCR-confirmed malaria were due to P. vivax, which occurred more frequently in adult males ≥ 20 years old, in agreement with the results reported in earlier studies conducted in Nouakchott [8, 12, 13, 22]. The difference in malaria prevalence between age groups and sex is probably associated with behavioural and occupational differences, but studies on social and behavioural aspects related to malaria risk and knowledge, attitude, and perception survey have not been performed in Mauritania to explain this difference. Although the current proportions of different ethnic groups are not known in the country, white Moors (71.8% vs 22.3% black Moors and 5.9% black Africans) were overrepresented in the patient population of the present study. White Moors were affected more frequently (76.8%) by P. vivax malaria than the other ethnic groups present in the country. This finding confirms that of earlier studies conducted not only in Nouakchott, but also in other sentinel sites in the country [9, 15, 22]. The results are consistent with the hypothesis that Duffy antigen is the major receptor necessary for the entry of P. vivax merozoites into reticulocytes and with the fact that > 50% of white Moors are Duffy-positive and 98% of black Africans are Duffy-negative [27, 28]. Furthermore, the data presented in the present study also suggested that P. vivax can also infect black Moors and black Africans, the majority of whom are known to be Duffy-negative [27, 29]. Further studies would be necessary to update epidemiological data on Duffy antigen in different ethnic groups in Mauritania and evaluate to what extent the presence of Duffy antigen is required for P. vivax malaria to occur.
The update presented in the present study fills an important gap in the epidemiological data on malaria in Mauritania and highlights the current malaria situation in the capital city of Mauritania. For centuries, Nouakchott had been a small fishing village with no particular history attached to it. Until the accession of Mauritania to independence in 1960, there were at most 500 permanent residents in Nouakchott (more precisely, in “Ksar,” a historical district of Nouakchott) [30]. Nouakchott was chosen as the capital of a new nation in 1960. Since then, the process of urbanization in Nouakchott has been intense and unbridled, to the point of becoming by far the largest city, not only in Mauritania, but also in the Sahara Desert today [31, 32]. The rapid pace of urbanization in Nouakchott was largely fueled by a massive rural exodus following the major long-lasting drought in the Sahel region in southern Mauritania during the 1970s and 1980s, resulting in an overcrowded city and leading to multiple challenges in urban organization. One of the most pressing problems that confronted the newly independent nation since 1960 concerned both scarcity and quality of safe potable water supply for the rapidly increasing urban populations. To palliate this problem, several public potable water fountains have been established to supply clean water that was pumped from the Idini aquifer’s well field, 60 km east of Nouakchott to newly settled populations residing in the slums surrounding the city centre. However, inadequate management of these water points have created stagnant water around them, resulting in the creation of breeding sites for mosquitoes, particularly Anopheles gambiae sensu lato (s.l.), but also Aedes aegypti [18, 33]. To bring about a more permanent solution to the problem of chronic water shortage in the capital city with an ever-growing population, in 2010, the Mauritanian government launched the Aftout Essahli project. While the project is at present in an advanced stage of implementation in Nouakchott and undoubtedly represents a major achievement for many Nouakchott residents, on the negative side of the ledger, the higher pressure and the age of the network have also increased the number of leaks and creation of additional mosquito larval habitats [18, 34].
During the last two decades, several studies had been conducted to assess the epidemiology of malaria in Nouakchott, mostly during the period between 2007 and 2013, which coincides with the commencement of the Aftout Essahli project. Some reported a high prevalence of malaria among febrile patients in health facilities in Nouakchott [8, 13, 22]. However, there has been no follow-up study since 2013 to produce an accurate and comprehensive estimate of malaria prevalence in the city, particularly after the completion of the Aftout Essahli project. The results of the present study address and fill this data gap, at least in part.
The longitudinal data from 2015 to 2020 suggest several key epidemiological features: (i) after the occurrence of a high prevalence of P. vivax infections to “epidemic proportions” in 2012–2013 among febrile patients, there is a clear trend towards a declining prevalence due to malaria (both P. vivax and P. falciparum) from 2015 onwards, even though the decline is not uniform year to year; (ii) P. vivax still remains predominant in Nouakchott; (iii) both P. falciparum and P. vivax malaria occur during a relatively short, well-defined period, i.e. at the end of, and just after the rainy season (September–December), but only P. vivax continues to be present throughout the rest of the year during the dry season, with the exception of very few cases of P. falciparum infection; and (iv) surprisingly, despite the absence of an entomological proof at present, the data presented in this study seem to suggest a local transmission of P. falciparum.
The two sentinel sites where many of the previous works on malaria had been conducted draw patients from the districts (Teyarett and Dar naim districts) where Anopheles mosquitoes abound and P. vivax epidemics were notified and demonstrated in 2012–2013 [7, 35, 36]. In an earlier study conducted in Nouakchott in 2009–2010, it was reported that 34.9% (105 of 301) of samples from febrile children included in three health centres, including Teyarett health centre, were PCR-positive (97.1% P. vivax and 2.9% P. falciparum) [22]. During the 2012–2013 malaria epidemics, almost 5,000 cases, representing 54% of clinically suspected malaria cases, were diagnosed as P. vivax malaria by RDT in Teyarett health centre alone, mostly during the months of September or October to December [16]. In that study, only P. vivax was detected, despite the use of government-approved PfHRP2/pLDH-based RDT and confirmation by PCR.
In another study conducted during the same period (2015–2016) as the present study in Atar, a city located in the Saharan zone to the north of Nouakchott, the performance of the RDT was assessed in comparison to nested PCR [9]. Malaria epidemiology in Atar is similar to that of Nouakchott, with a clear predominance of P. vivax over P. falciparum. The sensitivity and the specificity of the RDT to detect P. vivax were 63% and 99%, respectively. Despite the relatively low sensitivity to detect P. vivax using RDT, most probably due to generally low parasitaemia encountered in Mauritanian patients, data interpretation in the present work was based on the results of PCR to offset this technical limitation.
Based on the assumption that similar types of patients were screened and included in the present study from the same sentinel sites, it can be observed that the number of symptomatic patients with a clinical suspicion of malaria declined considerably in 2015–2020 (range, 103 to 459 patients per year; see Table 2 for details) despite a continuous, year-long monitoring for 6 years. More importantly, the mean proportion of PCR-confirmed malaria declined from 34.9% in 2009–2010 to 16.5% in 2015–2020. Several plausible explanations can be given.
First, as shown in the analysis of malaria prevalence and the amount of annual rainfall, a climate change characterized by a relatively reduced rainfall during the period 2015–2020 is highly correlated with decreasing malaria prevalence among febrile patients. By contrast, the unusually high amount of precipitation recorded in Nouakchott in 2009, 2010, and 2013 (130–195 mm, compared to the average annual precipitation of 79 mm in 2015–2020) probably contributed to increase malaria transmission in the early 2010s. In addition, floods occurred in the city in 2010 and 2013 [37]. Because of the absence of a water evacuation system, the city remained flooded for several weeks, or even months in some districts, during those years. The 2012–2013 P. vivax epidemics in Nouakchott can at least in part be ascribed to the large-scale flooding of the city just before the malaria transmission season [16]. Unusual variations and decrease in the annual rainfall have been recorded in Nouakchott in 2015–2020. The climate change, as well as the rising water table resulting in salty water pools seeping through the soil in areas below sea level, had a direct influence on the stability and productivity of anopheline larval habitats in Nouakchott, favouring the creation of highly saline water bodies [37]. Although some An. arabiensis populations have adapted to saline habitats, overall, the increased salinity resulted in decreased anopheline mosquitoes.
Second, a new water distribution system was installed in the capital city during the 2010s. In the early 2010s, public standpipes and water distribution using water tanks drawn by donkeys (a common scene in Teyarett-Dar naim districts) have been directly implicated in the creation of artificial mosquito larval habitats in these districts [18]. Despite initial technical problems leading to frequent leakage few years after the launching of the Aftout Essahli project in 2010, the completion of the potable water supply project in Nouakchott resulted in a considerable reduction in the number of public standpipes in Teyarett, Dar naim, and Ksar districts from 72 in 2013 (9 in Ksar, 23 in Teyrett, and 40 in Dar naim) to only 29 (i.e. 59.1% decrease) in 2020 (1 in Ksar, 10 in Teyarett and 18 in Ksar). During the same period, there was an overall increase in the number of households connected to the SNDE water supply network in Nouakchott, from 53,423 in 2013 to 135,723 in 2020, corresponding to an increase of 60.6%. The expansion and improvement of water distribution network, possibly along with increasing environmental pollution, have been associated with changing distribution of mosquito species. An epidemiological study conducted at that time showed that about 39–44% of indoor resting mosquitoes captured by pyrethrum spray catch in the morning or captured by light traps at night in Teyarett district in 2011–2013 belonged to Anopheles spp. (the rest of the specimens were Culex spp. or Aedes spp.) [14]. By contrast, in an entomological survey conducted at several sites near Teyarett and Dar naim districts in Nouakchott in 2016–2017, only 2.5% (41/1600) of indoor resting mosquitoes were An. arabiensis (the others included Aedes aegypti [26%], Anopheles multicolor [0.7%], and Culex spp. [70.7%]) [37]. The available entomological data tend to support the observation that the proportion of An. arabiensis among captured mosquito specimens in Nouakchott seems to have decreased during the 2010s.
Third, the Mauritanian Ministry of Agriculture and Ministry of Health had carried out a mass campaign of outdoor residual spraying of insecticides in different districts of Nouakchott in 2015 and 2016 in response to the mounting levels of complaints from residents due to nuisance caused by mosquitoes and houseflies [38]. Although numerous anecdotal accounts of the residents have suggested that spraying has resulted in a drastic diminution of mosquitoes and houseflies, unfortunately, rigorous scientific studies designed to measure the impact of spraying were not performed.
Fourth, because well-trained microscopists are rare in the country, the use of RDT has been resorted to massively since its introduction to the country in the early 2010s. The supply of RDT and ACT is subsidized through the Global Fund, and RDT-based diagnosis, as well as ACT for RDT-positive patients, is supposed to be provided free of charge to patients consulting public health centres [39]. The real impact of Global Fund programme for Mauritania has not been evaluated, precluding the possibility of drawing any conclusion with regards to the improvement of diagnosis and treatment as one of the possible causes of declining malaria prevalence in Nouakchott.
Fifth, another factor to take into consideration is the government initiative for a mass distribution of insecticide-impregnated bed nets and increased awareness of the residents through health education, possibly leading to an increased use of bed nets to protect pregnant women and young children and thereby reduce malaria prevalence. This initiative is also subsidized by the Global Fund, but its impact has not been assessed. In reality, however, bed net distribution in Nouakchott has been limited to pregnant women during a limited period of campaign period. This observation is indirectly supported by the results of the present study in which many of the participants (426/582, 73.2%), who did not include pregnant women, reported that they do not use or even own a bed net. Despite the emergence of malaria in Nouakchott and subsequent increase in prevalence during the 2000s and early 2010s [40], the Mauritanian NMCP has decided to distribute bed nets in priority areas situated in southern Mauritania, outside Nouakchott. However, even in the Sahelian zone with high prevalence of P. falciparum, close to one-third of febrile patients reported that they did not own or use bed nets in their household, and the reported use of bed nets did not reduce the probability of being infected by malaria [10]. The limited data available indicate that the coverage of bed net distribution has not reached an adequate level to have any impact on malaria prevalence in the country.
Sixth, another possible cause of the declining proportion of febrile illness attributable to laboratory-confirmed malaria is the emergence of other febrile illnesses that may mimic the clinical presentation of uncomplicated malaria in Nouakchott. For example, dengue fever was unknown in Nouakchott before 2014. In 2014 the presence of mosquito vector of dengue virus, Aedes aegypti, was identified for the first time in Nouakchott [33]. In the same year, the first known dengue epidemic broke out in Nouakchott [41]. According to anecdotal evidence, some Nouakchott adult residents have had febrile illness with a negative result for RDT for malaria but positive RDT for dengue. Furthermore, since 2014, dengue epidemics have been reported almost annually from different parts of the country. Another example is coronavirus disease 2019 (COVID-19) pandemic due to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The country has been affected by COVID-19 epidemic from March 2020 [42]. According to the data presented in Table 2, blood samples collected in 2020 had the lowest prevalence (2.1%) of PCR-confirmed malaria during the study period (2015–2020). Unfortunately, malaria RDT-negative febrile patients were not followed and medical information of the patients included in the present study on possible SARS-CoV-2 viral infection is not available since the latter was not covered by the study protocol and lied outside the scope of the ethical clearance and informed consent.
The reduction of malaria prevalence in Nouakchott in 2015–2020 can be ascribed mostly to the complex interaction of ecological, environmental, and meteorological modifications, with possibly an unknown impact due to the outdoor residual spraying in 2015–2016, more than due to specific malaria interventions based on prevention, diagnosis, and treatment. As observed in earlier studies conducted in Nouakchott [13, 14, 22], P. vivax is still the predominant malaria species in the capital city with no obvious change in the tendency of P. vivax-to-P. falciparum ratio (roughly 3:1 or 75% P. vivax). In Nouakchott, until proven otherwise, malaria parasites are transmitted exclusively by An. arabiensis, which can transmit both P. vivax and P. falciparum, as demonstrated in other African countries where P. vivax and P. falciparum are sympatric [43,44,45]. It is therefore not easy to understand why P. falciparum has been responsible for only a minority of malaria infections in Nouakchott until present. Elsewhere in southern Sahelian zone of the country (e.g., Kobeni, Kiffa, Rosso), where An. arabiensis is also the major malaria vector, P. falciparum predominates [8, 10, 11, 14, 46, 47].
Earlier works performed in Nouakchott in 2009–2013 have reported either total absence or few cases of P. falciparum in the capital city [8, 13, 14, 16, 22]. Moreover, the peak transmission period for P. falciparum was consistently 1 month earlier (i.e. in October) than that of P. vivax (peak in November) during the six-year study period. This time offset corresponds to the rainy season in the Sahelian south which begins earlier (June or July) than in the Saharan zone where Nouakchott is situated (August or September) [10]. This observation is also in agreement with (i) the results of the present study which showed that the majority of P. falciparum-infected patients (69%, including P. falciparum-P. vivax mixed infections) reported travel to regions in southern Mauritania where P. falciparum (also P. vivax to a lesser extent) is endemic less than 1 month before consultation and (ii) also with the hypothesis that P. falciparum malaria is allochthonous in Nouakchott. This unproven hypothesis is plausible due to the regular travels of local populations to and from the capital city to localities in southern region, both for personal reasons related to family ties and business. The road transportation system that links the capital city to various cities and towns in P. falciparum endemic zone is not dense and extensive, but it is efficient for relatively rapid transportation. However, on the contrary, the data in the present study also suggest the occurrence of a local transmission of P. falciparum in Nouakchott. Close to one-third of patients with PCR-confirmed P. falciparum, alone or mixed with P. vivax, denied having travelled outside Nouakchott. Because P. falciparum is undoubtedly present in Nouakchott and few cases may be mixed P. falciparum-P. vivax infections, which may possibly be misdiagnosed by RDT or microscopy, an accurate diagnosis of malaria species and appropriate treatment are required even though, from the medical and ethical viewpoint, a pragmatic approach to case management may be to consider all laboratory-confirmed malaria cases as potentially fatal falciparum malaria. Based on the available evidence, it can be hypothesized that P. falciparum malaria diagnosed in Nouakchott has been cases of imported malaria from the Sahelian southern regions of the country and that this malaria species is still in the process of becoming endemic in the capital city. Further studies will be required to prove this hypothesis, in particular by detecting P. falciparum in the salivary glands of An. arabiensis captured in Nouakchott. In addition to entomological surveys, a long-term regular parasitological surveillance would be necessary to assess malaria burden due to P. falciparum in Nouakchott.
The surveillance system to gather epidemiological data, as described in this work, has several limitations. The evaluation of malaria burden was limited to symptomatic patients presenting spontaneously at the sentinel sites in Nouakchott. Because patient recruitment was performed in an out-patient setting, a possible occurrence of severe and complicated P. vivax malaria was not evaluated in the present study. In addition, in a passive surveillance the malaria situation in the general population in Nouakchott, in particular asymptomatic relapse or carriage of malaria parasites, cannot be evaluated. The patient’s clinical record in health centres and hospitals is not computerized, and the patients do not possess any written clinical record of past illnesses. A patient’s past history on malaria infections and information on possible malaria reinfections or relapse during the six-year surveillance period cannot be assessed. Moreover, data between March and July 2020 are incomplete due to the temporary suspension of patient recruitment during the national lockdown decreed by the government to limit the propagation of SARS-CoV-2 virus responsible for COVID-19 pandemic. However, since the period of lockdown lies outside the malaria transmission period, the missing data in 2020 would likely have a limited impact on the analysis of data in the present work. On the contrary, the strengths of the present study lie on a longitudinal surveillance in the same sentinel sites as in the past studies. The inclusion criteria of the patients were the same as those of the past studies. Laboratory diagnosis was confirmed by PCR. Despite the inherent weakness of a passive disease surveillance, the present study provides various insights into the dynamics of malaria transmission in Nouakchott in recent years.