Complexity of Vector Control and Entomological Surveillance in Endemic Sentinel Sites of the National Malaria Control Program (NMCP) in the Democratic Republic of Congo (DRC)

Abstract

In order to represent the different epidemiological facies that abound in the Democratic Republic of Congo (DRC), new sentinel sites were created. Before their operationalization, baseline evaluations of the bionomics and the insecticide resistance status of malaria vectors were conducted. Using Human Landing Catches (HLCs) and Pyrethrum Spray Catches (PSCs), sampled Anopheles gambiae s.l. were screened for the presence of Plasmodium falciparum. Larval surveys were organized to assess the sensitivity of wild An. gambiae to selected insecticides. Surveys on the community use of Insecticide-Treated Nets (ITNs), Surveys on the community use of Insecticide-Treated Nets (ITNs), were conducted. A total of 2238 Anopheles were collected. Including, 1802 (80.5%) by HLC and 436 (19.5%) by PSC. The majority of the samples were An. gambiae (98%) with very high average transmission entomological indices (density, Human Biting Rates (HBRs) and Entomological Inoculation Rates (EIRs)). These An. gambiae were resistant to selected insecticides at all sites. Households close to breeding sites were at high risk. Overall, ITN coverage was low (41.7%). Of these three sites, only Mweka presented a good coverage of 90%. Only Mweka presented a good coverage of 90%. The sentinel sites are located in the same epidemiological facies where the conditions for transmission of the disease and the incidence are identical. This transmission is ensured by An. gambiae with high resistance statuses vis-à-vis pyrethroids. The ecological choice is necessary for a good representation.

1. Introduction

Malaria is a parasitosis that constitutes a global public health problem and is still endemic in many parts of the world with an estimated 247 million cases in 2021 [1]. It is particularly rife in Sub-Saharan Africa where 95% (234 million) of the global malaria cases and subsequently 96% (593,000) of all deaths occur. In this region of Africa, four countries (Nigeria 27%, the Democratic Republic of Congo (DRC) 12%, Uganda 5% and Mozambique 4%) pay a particularly heavy toll from malaria [1]. Nigeria and the DRC alone carry 39% of the global burden of malaria [1].

In the DRC, malaria remains a major public health problem, despite significant progress made [2, 3]. In 2021, the number of deaths associated with malaria was estimated at 22,729, of which 15,297 occurred of children under the age of 5, i.e., 67% [4, 5]. Around 97% of the population lives in areas with stable, endemic malaria transmission characterized by the equatorial and tropical facies [4]. This country has five distinct climatic zones (Equatorial zone, South tropical wet zone, South tropical dry zone, North tropical wet zone and Temperate tropical zone) [6]. The equatorial climate occupies the central basin and the tropical climate extends from north to south, thus occupying most of the country. The DRC stretches over an area of about 2,345,000 km², and it alone presents almost all the ecologies and bioclimatic strata encountered in Africa, from the Sahelian savannah regions to the equatorial forest regions [4]. This environmental heterogeneity makes combating malaria in the DRC particularly complex [6, 7]. Moreover, despite several rounds of mass distribution of ITNs, the country is struggling to achieve 60% ITN coverage [7].

This DRC has adhered to the World Health Organization’s (WHO’s) global technical strategy for the fight against malaria (2016–2030), which provides comprehensive technical guidance with the aim of reducing the incidence and mortality rates linked to malaria by at least 90% compared to 2015 reference year [1].

Large-scale use of ITNs offers good protection to populations at risk of malaria and has been the primary tool for vector control for nearly 40 years [8]. Taking this global strategy into account, for almost a decade the DRC has opted for building universal coverage of malaria control interventions within the National Strategic Plan (NSP) [4]. The current NSP adopted at the national level is aligned with the global technical strategy for the fight against malaria, with the objectives of reducing the rate of morbidity and mortality linked to malaria by at least 40% by the end of 2023 with coverage ≥80% in ITNs [4].

One of the guiding principles the NSP has set itself is the implementation of vector control based on evidence and convincing results [4]. To do this, the National Malaria Control Program (NMCP) relies on data from routine systematic surveys and studies [4, 7, 9] requiring the establishment of sentinel sites. Together, these sentinel sites generate inexpensive, complete and quality data [4]. They are distributed according to the health subdivision, Health Division (HD), Health Zone (HZ) and Health Area (HA), which is similar to the administrative subdivision (province, territory, community, village) seen elsewhere.

Parasitological, pharmacological and environmental data are needed to assess the evolution of malaria [4] and this guided how sentinel surveillance sites were created and located across the country on the basis of the Health Division (HD). Sentinel sites correspond to health surveillance zones where parasitological and entomological surveys are carried out to provide basic data for evaluating malaria control activities, and which represent the Province Health Division (PHD).

Before decentralization in 2010, the DRC had 11 provinces giving 11 sentinel sites. After decentralization, 26 provinces were obtained. Thus, the DRC will gradually add 15 new sites corresponding to one site in each of the 26 Provincial Health Divisions (PHDs) [10]. These sentinel sites represent the different epidemiological facies of malaria transmission found in the DRC.

It is important to note that the distribution of each Anopheles species relies on ecological characteristics favorable to their reproduction and multiplication [11, 12]. In the DRC, the most encountered vectors are Anopheles gambiae, Anopheles funestus, Anopheles nili, Anopheles moucheti and Anopheles paludis [4]. These species are not evenly distributed in all PHDs.

Surveys examining vector bionomics (behaviors including host preference, host-seeking, peak activity period as well as larval site characteristics) and insecticide resistance (IR) monitoring are carried out in the various sentinel sites for evidence that helps the NMCP choose the most appropriate vector control tools. The emergence of IR in the African region raises many questions about the ongoing effectiveness of ITNs in helping to reduce the incidence of malaria [8]. Increasing evidence of Anopheles resistance to different classes of insecticides is reported by several authors [9, 13, 14]. Thus, it is necessary that the choice of sentinel sites take into account the ecology and the bionomics of the vectors, as well as the profiles of the Anopheles vis-à-vis the insecticides used.

The current choice of NMCP sentinel sites is based on administrative criteria alone and does not address the biodiversity of vector species and various epidemic facies. Indeed, the work of Rahm and Vermilin [15] on the distribution of anopheline species in all Congolese territories identified 66 species of Anopheles, whereas Watsenga et al.’s [9] work, based on sentinel sites in the DRC, identified fewer species. Yet, there is evidence elsewhere (Cameroon) that a greater diversity of species can be identified when the choice of sentinel sites is based on ecological criteria [16].

To enable the new sentinel sites to fully play their role by providing reliable data and reflecting the local reality of malaria transmission, this pilot study was carried out to assess entomological surveillance in the nine additional sentinel sites. Only four sites, namely Boende in the Province of Tshuapa, Lisala in the Province of Mongala, Mweka in Kasai and Nyakunde in Ituri, were selected.

The present study aims to contribute to our understanding of the local epidemiology of malaria by identifying vector species, determining entomological parameters (species, density, HBR and EIR) and the profile of An. gambiae s.l.’s susceptibility to the usual insecticides and the factors modulating the transmission of malaria in the sentinel sites.

2. Complexity of vector control and entomological surveillance in endemic sentinel sites of the NMCP in DRC

2.1 Methods

2.1.1 Study sites

The study was conducted in four new sentinel sites, from June 17 to July 17, 2019 for the first three sites (Boende, Lisala and Nyakunde) and from October 15 to November 15, 2019 for the Mweka site (Figure 1) by three teams of entomologists from Laboratory of Bio-ecology and Vector Control at the Kinshasa School of Public Health (BIOLAV-KSPH). Location of entomological sentinel sites and frequency of sensitivity testing and trapping to determine human bite rates are given in Table 1.

Figure 1.

Sentinel mosquito collection sites (in green, the provinces and in red the capture point).

Sentinel site	Province	1GPS coordinates	Selected 2HA	Frequency and period of data collection	Climatic zone
Bende	Tshuapa	0°16′47″S	Motema mosantu	4 times during a week (June or July 2019)	Equatorial
		20°52′39″E	Kimbangu II	4 times during a week (June or July 2019)
Lisala	Mongala	2°9′0” N	Antenne	4 times during a week (June or July 2019)	Equatorial
		21°31′0″ E	Fisherman	4 times during a week (June or July 2019)
Nyakunde	Ituri	1°25′60” N	Nyakunde	4 times during a week (June or July 2019)	Equatorial
		30°1′60″ E	Birinyama	4 times during a week (June or July 2019)
Mweka	Kasaï	4°50′60”S	Mweka 1	4 times during a week (October or November 2019)	3S Too w
		2°34′0″ E	Bulongo	4 times during a week (October or November 2019)

Table 1.

Entomological sentinel site location, period of data collection and climatic zone.

¹

GPS, Global Positioning System.

²

HA, Health Area.

³

S Too w, South Tropical Wet.

2.1.1.1 Boende

Boende is located in the Province of Tshuapa, on the Tshuapa and Lomami rivers, east of Mbandaka, 1400 km north-east of the capital Kinshasa. It is served by Route Nationale 8.

2.1.1.2 Lisala

Lisala is located in Mongala province, on the right bank of the Congo River, in a forested area with an equatorial climate. It lies 2650 km north-east of Kinshasa and is served by the Route Nationale 6.

2.1.1.3 Nyakunde

Nyakunde is the seat of the Andisoma chiefdom of the Bira, located some 30 km south-west of Bunia, in the Irumu territory, Ituri district, Orientale Province in northeastern Democratic Republic of Congo. It comprises three regions with distinct climatic characteristics: a very rainy region like the equatorial basin, an intermediate zone where rainfall decreases during the dry season like the tropical zone, and a territory with little rainfall but alternating between the two seasons.

2.1.1.4 Mweka

Mweka is a decentralized entity in the Province of Kasai in the Democratic Republic of Congo. Located in a savannah zone with a humid tropical climate with two seasons (dry from May to September and rainy September to December), the territory is characterized by a slightly warm temperature with a maximum estimated at 27°C. The climatic elements (temperature and rainfall) for all three sites sampled at different stations (Lisala, Tshuapa and Tshikapa) are shown in Figure 2.

Figure 2.

Ombrothermic curve of Mweka, Lisala and Boende in 2019 (source: Lisala, Tshuapa and Tshikapa weather station).

2.1.2 Study population and sampling technique

The statistical unit of the study was the Anopheles mosquito specimens, which were collected from selected houses using a four-stage probability sampling technique:

• Four sentinel sites corresponding to four Provincial Health Divisions (PHDs) were randomly selected from nine newly created NMCP sentinel sites.

• Two health areas (HAs) were randomly selected from those carrying out sentinel surveillance in the site.

• Two villages were selected among all those constituting the selected health area.

• Ten houses per village were selected, according to the World Health Organization (WHO) recommendation, to have a good representation [17].

However, one of the surveys was underway and it was possible to sample only in three sites (Boende, Lisala and Mweka) out of the four initially planned, i.e., a total of 60 houses instead of the 80. Unfortunately, security threats at the sentinel sites within Nyakunde prevented the surveys from going ahead as intended. Thus, this site was removed from our analyses.

2.1.3 Data collection study procedure

2.1.3.1 Mosquito collection

2.1.3.1.1 Larva collection

Larvae and pupae of An. gambiae s.l. were collected from different larval sites. After sorting, the selected larvae were reared in tanks covered with mosquito nets and then kept under conditions likely to favor the emergence of mosquitoes, without contamination of the rearing by external mosquitoes and without infestation of the environment. The daily monitoring of the farm consisted in counting the numbers of larvae in each tank, on the one hand, and living and dead imagoes in each cage, on the other hand [18].

2.1.3.1.2 Adult mosquitoes (anopheles)

Two collection methods were used: Capture by indoor spraying with pyrethrums (PSCs) allowing the residual density of resting anophelines and their trophic states to be established; and human landing catches (HLCs) to assess the risk of infective/human/night bites, Anopheles behavior and the cycle of aggression (peak biting).

2.1.3.1.3 Human landing catches (HLCs)

Human landing catches (HLCs) were conducted in 10 houses per village from 6 p.m. to 6 a.m., and the data reported cover only 12 h of nocturnal captures. Thus, in each village, 10 houses were selected for capture, giving a total of 20 houses per site and 60 houses for all sites. As mentioned above, the Nyakunde data were not taken into account. Four mosquito collectors were assigned to each house, i.e., two outdoor and two indoor. After sorting the mosquitoes morphologically to identify the anophelines in the field, the mosquitoes were individually preserved in Eppendorf tubes with silica gel and brought back to the BIOLAV/KSPH laboratory for further analysis (identification, determination of infectivity and WHO Tube Pass sensitivity test) [19].

2.1.3.1.4 Capture by indoor pyrethrum spray (PSC)

The PSC was carried out by spraying a pyrethrum-based insecticide in 10 non-HLC houses between 6 and 10 a.m. after removing cooking utensils, drinks and food. White sheets were spread out on the floor in all rooms of the house and spraying began outside the house, in front of doors and windows, then inside. A commercially available pyrethroid spray (Baygon, Bayer) was sprayed in the house and doors were closed for 15 min. After which, doors and windows were opened to collect mosquitoes that had fallen onto the sheets. Female mosquitoes were classified according to four trophic statuses (unfed, freshly fed, half-gravid and gravid) [17, 20].

2.1.4 Household survey

The household surveys were carried out in each of the houses where the mosquitoes were captured. Household heads were interviewed based on the questions administered in the Malaria Indicator Surveys (MIS). The questionnaire included questions on sociodemographic status, the notion of fever in children under 5 years of age, the use or not of mosquito nets and the physical condition of ITNs. The type of construction materials of the houses (blackberries and roofs) was also characterized in order to determine the quality of the habitat allowing the access or not of the mosquitoes [7].

2.1.5 Data and studies procedure

2.1.5.1 Morphological identification of Culicidae

The mosquitoes were sorted based on morphological criteria to gender and sex in the field. They were then stored individually in Eppendorf tubes with silica gel and sent to the entomology laboratory of the Kinshasa School of Public Health (KSPH). Once in the laboratory, morphological identification under the microscope was conducted using dichotomous keys [21, 22].

2.1.5.2 Insecticide susceptibility tests and synergist bioassays

Sensitivity tests were carried out on wild populations of Anopheles in each of the four sentinel sites selected. They made it possible to better know the levels of sensitivity of the vectors to the two Pyrethroids (Alphacypermethrin and Deltamethrin). And in the event of resistance to pyrethroids, a pre-exposure to Piperonyl Butoxide (PBO) was carried out for the search for oxidases. The effectiveness of these compounds was compared with that of DDT and Bendiocarb, to detect the existence or not of cross-resistance between the three chemical families. And this insecticide’s effectiveness was measured according to their knockdown effect (Kdt₅₀ and Kdt₉₅) and the mortality they cause after 24 h of observation.

Larval collections were made in each study site using larval dippers. Larval sampling was done in transient, sun-exposed puddles to maximize the likelihood of sampling An. gambiae s.l.. Larvae were subsequently transported to a field insectary for rearing. Emerging adult mosquitoes were provided with sugar solution until they were used for insecticide susceptibility tests [9].

Adult females 2–5 days old were selected and tested according to the WHO protocol [18]. The tests were carried out with the papers impregnated with insecticides from the WHO Kit composed of four types of insecticides and the PBO synergist of different concentrations: Deltamethrin 0.05%, Alphacypermethrin 0.05%, Bendiocarb 0.1%, DDT 4% and PBO 5%. And the effectiveness of insecticide was measured according to their knockdown effect for 60 min of contact and the mortality they cause after 24 h of observation.

The tests were carried out in accordance with the general conditions of the test procedures recommended by the WHO, particularly with regard to: temperature (between 23 and 27°C), relative humidity (between 70 and 90%), the age of the adult female mosquitoes (2 to 5 days), the number of times an insecticide-treated paper should be used (Not more than 6 times/At most 150 mosquitoes per paper for pyrethroids), the qualities and the number of uses of control test papers [18].

For each insecticide retained, four test replications, each involving 25 mosquitoes, were carried out. The first series of tests was carried out to observe the behavior of An. gambiae s.l. after 60 min of contact with insecticides. The numbers of mosquitoes shocked after 10, 15, 20, 30, 40, 50 and 60 min of exposure or contact with the insecticide were recorded, in order to calculate the shock effect (Kdt₅₀ and Kdt₉₅). Twenty-five mosquitoes were used for the negative control. After exposure, mosquitoes were transferred to clean holding tubes and provided with sugar solution. Mortality was recorded 24 h after exposure.

In the event of resistance to pyrethroids, a second series of tests was carried out with anophelines pre-exposed to PBO for 60 min.

The test results were interpreted according to the WHO criteria [18]:

• Susceptible (S), if mortality is between 98 and 100%;

• Probably Resistant (PR), if mortality is between 90 and 97%.

• Resistant (R), if mortality is less than 90%;

2.1.5.3 Determination of the infectivity rate by the enzyme-linked immunosorbent assay (ELISA)-plasmodium falciparum circumsporozoite protein (ELISA-CSP) pf technique

After morphological identification, a subsample of individually stored An. gambiae s.l. collected by HLC were analyzed at the Institut National de la Recherche Biomédicale (INRB). The An. gambiae s.l. female were ground for analysis in the pf CSP ELISA test according to the Wurtz protocol.

The sporozoite index (SI) was determined after ELISA-CSP to detect the presence of the circumsporozoite protein of P. falciparum. However, after grinding Blocking Buffer (BB-IGEPAL®) using a pestle, the Pf monoclonal antibodies (mAb Pf) were fixed on the microplates, and the ground material was brought into contact with the mAb Pf. In addition, the enzyme-coupled conjugate (mAb capture, peroxidase) was added before adding the enzyme substrate (hydrogen peroxide (H₂O₂)). Then, the reading was made with an ELISA reader at 405 nm.

2.1.5.4 Determination of entomological indices of transmission

2.1.5.4.1 Indoor resting density (IRD)

Density of resting mosquitoes inside houses was estimated by the total number of female mosquitoes resting indoors on the total number of houses inspected.

2.1.5.4.2 Human biting rate (HBR)

This rate designates the number of mosquito bites received by a person during one night (bite/human/night). It was estimated on the basis of the number of Anopheles caught on human bait, from 6 p.m. to 6 a.m.

2.1.5.4.3 Entomological inoculation rate (EIR)

The entomological inoculation rate (EIR) represents the average number of infective bites per human per night (ib/h/n). It is calculated as follows:

$$EIR=HBR\times \text{Sporozoite index}\left(\%\right)/100.$$E1

2.1.6 Data analysis

Data were entered using Microsoft Excel® 2016. Statistical analyses of study data were performed using R® software in version 4.1.2. A statistical analysis was carried out on certain parameters with the usual statistical tests.

The data were essentially summarized in the form of a proportion because it was mainly necessary to calculate the entomological indices. Additionally, data were summarized in tables and graphs using Microsoft Excel® 2016 and R® 4.1.2. As the Shapiro test and the quantile-quantile (QQ) graph showed that the distribution of our data was not normal, it was more the nonparametric tests that were used in the rest of the analyses. The results obtained in the six health areas during the different surveys were compared using Kruskal-Wallis tests on the one hand, for the comparison of the medians and Fisher exact to compare the proportions. To determine factors associated with anopheline density, a nonparametric Theil-Sen regression model was used. Before performing the regression, exploratory univariate analyses were performed to identify independent variables significantly associated with anopheline density. A two-sided alpha level of 0.05 was considered statistically significant.

2.2 Results

2.2.1 Mosquito species composition and abundance

A total of 6697 mosquitoes were collected by HLC and PSC. Culex were most abundant with 4459 (66.58%), compared to 2238 Anopheles (33.41%). Of the 2238 Anopheles collected, 1802 were collected by HLC and 436 by PSC.

Of the 436 Anopheles collected, An. gambiae s.l. was predominant with 430 specimens (98%). An. funestus and Anopheles ziemani were poorly represented with, respectively, five and one specimens (1.15 and 0.23%) (Figure 3). Mweka presented a higher abundance of 156 Anopheles specimens (35.78%) than the other two sites.

Figure 3.

Distribution of anopheles collected by PSC and HLC according to three sentinel sites.

In view of the observations made by HLC, it emerges from Figure 4 that 1802 mosquitoes were captured in the three sites. Indeed, An. gambiae s.l. showed a remarkable abundance of 1796 specimens (or 99.67%). The remaining two species (An. funestus and An. ziemani) were poorly represented with five and one specimens (i.e., 0.28 and 0.06%) (Figure 3). The Mweka site was the most abundant site inAnopheles, with 664 specimens (36.85%).

Figure 4.

Biting time of an. Gambiae s.l. collected by HLC according to sentinel sites. (a) Lisala (As pêcheur), (b) Boende (As motema-mosantu), (c) Mweka (As Mweka), (d) Lisala (As antenne), (e) Boende (As Boende II), (f) Mweka (As Mweka). Out = outdoor; In = indoor.

2.2.2 Determination of entomological parameters

2.2.2.1 PSC collection

With regard to the results presented in Table 2, the Resting Density Index (RDI) and the Indirect Human Biting Rate (IHBR) observed vary greatly from one site to another, and even from one village to another. Thus, the average values observed varied from 4.9 to 10 An./house for the RDI and from 0.08 to 1.55An. fed/human/night for the indirect bite.

Sentinel site, Capture district or village (Health area)		RDI* (An./house)	I HBR* (An. fed/human/night)
Boende (Tshuapa)	Hôpital (Motema mosantu)	4.9	0.51
	Buza (Boende II/Kimbangu)	10	1.55
Lisala (Mongala)	Lokele (Pêcheur)	7.5	0.27
	3C (Antenne)	6.1	0.08
Mweka (Kasai)	Congo I (Mweka)	9	0.98
	Village Bulongo (Bulongo)	6.1	0.92

Table 2.

Entomological evaluation indices by PSC in sentinel sites.

*

RDI: Relative density indoor (mean numbers of Anopheles per house). IHBI: Indirect Human Biting Rate (Anopheles fed/human/night).

The highest and lowest RDIs were, respectively, observed in the Buza district (Boende II/Kimbangu) and Hôpital (Motema-Mosantu) from the Boende site (Tshuapa) with 10 and 4.9An. female/house, respectively.

As for indirect Human Biting Rate, the Buza district (Boende II/Kimbangu) presented the highest indirect bite rate next to the 3C district (Antenne), which presented the lowest indirect bite rate with, respectively, 1.55 and 0.08An. fed/human/night.

2.2.2.2 Human landing catch (HLC)

2.2.2.2.1 Biting time

Overall, anopheline biting activity started very early at 6 p.m., initially increasing gradually until 10 p.m. where the numbers captured increased at a higher rate before peaking between 24 a.m. and 2 a.m. The exception was Bulongo village where biting activity peaked notably later at 2 a.m. (indoors) and numbers captured began to decrease around 5 a.m. (Figure 4). The behavior of Anopheles sp. was more endophagic at all sites.

2.2.3 Entomological clues of transmission

It appears from Table 3 that the average of the entomological indices of transmission (density, HBR, SI and EIR) was high and varied from one site to another. High density was recorded at Tshuapa Boende in Boende II health area (10 Anopheles/house), followed by Kasai Mweka in Mweka 1 health area (9 Anopheles/house), and low density was recorded in Tshuapa Boende in the Motema-Mosantu health area (4.9 years/house). The HBR was more recorded indoor (13.1 b/h/n) against 9.3 b/h/n outdoor. However, in Kasai Mweka in the health area of Mweka 1, the high rate of bites was recorded (19.8 b/h/n) and the same EIR (2.73 ib./h/n).

Sites	Health area	Density An./House	*HBR outdoor b/h/n	HBR indoor b/h/n	HBR b/h/n	*SI%	*EIR ib./h/n
Tshuapa Boende	Motema mosantu	4.9	9.3	12.3	10.8	11.5	1.2
	Boende II	10	4.3	6.9	5.6
Mongala Lisala	Pêcheur	7.5	6.4	10.9	8.7	9.8	0.85
	Antenne	6.1	11.9	21.3	16.6
Kasai Mweka	Mweka 1	9	19.3	20.3	19.8	13.8	2.73
	Bulongo	6.1	4.3	6.9	5.6
Mean		7.3	9.3	13.1	11.2	11.7	1.59

Table 3.

Synthesis of entomological indices of transmission.

*

HBR: Human Biting Rate (bite/human/night); SI: sporozoite index; EIR: entomological inoculation rate (infective bites/human/night).

2.2.4 Anopheles susceptibility profile of gambiae s.l. against common insecticides

The results of WHO susceptibility tests carried out on wild populations of An. gambiae s.l. in the four sites (Tshuapa/Boende, Mongala/Lisala, Mweka/Kasaï and Ituri/Nyakunde) to the two pyrethroids (Alphacypermethrin and Deltamethrin), to the organochlorines (DDT) and to the carbamates (Bendiocarb) are recorded in Figure 5.

Figure 5.

Percentage mortality of an. Gambiae s.l. after pre-exposure to PBO followed by permethrin 1X and Alphacypermethrin 1X in WHO tube tests in four sites.

Susceptibility tests in indicated generally reveal that, across all collection sites, An. gambliae s.l. showed resistance (≤90%) of varying degrees to Alphacypermethrin and Deltamethrin. Separately, Deltamethrin mortality was slightly elevated at all sites in contrast to that of Alphacypermethrin. After pre-exposure of An. gambiae s.l. to PBO, the efficacy of Deltamethrin and Alphacypermethrin was fully restored at all sites to 98–100%.

Figure 5 shows that An. gambiae s.l. tested were more resistant to DDT with very low mortality of 11% in Mweka and 19% tied in Boende and Lisala. Only Bendiocarb was 100% effective at all sites.

Table 4 presents the analysis that determined the Kdt₅₀ and Kdt₉₅, the value of these two parameters Kdt (knockdown time) corresponds to the time after which, respectively, 50 and 95% of An. gambiae s.l. were knocked out after 60 min of insecticide contact. Kdt₅₀ and Kdt₉₅ have never been reached in all sites with DDT. The Kdt₅₀ with Deltamethrin and Alphacypermethrin was not reached in the only site of Nyakunde and elsewhere, it was reached late.

Sites	Insecticides	n	Kdt50 (min)	Kdt95 (minute)	Mortality 24 h (%)	Statut*
			IC 95%	IC 95%
Tshuapa-Boende	Deltamethrine 0,05%	100	27,8 (26,2–29,4)	44,7 (41,9–-48,3)	96	PR
	Deltamethrine 0,05% + PBO 5%	100	27,2 (26,0–-28,5)	39,3 (37,3–-42,1)	100	S
	Alphacypermethrine 0,5%	100	28,9 (26,5–-31,2)	49.6(45.6–-53.3)	90	R
	Alphacypermethrine 0,5% + PBO 5%	100	26,1 (25,2–-27,0)	39,7 (38,1–-41,8)	100	S
	DDT 4%	100	n/a	n / a	19	R
Mongala-Lisala	Deltamethrine 0,05%	100	37,3 (29,7–47,46)	n/a	73	R
	Deltamethrine 0,05% + PBO 5%	100	19.1 (16.7–21.1)	32,4 (28,1–40,3)	100	S
	Alphacypermethrine 0,5%	100	42,1 (35,5–51,3)	n/a	69	R
	Alphacypermethrine 0,5% + PBO 5%	100	20,8 (15,8–-27,7)	37,1 (29,5–-59,3)	98	S
	DDT 4%	100	n/a	n/a	19	R
Nyakunde-Ituri	Deltamethrine 0,05%	100	n/a	n/a	51	R
	Alphacypermethrine 0,5%	100	n/a	n/a	42	R
Mweka/Kasaï	Deltamethrine 0,05%	100	47,68 (44,5–51,6)	n/a	56	R
	Deltamethrine 0,05% + PBO 5%	100	23,5 (21,3–25,9)	43,6 (37,7–54,2)	98	S
	A	100	47,68 (44,5–51,6)	50,3 (45,9–56,1)	61	R
	Alphacypermethrine 0.,5% + PBO 5%	100	18,8 (16,9–20,6)	34,5 (29,9–42,6)	100	S
	DDT 4%	100	n/a	n/a	11	R

Table 4.

Knockdown time of an. Gambiae s.l. after 60 min exposure to insecticides and mortality after 24 h.

*

R = resistance (24 h mortality ≤90%); S = sensibility (mortality 24 h 98–100%), RP = Resistance probability (mortality 24 h between 90 and 97%), n/a = not available.

Kdt₉₅ with Deltamethrin and Alphacypermethrin was reached with difficulty and late at the Tshuapa/Boende site, respectively, at 44.7 min 95% confidence interval (CI) (41.9–48.3 min) and 49.6 min 95% CI (45.6–53.3 min). After pre-exposure to PBO, the efficacy of these two pyrethroids was restored and Kdt₅₀ and Kdt₉₅ were reached.

It is observed in Mweka, Kdt₉₅ was not available to Deltamethrin and reached late forAlphacypermethrin by 50.3 (45.9–56.1) min, which reflected a high degree of resistance of An. gambiae s.l. to these two pyrethroids. However, in Nyakunde, Kdt₉₅ was not available to Deltamethrin and Alphacypermethrin.

2.2.5 Infectivity rate

Sporozoite rates for all three sentinel sites are presented of 440 An. gambiae s.l. analyzed, the mean sporozoite rate in all sites being 11.6% (95% CI 3.4–5.2) for An. gambiae s.l. Moreover, sporozoite rates were relatively low at all sites. The lowest sporozoite rate was observed in Lisala at 9.8% (95% CI 1.1–4.6) and the highest at Mweka at 13.1 (95% CI 1.1–4.6).

2.2.6 Determination of factors modulating malaria transmission in sentinel sites

Table 5 shows that in the Antenne health area, out of the 10 houses surveyed, no engorged Anopheles were found, whereas the median value was higher in Motema-Mosantu, Boende 2 and Pêcheur. The difference in the medians in the six health areas is statistically significant (p = 0.038). The distance between houses and larval sites was shorter in the health area of Motema-Mosantu compared to the other health areas (p < 0.001). The quality of mosquito nets shows a significant difference between HAs (p < 0.001), with 0 and 1% for good quality of mosquito nets in Antenne and Mweka 1 HAs. The survey showed a difference in the proportions of metallic houses (p < 0.001), with 0% in AS Pêcheur, Bulongo, Mweka 1 and Boende 2 (Table 5). Possession of ITNs was very low in Lisala and Boende at 15 and 20%, respectively, and in Mweka there was good coverage of 90% with an average possession in the sites of 41.7%.

Variables	A	Pêcheur	Bulongo	Mweka_1	Boende_2	Motema_Mosantu	Ooverall	Test
Anopheles_density (Count)	10	10	10	10	10	10	60	Kruskal –-Wallis (0.353)
Median (IQR)	2.50 (6.25)	3.50 (10.50)	2.00 (3.50)	1.50 (6.50)	4.00 (11.75)	6.50 (8.75)	3.00(8.00)
Inphed (Count)	10	10	10	10	10	10	60	Kruskal –-Wallis (0.038)
Median (IQR)	0.00 (0.00)	3.00 (5.75)	2.00 (0.75)	1.50 (4.00)	4.00 (6.75)	1.50 (2.75)	2.00(3.25)
Household_persons	10	10	10	10	10	10	60	Kruskal -–Wallis (0.047)
Median (IQR)	3.50 (2.00)	4.50 (3.75)	5.50 (3.25)	7.00 (6.25)	5.00 (2.25)	8.00 (5.50)	5.00(4.00)
Dist_house_hablarva	10	10	10	10	10	10	60	Kruskal -–Wallis (<0.001)
Median (IQR)	62.50 (71.50)	35.00 (27.50)	42.50 (17.50)	50.00 (27.50)	50.00 (17.50)	8.00 (1.00)	40.00 (36.25)
ITN_Possession (%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	60	Fisher exact (0.152)
Yes	6 (40.00%) (60.00%)	2 (13.33%) (20.00%)	1 (6.67%) (10.00%)	2 (13.33%) (20.00%)	1 (6.67%) (10.00%)	3 (20.00%) (30.00%)	15 (100.00%) (25.00%)
No	4 (8.89%) (40.00%)	8 (17.78%) (80.00%)	9 (20.00%) (90.00%)	8 (17.78%) (80.00%)	9 (20.00%) (90.00%)	7 (15.56%) (70.00%)	45 (100.00%) (75.00%)
ITN_condition (%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	60	Fisher exact (<0.001)
Bad	10 (28.57%) (100.00%)	7 (20.00%) (70.00%)	1 (2.86%) (10.00%)	9 (25.71%) (90.00%)	1 (2.86%) (10.00%)	7 (20.00%) (70.00%)	35 (100.00%) (58.33%)
Good	0 (0.00%) (0.00%)	3 (12.00%) (30.00%)	9 (36.00%) (90.00%)	1 (4.00%) (10.00%)	9 (36.00%) (90.00%)	3 (12.00%) (30.00%)	25 (100.00%) (41.67%)
Fever (cCount (%))	10 (16.67%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	60	Fisher exact (0.114)
Yes	8 (20.00%) (80.00%)	7 (17.50%) (70.00%)	7 (17.50%) (70.00%)	3 (7.50%) (30.00%)	6 (15.00%) (60.00%)	9 (22.50%) (90.00%)	40 (100.00%) (66.67%)
No	2 (10.00%) (20.00%)	3 (15.00%) (30.00%)	3 (15.00%) (30.00%)	7 (35.00%) (70.00%)	4 (20.00%) (40.00%)	1 (5.00%) (10.00%)	20 (100.00%) (33.33%)
House_type (%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	10 (16.67%)	60	Fisher exact (<0.001)
s	0 (0.00%) (0.00%)	0 (0.00%)(0.00%)	3(42.86%)(30.00%)	0 (0.00%)(0.00%)	4 (57.14%) (40.00%)	0 (0.00%) (0.00%)	7 (100.00%) (11.67%)
Sheet_metal	10 (50.00%) (100.00%)	0 (0.00%) (0.00%)	0 (0.00%) (0.00%)	0(0.00%)(0.00%)	0 (0.00%) (0.00%)	10 (50.00%) (100.00%)	20 (100.00%) (33.33%)
s	0 (0.00%) (0.00%)	10(30.3%)(100.%)	7(21.21%)(70%)	10(30.30%) (100%)	6 (18.18%) (60%)	0 (0.00%) (0.00%)	33 (100.00%) (55.00%)

Table 5.

Factors influencing malaria transmission in sites.

Table 6 shows that the anopheline density, according to our data, is significantly lower for houses located far from breeding larval sites (protective role because the odds ratio is less than 1). The other two variables were not statistically significant. The anopheline density with engorgement, according to our data, is significantly high for houses with a high number of ITNs.

Characteristic	(Intercept)	Coefficient*	Odds ratio	Lower 95% CI limit	Upper 95% CI limit	P-value
Site distance	6.43243243	−0.05882353	0.9428731	0.9013153	0.9535306	0.04073
Number of people	2.00000000	0.08333333	1.086904	0.9512268	1.673181	0.196
House type	1.85624153	0.07423761	1.077063	0.9492531	1.638152	0.1822
ITN number	1.499946	1.000010	2.718282	2.518309	7.389241	0.02652
Child ≤5 years old	1.5000088	0.01287691	1.00296	1.00121	2.954372	0.6716
Fever ≤14 days	1.4759213	0.01174856	1.011818	1.010123	1.638152	0.4186

Table 6.

Determination of factors influencing anapheline density.

*

Coefficients of the linear predictor of the resulting line of the logistic model. The exponential of these coefficients gives the odds ratio, the distance between the household and the breeding site; the number of people in the household; and house type.

The exponential of these coefficients gives the odds ratio, the distance between the household and the breeding site; the number of people in the household; and house type.

2.3 Discussion

2.3.1 Identification of vector species

After the staggered capture over the duration of the study, three species of Anopheles were identified overall but with a remarkably high density of An. gambiae s.l. found at all sites. However, according to Mouchet et al. [23], the different species of Anopheles exploit a wide variety of water collections as roosts, in particular the residual pools of sunny stagnant surfaces, pools with erect vegetation, brackish water, etc. But An. gambiae s.l. preferentially uses residual pools of sunlit stagnant surfaces in several regions of Africa [24]. It is important to point out in the DRC with the advent of decentralization which increased the number of provinces from 11 to 26, a lot of modernization and urbanization work has been carried out in the new provinces to accommodate the provincial institutions.

Overall, correct urbanization reduces the risk of malaria transmission [1]. However, this urbanization has been anarchic in many provinces and has favored focal transmission of malaria, which results in a high disease burden [1, 4]. This anarchic urbanization is exacerbated by human activities which create temporary and artificial shelters favorable to the development of An. gambaie sl throughout the year [11, 12].

All urban areas of the DRC have almost the same ecology (large sunny space and the larval sites created by human activities). This ecology is favorable for the development of An. gambiae s.l. These results would reflect a low diversity observed. On the other hand, these results do not corroborate those of Rahm and Vermylen [15] and Sinka et al. [25], who found 61 species at the ecological scale of all the territories of the DRC.

2.3.2 Determination of entomological parameters

In general, the entomological indices of transmission are very high in the three capture sites. Despite high anopheline densities observed in Boende (Boende 2 health area) at 10 An./house and in Mweka (Mweka health area) at 9 An./house, compared to the Lisala site, no significant difference was observed. These highly observed densities can be justified by the poor quality of the habitat observed in these sites [26]. In addition, the absence of ITNs and the presence of larval sites also positively influenced anopheline density.

And we observed a high rate of Anopheles engorgement in the different households. The Boende site recorded more gorged Anopheles due to the absence of mosquito nets and a lot of shelter. There is a significant difference between distance from houses and larval sites. The houses of the Antenne site were very distant from the larval sites. Anopheline density was lower in houses located far from larval sites (protective effect).

The anopheline density of gorged Anopheles mosquitoes was twice as high in houses with more impregnated mosquito nets (poor quality of mosquito nets found).

The contraction on the effectiveness of ITNs is justified by the fact that ITNs were dilapidated and also the observed low durability [27]. In the Boende and Lisala sites, the ITNs’ distribution campaign took place in 2016 (problem with the quality of ITNs found in households). The other two variables were not statistically significant.

It is observed that malaria transmission is intense in Mweka with a very high entomological HBR index of 19.8 bites/human/night and EIR index of 2.73 infectious bites/human/night. and This contrasts with the good coverage in ITNs of 90% recorded in 1 year after the mass distribution of ITNs. There is also a strong resistance of An. gambiae s.l. to Alphacypermethrin (61%) and Deltamethrin (56%). These observations corroborate those of Metelo et al. [28] in Bandundu-ville and Apinjoh et al. [29]. These observations require a study to evaluate the effectiveness of ITNs in the context of malaria endemicity and resistance of An. gambiae s.l. to insecticides.

Anopheles behaviors were more endophilic and endophagic in all collection sites. The aggressive cycle started early around 6 p.m. and peaked around 12 a.m. This activity remained intense all night with high rates of aggressiveness, infectivity and entomological inoculation rate in all sites. It emerges that these sentinel sites, which are located at different locations and climatic conditions, thus present the same condition of transmission of malaria and the same incidence.

2.3.3 Anopheles susceptibility profile of gambiae s.l. against common insecticides

For more than a decade, the NMCP has been carrying out mass distribution campaigns for ITNs in order to obtain wide coverage (≥80%) for effective protection of the vulnerable population [4]. The widespread use of ITNs must require regular monitoring of the sensitivity of wild Anopheles populations to the insecticides used in the fight for resistance management [30]. This is the subject of the creation of sentinel sites for monitoring malaria vector bionomics and resistance [4].

It emerges from our study that DDT was ineffective in all sites and no knockdown effect (Kdt₅₀ and Kdt₉₅) was achieved. The populations of An. gambiae s.l. in all sites exhibited high resistance (11–19%). These results corroborate those of Basilua et al. [13], Watsenga et al. [9] and Metelo et al. [31]. This is justified by the fact that these sites were agricultural sites at the time and benefited from the use of pesticides.

Deltamethrin and Alphacypermethrin reached Kdt₅₀ late almost everywhere, except in Nyakunde. Deltamethrin was less effective compared to Alphacypermethrin. Anyway, the two pyrethroids (Deltamethrin and Alphacypermethrin) were less effective and the wild populations of An. gambiae s.l. have been resistant to them. Curiously after pre-exposure of An. gambiae s.l. to PBO, the efficacy of these two pyrethroids was restored at all sites.

3. Conclusions

The transmission of malaria is intense in the sentinel sites of the NMCP and is located in the same epidemiological facies where the conditions of transmission of the disease and the incidence are identical. This situation is aggravated by noncompliance with ITN distribution cycles and poor durability. This transmission is ensured by the An. gambiae s.l. with high resistance statuses to pyrethroids (Deltamethrin and Alphacypermethrin). The ecological choice is necessary for a good representation.

In addition, this study will allow the development of mapping and will serve as a benchmark for future entomological assessments to improve malaria surveillance in the DRC.

Acknowledgments

Our sincere thanks to Prof. Dr. Jean-Jacques Muyembe, Director General of the National Institute of Biomedical Research, for his support. Thank you teams of entomologists from INRB and KPSK for their support. We are very grateful to Marianne Sinka for correcting the English grammar.

Conflict of interest

The authors declare no conflict of interest.