Study design and site
In this report, we describe the first implementation of the cluster grid response strategy carried out in the community during the cholera outbreak in Kinshasa, in which water and hygiene interventions were carried out in cholera case clusters. We also performed a preliminary community trial study to assess the temporal pattern of the outbreak in Kinshasa, in terms of number of cases per health zone and Kinshasa Province overall, before and after response interventions were implemented. Cholera surveillance databases from the Ministry of Health were analyzed to assess the spatiotemporal dynamics of the outbreak using epidemic curves and maps.
Kinshasa Province is one of 26 provinces in the DRC and is coterminous with the national capital. The city-province is divided into 35 administrative health zones. Kinshasa is located in the far west of the country on the banks of the Congo River (Figure 1). The province covers approximately 9,965 km2, with an estimated population of nearly 12 million. Infrastructure measures in the city have not kept pace with urbanization and the increasing population (14). As a result, neighborhoods have been established in flood-prone areas where water drainage is a challenge, thus increasing the risk and severity of flooding, especially during heavy rains in November and April (15).
Figure 1. Map of study area: Kinshasa Province. DRC, Democratic Republic of the Congo.
Surveillance data sources
The National Integrated Disease Surveillance and Response System was established in 2000 by the DRC Ministry of Health in conjunction with the World Health Organization (WHO). The Integrated Disease Surveillance and Response System targets thirteen infectious diseases with epidemic potential (including cholera) for passive surveillance (16). In each cholera treatment center (CTC), suspected cases and deaths due to moderate or severe cholera infection are documented via line list (17), which includes the patient’s address, age, sex, date of admission, date of onset, travel history during 14 days prior to symptom onset, and observation of any other individuals with diarrhea living in the patient’s home. Trained Ministry of Health officials aggregate and anonymize these data at the health zone level and report the data to the Ministry of Health in Kinshasa on a weekly basis.
Cholera case definition
According to WHO policy, a suspected case of cholera is defined as ‘‘any person two years of age or older in whom acute watery diarrhea with or without vomiting develops” during a cholera outbreak (18). The age limit is increased to five years and older during inter-epidemic periods to reduce the number of false positives. At the beginning of an outbreak, five to ten stool samples from each health zone are laboratory-confirmed through identification of Vibrio cholerae in culture. Subsequent cases of acute watery diarrhea in the same geographic region are presumed to be cholera.
Management and analysis of epidemiological data
Secondary data was extracted from surveillance databases organized by staff of the National Program for Cholera Elimination and Diarrheal Disease Control (Programme National d’Elimination du Choléra et de lutte contre les Maladies Diarrhéiques [PNECHOL-MD]). The database was verified for consistency and analyzed to determine weekly case numbers per health zone using Microsoft Excel. Epidemic curves per health zone were generated to assess the temporal evolution of the outbreak in Kinshasa as well as outbreaks in each affected health zone in Kinshasa Province, covering the period week 1 of 2017 to week 45 of 2018 (the epidemic curve shows cases starting from week 15 of 2017 because few cases were reported earlier in the year). Total weekly case numbers per health zone and in Kinshasa Province were also calculated to assess the temporal pattern of the outbreak before and after response interventions were implemented. The total suspected cholera case numbers reported in each health zone from November 1st 2017 to March 31st 2018 were used to represent the geographic distribution of cholera cases during the main outbreak period shown in Figure 4.
Cartography
The maps of Kinshasa and the DRC were generated using QGIS V3.4.3 Madeira with shapefiles provided by the DRC Ministry of Health (DRC health zones, DRC provinces, rivers and lakes). Additionally, shapefiles of Republic of the Congo administrative boundaries and transportation network features (rail and road) were retrieved from DIVA-GIS (http://www.diva-gis.org/gdata). The GPS coordinates of the CTCs were provided by the Kinshasa Ministry of Health.
Precipitation data
Precipitation levels were derived from the Climate Hazards Group InfraRed Precipitation with Station (CHIRPS) dataset (product: Daily UCSB CHIRPS v2p0 daily-improved global 0p25). The CHIRPS precipitation data is a 30-year quasi-global rainfall dataset supported by the University of California at Santa Barbara (CA, USA). Daily values were extracted and aggregated by health zone (R environment for statistical computing and graphics). Spatial aggregation from gridded data at the province level for Kinshasa was carried out using R. Daily precipitation levels (mm) were then aggregated by week using Microsoft Excel.
Field visits
Field visits were conducted by joint investigation teams composed of representatives of the PNECHOL-MD, Provincial Health Directorates and community agents in each affected health zone. Investigation teams met with local surveillance departments and health facilities. Information was collected concerning potential sources of infection, risk factors and possible links between cases (17). In each case cluster, the investigation teams also evaluated local demographic data, WASH (water, sanitation and hygiene) indicators and other factors that may play a role in cholera dynamics (19):
- Local demographic data (number of people per household, occupation and workplace of adults in the household).
- Factors contributing to amplification or persistence of an outbreak: high population density, potentially contaminated drinking water sources, poor water quality, poor sanitation (open defecation, broken sewer pipes, etc.), poor food hygiene, crowded or high-risk gathering places (markets, transportation hubs, schools, waterbodies used for bathing, etc.), and flooding.
Description of the targeted response using the cluster grid strategy
The main objective of the cluster grid response strategy was to quickly target case clusters (including affected households and at-risk populations in the community), in which interventions focused on emergency water supply, household water treatment and safe storage, home disinfection and hygiene promotion (19–23).
To inform and guide the targeted response, the epidemiological data was analyzed to first identify the most affected health zones (reporting more than 10% of the total suspected cholera cases during the previous three-week period). Each new affected health zone that experienced a laboratory-confirmed outbreak and displayed risk factors that may trigger an increase in cases according to field investigations was also included in the response. As a result, five health zones (Binza Météo, Limeté, Kintambo, Kingabwa and Bumbu) were selected for targeted water supply and hygiene interventions. Over the course of the outbreak, daily case admission trends, including origin of patients, were monitored to assess the epidemiological evolution of the outbreak in real time and adjust response activities accordingly.
To target case clusters within each of the five health zones, the line lists of suspect cases were consulted in the CTCs to obtain the address of patients admitted over the previous 14-day period (CTCs were established in Limeté (Pakadjuma) and Binza Météo (Camp Luka)).
A response team then visited each patient residence to obtain GPS coordinates of the case households. The location of the most recent cases (<14 days) were mapped, and a circle (500-meter radius) was delineated around each cluster, which was then subdivided into a grid. Each grid unit represented an average of 20-30 households, which varied depending on the geographical characteristics of the area (Figure 2).
Figure 2. Schematic diagram of the cluster grid response strategy. The case cluster is shown in green, case residences are represented by red dots, nearby neighbors (≤50 m from a case household) are represented in red squares and peripheral neighbors (>50 m from a case household, within the case cluster) are represented in orange squares.
To reduce cholera transmission within the case clusters, the appropriate water and hygiene interventions were conducted depending on the transmission context (e.g., case households, public places in the community) (24).
The following response activities were carried out at case residences, close neighboring residences (≤50 m from a case household), and high-risk peripheral residences (>50 m from a case household) within case clusters:
- Household drinking water was systematically chlorinated everyday over a 14-day period using either water purification tablets (sodium dichloroisocyanurate, 7 mg) or 1% chlorine solution. To treat water of low turbidity (<5 NTU), 2 ml of 1% chlorine stock solution was added to 10 liters of water (final concentration 2 mg/L). To treat turbid water (>5 NTU), a jar test was used to determine the treatment necessary (19,20).
- Each household was provided with a household hygiene kit, containing soap, a 20-liter water storage container and ready-to-use chlorine for disinfection of drinking water containers (22,23). Water storage containers were distributed together with instructions to safely store household drinking water (25).
- To enhance health awareness and encourage safe practices, hygiene and health messages were delivered to the household members (19,21).
- For cholera case residences only, surfaces likely to be contaminated with vomit or diarrhea from a cholera patient were disinfected with a 0.2% chlorine solution (21) within less than 72 hours after patient registration at the CTC.
The activities carried out at in public places in the community within case clusters, over the course of 14 days, are described below. Additional details concerning the water and hygiene interventions in each health zone are displayed in Table 1.
- Water bladders (10-m3) were installed in public places. The water bladders are self-supporting closed flexible tanks composed of a coated polyester mesh, which are used to store drinking water for distribution (Labaronne Citaf, Pont-Eveque, France; https://www.labaronne-citaf.com/products/self-supporting-closed-flexible-tank/). Water bladders were installed in neighborhoods of high population density without a source of safe drinking water nearby,Two water bladders were installed in Limeté near the CTC, where they also served to provide safe drinking water to the local population, and one water bladder was installed in Kingabwa. refilled every 48 to 72 hours by national water company tankers (Regideso). The residual chlorine levels were verified prior to distribution.
- Fixed water chlorination points (bucket chlorination) were installed in public places in the health zones most affected early during the outbreak: Binza Météo and Kintambo.
- Handwashing points were installed in public places in Binza Météo and Kintambo.
- Hygiene education messages were disseminated to the community to promote health-seeking behaviors and protection mechanisms via health promotion campaigns in public places (i.e., markets, schools, transport stations, water points). Messages were communicated via radio, TV, posters and town criers, and topics included the modes of transmission, water treatment, and the importance of reporting cases of severe diarrhea (20).
- Public health rules were enforced by the local authorities - swimming in waterbodies (e.g., lakes, rivers, streams) was banned during the 14-day period (27).
Table 1. Response intervention details per health zone.
Health zone
|
Details of water and hygiene interventions in the community
|
Intervention duration
|
Number of
water bladders
|
Number of fixed water chlorination points
|
Number of handwashing points
|
Binza Météo
|
60 days
|
0
|
15
|
4
|
Limeté
|
30 days
|
2
|
0
|
0
|
Kintambo
|
30 days
|
0
|
2
|
3
|
Kingabwa
|
30 days
|
1
|
0
|
0
|
Bumbu
|
30 days
|
0
|
0
|
0
|
Field response teams consisted of a supervisor, two hygiene promotion educators (a crier and a door-to-door educator), four team members who performed water chlorination (two for fixed sites and two for door-to-door household visits), two team members who performed home disinfection and two attendants at handwashing points. Each team covered at least two 30-household grid units. The number of personnel involved per intervention type in each health zone is detailed in Table 2.
Table 2. Number of personnel involved per intervention type for each health zone.
Health zone
|
Number of personnel per intervention type and role
|
Hygiene
promotion
|
Water
chlorination
|
Home
disinfection
|
Supervision
|
Total Personnel
|
Binza Météo
|
135
|
60
|
42
|
20
|
257
|
Limeté
|
8
|
10
|
10
|
4
|
32
|
Kintambo
|
40
|
8
|
10
|
5
|
63
|
Kingabwa
|
18
|
17
|
10
|
5
|
50
|
Bumbu
|
40
|
40
|
30
|
8
|
118
|
Total
|
241
|
135
|
102
|
42
|
520
|
Additional response activities conducted in case clusters
In parallel with water and hygiene activities, active case search was carried out in the community, prioritizing the immediate entourage of probable and confirmed cases identified or treated at the CTC (20). Chemoprophylaxis of all immediate contacts of cholera cases was also conducted during household visits to provide short-term protection against infection (28,29). Adults received a single dose of doxycycline (300 mg), pregnant women received a single dose of ciprofloxacin (1 g) and children received a single dose of ciprofloxacin (20-30 mg/kg) (19).
Ethics
Ethics approval was not required for this study because cholera surveillance and response are covered by national public health laws as an integral part of the public health mandate of the DRC Ministry of Health.