Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Jun 12;16(1):248.
doi: 10.1186/s12936-017-1903-z.

Effectiveness of Reactive Case Detection for Malaria Elimination in Three Archetypical Transmission Settings: A Modelling Study

Affiliations
Free PMC article

Effectiveness of Reactive Case Detection for Malaria Elimination in Three Archetypical Transmission Settings: A Modelling Study

Jaline Gerardin et al. Malar J. .
Free PMC article

Abstract

Background: Reactive case detection could be a powerful tool in malaria elimination, as it selectively targets transmission pockets. However, field operations have yet to demonstrate under which conditions, if any, reactive case detection is best poised to push a region to elimination. This study uses mathematical modelling to assess how baseline transmission intensity and local interconnectedness affect the impact of reactive activities in the context of other possible intervention packages.

Methods: Communities in Southern Province, Zambia, where elimination operations are currently underway, were used as representatives of three archetypes of malaria transmission: low-transmission, high household density; high-transmission, low household density; and high-transmission, high household density. Transmission at the spatially-connected household level was simulated with a dynamical model of malaria transmission, and local variation in vectorial capacity and intervention coverage were parameterized according to data collected from the area. Various potential intervention packages were imposed on each of the archetypical settings and the resulting likelihoods of elimination by the end of 2020 were compared.

Results: Simulations predict that success of elimination campaigns in both low- and high-transmission areas is strongly dependent on stemming the flow of imported infections, underscoring the need for regional-scale strategies capable of reducing transmission concurrently across many connected areas. In historically low-transmission areas, treatment of clinical malaria should form the cornerstone of elimination operations, as most malaria infections in these areas are symptomatic and onward transmission would be mitigated through health system strengthening; reactive case detection has minimal impact in these settings. In historically high-transmission areas, vector control and case management are crucial for limiting outbreak size, and the asymptomatic reservoir must be addressed through reactive case detection or mass drug campaigns.

Conclusions: Reactive case detection is recommended only for settings where transmission has recently been reduced rather than all low-transmission settings. This is demonstrated in a modelling framework with strong out-of-sample accuracy across a range of transmission settings while including methodologies for understanding the most resource-effective allocations of health workers. This approach generalizes to providing a platform for planning rational scale-up of health systems based on locally-optimized impact according to simplified stratification.

Keywords: Human movement; Malaria elimination; Mathematical modeling; Reactive case detection; Stratification.

Figures

Fig. 1
Fig. 1
Three health facility catchment areas (HFCAs) in the Lake Kariba region of Southern Province, Zambia, span a range of transmission intensities and population densities. a Village-level prevalence of RDT-positive infections in June 2012 prior to mass drug campaigns shows higher transmission in lakeside areas and lower transmission in higher-altitude villages. Circle size is proportional to village population. b In Bbondo HFCA, households are highly clustered and baseline prevalence is low. c In Luumbo HFCA, households are dispersed and baseline prevalence is high. d In Munyumbwe HFCA, households are predominantly clustered around major roads, and baseline prevalence is mixed, with higher prevalence in the southwest valley and eastern roadside areas
Fig. 2
Fig. 2
Timeline of interventions carried out in Bbondo, Luumbo, and Munyumbwe HFCAs from 2007 to 2016
Fig. 3
Fig. 3
Simulated transmission intensity in a Bbondo b Luumbo and c Munyumbwe HFCAs compared with RDT prevalence measurements taken during MTAT and MDA/fMDA rounds and RDT-confirmed fevers reported by health facilities and CHWs. In the bottom row of each panel, simulated clinical cases included both treated and untreated symptomatic cases of malaria, while simulated treated cases are cases that have contacted the health system and would potentially show up in clinic and CHW reporting. The relevant comparison is therefore between the gray bars and the yellow line and area. Each simulation trace shows the mean and range of 1000 realizations, with the yellow shaded area expanded by 150% to indicate additional uncertainty in treatment and reporting rates
Fig. 4
Fig. 4
Success of elimination programmes in Bbondo HFCA is dominated by effects of reducing importation and improving case management. a Transmission in Bbondo HFCA was monitored under 1000 realizations of each intervention scenario from 2014 through 2020. Simulations with zero locally-acquired infections in 2020 were considered to have achieved elimination. b Five potential intervention packages were modelled in the external high-transmission area to which Bbondo residents may travel. See “Methods” for details on these intervention packages. c Probability of elimination in Bbondo under three potential case management rates and five potential intervention packages implemented in the external high-transmission area. d Probability of elimination in Bbondo under various case management rates and potential RCD implementations. Perfect RCD indicates 100% of index cases receiving follow-up and 100% of residents at home and receptive to follow-up activities. Scenarios were simulated under three importation rates, corresponding to intervention package #5 in the external high-transmission area (no importations), package #4 (low importation), and package #2 (high importation)
Fig. 5
Fig. 5
Elimination in Luumbo HFCA requires limiting importations, increasing case management rate, maintaining vector control, and either RCD or MDA campaigns. a Probability of elimination in Luumbo under three potential case management rates and five potential intervention packages implemented in the external high-transmission area. b Probability of elimination in Luumbo under various case management rates and potential RCD implementations. Importation rates correspond to intervention packages in the external high-transmission area as described in the caption to Fig. 4d. c Probability of elimination in Luumbo under various vector control packages implemented in Luumbo. All intervention scenarios were simulated under low importation (intervention package #4 in external high-transmission area)
Fig. 6
Fig. 6
Elimination in Munyumbwe HFCA requires limiting importations, increasing case management rate, enhancing vector control with targeted high-coverage campaigns, and both RCD and MDA. a Probability of elimination in Munyumbwe under three potential case management rates and five potential intervention packages implemented in the external high-transmission area. b Probability of elimination in Munyumbwe under various case management rates and potential RCD implementations. Importation rates correspond to intervention packages in the external high-transmission area as described in the caption to Fig. 4d. c Probability of elimination in Munyumbwe under various vector control packages implemented in Munyumbwe. All intervention scenarios were simulated under no importation (intervention package #5 in external high-transmission area)
Fig. 7
Fig. 7
Contribution of recently symptomatic individuals to the infectious reservoir in Bbondo, Luumbo, and Munyumbwe HFCAs. For each HFCA, a representative simulation was run over the year 2015 at baseline case management rate with no RCD but including any vector control and MDA or fMDA campaigns. The external high-transmission area received intervention package #1. a Prevalence of infected individuals, currently symptomatic individuals, and individuals who were symptomatic within the last 30 days but not currently symptomatic. b Relative contribution to the infectious reservoir in each HFCA from current and recent symptomatics. c Daily number of vectors infected by currently symptomatic, recently symptomatic, and all other individuals in each HFCA
Fig. 8
Fig. 8
Recommended intervention mixes for elimination in a low- and b high-importation settings, stratified by local baseline transmission intensity and household clustering. High-clustering, low baseline transmission based on results from Bbondo HFCA; low-clustering, high baseline transmission based on Luumbo HFCA; high-clustering, high baseline transmission based on Munyumbwe HFCA

Similar articles

See all similar articles

Cited by 15 articles

See all "Cited by" articles

References

    1. Cotter C, Sturrock HJ, Hsiang MS, Liu J, Phillips AA, Hwang J, et al. The changing epidemiology of malaria elimination: new strategies for new challenges. Lancet. 2013;382:900–911. doi: 10.1016/S0140-6736(13)60310-4. - DOI - PubMed
    1. Sturrock HJW, Hsiang MS, Cohen JM, Smith DL, Greenhouse B, Bousema T, et al. Targeting asymptomatic malaria infections: active surveillance in control and elimination. PLoS Med. 2013;10:e1001467. doi: 10.1371/journal.pmed.1001467. - DOI - PMC - PubMed
    1. Bousema T, Griffin JT, Sauerwein RW, Smith DL, Churcher TS, Takken W, et al. Hitting hotspots: spatial targeting of malaria for control and elimination. PLoS Med. 2012;9:e1001165. doi: 10.1371/journal.pmed.1001165. - DOI - PMC - PubMed
    1. Ndiath M, Faye B, Cisse B, Ndiaye J, Gomis J, Dia A, et al. Identifying malaria hotspots in Keur Soce health and demographic surveillance site in context of low transmission. Malar J. 2014;13:453. doi: 10.1186/1475-2875-13-453. - DOI - PMC - PubMed
    1. Rosas-Aguirre A, Speybroeck N, Llanos-Cuentas A, Rosanas-Urgell A, Carrasco-Escobar G, Rodriguez H, et al. Hotspots of malaria transmission in the Peruvian Amazon: rapid assessment through a parasitological and serological survey. PLoS ONE. 2015;10:e0137458. doi: 10.1371/journal.pone.0137458. - DOI - PMC - PubMed

Publication types

Feedback