How Real-Time Case-Based Malaria Surveillance Helps Zanzibar Get a Step Closer to Malaria Elimination: Description of Operational Platform and Resources.

Testing and treating asymptomatic populations have the potential to reduce the population's parasite reservoir and reduce malaria transmission. Zanzibar's malaria case notification (MCN) platform collects detailed sociodemographic and epidemiological data from all confirmed malaria cases to inform programmatic decision-making. We describe the design and operationalization process of the platform and other malaria surveillance resources that are enabling Zanzibar's progress toward malaria elimination.The MCN platform consists of an interactive short message service (SMS) system for case notification, a software application for Android mobile devices, a visual question set and workflow manager, a back-end database server, and a web browser-based application for data analytics, configuration, and management. Malaria case data were collected from August 2012 to December 2021 and reported via SMS from all public and private health facilities to a central database and then to district malaria surveillance officers' mobile devices. Data included patient names, shehia (administrative area), and date of diagnosis, enabling officers to track patients, ideally within 24 hours of reporting. Patients' household members were tested for malaria using conventional rapid diagnostic tests (RDTs). Treatment using artemisinin-based combination therapy was provided for persons testing positive.Between 2012 and 2021, a total of 48,899 index malaria cases were confirmed at health facilities, 22,152 (45.3%) within 24 hours of reporting; 41,886 (85.7%) cases were fully investigated and followed up to the household level. A total of 111,811 additional household members were tested with RDTs, of whom 10,602 (9.5%) were malaria positive.The MCN platform reports malaria case data in near real time, enabling prompt follow-up of index cases and prompt testing and treatment of members in index case households. Along with routine testing and treatment and other preventive interventions, the MCN platform is foundational to the programmatic efforts in further reducing malaria and ultimately eliminating autochthonous malaria transmission in Zanzibar.

Improvements in malaria surveillance through the electronic Integrated Disease Surveillance and Response (eIDSR) system in mainland Tanzania, 2013-2021.

BACKGROUND: Tanzania has made remarkable progress in reducing malaria burden and aims to transition from malaria control to sub-national elimination. In 2013, electronic weekly and monthly reporting platforms using the District Health Information System 2 (DHIS2) were introduced. Weekly reporting was implemented through the mobile phone-based Integrated Disease Surveillance and Response (eIDSR) platform and progressively scaled-up from 67 to 7471 (100%) public and private health facilities between 2013 and 2020. This study describes the roll-out and large-scale implementation of eIDSR and compares the consistency between weekly eIDSR and monthly DHIS2 malaria indicator data reporting, including an assessment of its usefulness for malaria outbreak detection and case-based surveillance (CBS) in low transmission areas. METHODS: The indicators included in the analysis were number of patients tested for malaria, number of confirmed malaria cases, and clinical cases (treated presumptively for malaria). The analysis described the time trends of reporting, testing, test positivity, and malaria cases between 2013 and 2021. For both weekly eIDSR and monthly DHIS2 data, comparisons of annual reporting completeness, malaria cases and annualized incidence were performed for 2020 and 2021; additionally, comparisons were stratified by malaria epidemiological strata (parasite prevalence: very low < 1%, low 1 ≤ 5%, moderate 5 ≤ 30%, and high > 30%). RESULTS: Weekly eIDSR reporting completeness steadily improved over time, with completeness being 90.2% in 2020 and 93.9% in 2021; conversely, monthly DHIS2 reporting completeness was 98.9% and 98.7% in 2020 and 2021, respectively. Weekly eIDSR reporting completeness and timeliness were highest in the very low epidemiological stratum. Annualized malaria incidence as reported by weekly eIDSR was 17.5% and 12.4% lower than reported by monthly DHIS2 in 2020 and 2021; for both 2020 and 2021, annualized incidence was similar across weekly and monthly data in the very low stratum. CONCLUSION: The concurrence of annualized weekly eIDSR and monthly DHIS2 reporting completeness, malaria cases and incidence in very low strata suggests that eIDSR could be useful tool for early outbreak detection, and the eIDSR platform could reliably be expanded by adding more indicators and modules for CBS in the very low epidemiological stratum.

Risk factors associated with malaria infection identified through reactive case detection in Zanzibar, 2012-2019.

BACKGROUND: Over the past two decades, Zanzibar substantially reduced malaria burden. As malaria decreases, sustainable improvements in control interventions may increasingly depend on accurate knowledge of malaria risk factors to further target interventions. This study aimed to investigate the risk factors associated with malaria infection in Zanzibar. METHODS: Surveillance data from Zanzibar's Malaria Case Notification system from August 2012 and December 2019 were analyzed. This system collects data on malaria cases passively detected and reported by all health facilities (index cases), and household-based reactive case detection (RCD) activities linked to those primary cases. All members of households of the index cases were screened for malaria using a malaria rapid diagnostic test (RDT). Individuals with a positive RDT were treated with artemisinin-based combination therapy. Univariate and multivariate logistic regression analyses were done to investigate the association between RDT positivity among the household members and explanatory factors with adjustment for seasonality and clustering at Shehia level. RESULTS: A total of 30,647 cases were reported of whom household RCD was completed for 21,443 (63%) index case households and 85,318 household members tested for malaria. The findings show that younger age (p-value for trend [Ptrend] < 0.001), history of fever in the last 2 weeks (odds ratio [OR] = 35.7; 95% CI 32.3-39.5), travel outside Zanzibar in the last 30 days (OR = 2.5; 95% CI 2.3-2.8) and living in Unguja (OR = 1.2; 95% CI 1.0-1.5) were independently associated with increased odds of RDT positivity. In contrast, male gender (OR=0.8; 95% CI 0.7-0.9), sleeping under an LLIN the previous night (OR = 0.9; 95% CI 0.7-0.9), having higher household net access (Ptrend < 0.001), and living in a household that received IRS in the last 12 months (OR = 0.8; 95% CI 0.7-0.9) were independently associated with reduced odds of RDT positivity. A significant effect modification of combining IRS and LLIN was also noted (OR = 0.7; 95% CI 0.6-0.8). CONCLUSIONS: The findings suggest that vector control remains an important malaria prevention intervention: they underscore the need to maintain universal access to LLINs, the persistent promotion of LLIN use, and application of IRS. Additionally, enhanced behavioural change and preventive strategies targeting children aged 5-14 years and travellers are needed.

Improved spatial ecological sampling using open data and standardization: an example from malaria mosquito surveillance.

Vector-borne disease control relies on efficient vector surveillance, mostly carried out using traps whose number and locations are often determined by expert opinion rather than a rigorous quantitative sampling design. In this work we propose a framework for ecological sampling design which in its preliminary stages can take into account environmental conditions obtained from open data (i.e. remote sensing and meteorological stations) not necessarily designed for ecological analysis. These environmental data are used to delimit the area into ecologically homogeneous strata. By employing Bayesian statistics within a model-based sampling design, the traps are deployed among the strata using a mixture of random and grid locations which allows balancing predictions and model-fitting accuracies. Sample sizes and the effect of ecological strata on sample sizes are estimated from previous mosquito sampling campaigns open data. Notably, we found that a configuration of 30 locations with four households each (120 samples) will have a similar accuracy in the predictions of mosquito abundance as 200 random samples. In addition, we show that random sampling independently from ecological strata, produces biased estimates of the mosquito abundance. Finally, we propose standardizing reporting of sampling designs to allow transparency and repetition/re-use in subsequent sampling campaigns.