Towards malaria elimination in Zanzibar: identifying opportunities for improving surveillance and response strategies

Malaria is a potentially deadly parasitic disease with recent estimates of 249 million cases, and approximately 608,000 deaths. The malaria burden is most severe in Sub-Saharan Africa, which accounts for 94% of malaria cases and 95% of all malaria related deaths. Over the past twenty years, significant progress in malaria control has been made as a result of the large-scale rollout of vector control interventions, improved diagnosis, and highly effective artemisinin-based combination therapy. The WHO Global Technical Strategy for Malaria 2016–2030 set targets for the reduction in malaria incidence and mortality rates worldwide, including at least a 90% decrease by 2030 compared to 2015.
Despite substantial advancements in malaria control, recent reports indicate a resurgence of malaria, suggesting that the global progress in malaria control and elimination is off-target. This resurgence shows the limitations of current approaches and underlines the critical need for intensifying surveillance approaches in countries at all levels of transmission.
The government of Zanzibar renewed its malaria control strategy in 2002 aiming at elimination. However, malaria infections still persist on the Zanzibar islands, linked to importation of cases from high transmission regions, primarily in mainland Tanzania, and local transmission. A well-functioning surveillance and response system is critical in this context to understand the dynamics of transmission and target response measures. Case-importation was shown to contribute to the ongoing transmission but there is limited knowledge about the specific population groups that import malaria infections to Zanzibar – information that is key to a targeted response. Similarly, while there is ongoing local transmission and malaria foci can be routinely identified, there is little knowledge of local risk factors and high-risk populations that would facilitate targeted measures.
This PhD thesis had the following three specific objectives: a) to understand the malaria surveillance-response system in Zanzibar through exploring its challenges and opportunities; b) to investigate the factors contributing to the ongoing importation of malaria parasites, identify the patterns of travel, and propose effective strategies to prevent importation; c) to assess the risk factors associated with local malaria infections among the population group with no recent history of travel outside Zanzibar.
In a qualitative study that included semi-structured interviews with key-informants and community members, the performance of the surveillance system implemented by the Zanzibar Malaria Elimination Programme was explored. The study found that the surveillance system is faced with a number of flaws, including challenges related to logistics, electronic system functionality and health service delivery related to both providers and supplies, which threaten its smooth operations. At the same time, the system was found to be resilient to challenges, benefiting from numerous strengths and substantial programme support. To accelerate the pace of malaria elimination in Zanzibar constant supplies should maintained, bugs in the electronic system should be addressed in time, and appropriate incentives should be provided to the health workforce.
Using cross-sectional survey data (collected in 2018 - 2019) with travel details and malaria test results of Zanzibar residents who travelled to mainland Tanzania in the last eight weeks, risk factors for imported malaria infections were investigated. The results show that among travellers to mainland Tanzania, 64% reported traveling to districts with a ‘high’ malaria endemicity and 12% to a district with ‘moderate’ malaria endemicity. Travelers visiting highly endemic districts had a substantially higher odds of being infected with malaria (diagnosed by qPCR) compared to those who travelled only to districts where malaria-endemicity was classified as low or very low adjusted odd ratio [aOR] = 7.0, 95% Confidence Interval – [CI] 1.9 - 25.5). To reduce malaria importation to Zanzibar, identifying population groups traveling to highly endemic areas in mainland and targeting interventions to these groups may prevent the importation of a substantial number of infections.
Using data from the same cross-sectional survey, the association of malaria infection and risk factors (related to socio-demographic and household characteristics and bed net use) was investigated for those participants who reported no travel outside Zanzibar in the previous eight weeks. Individuals living in households in which a clinical malaria case was reported, school age children and youths had a greater likelihood of being infected with malaria (aOR = 18.4, 95% CI: 11.4 – 29.7; aOR = 2.9, 95% CI: 1.9 – 4.5;, aOR = 2.0, 95% CI: 1.3 – 3.1, respectively), compared to individuals living in non-index households, and adults aged 26 years and above respectively. Individuals whose job was reported as entrepreneurship (which includes any type of self-employed business) had higher odds of malaria infections (aOR = 1.8, 95% CI: 1.1 – 2.7), compared to those who were unemployed. Going to sleep between 20:00 and 22:00 hours (aOR = 0.4, 95% CI 0.3 – 0.6) and later than 22:00 (aOR = 0.5, 95% CI 0.3 – 0.9) exhibited a decreased likelihood of being infected =, compared to going to sleep earlier between 18:00 and 20:00 hours. Residing in poorer house structures with window openings that cannot be closed or are not screened was associated with higher odds of infection (aOR = 2.1, 95% CI 1.4 - 3.1) compared to living in houses with windows that could prevent entry of mosquitoes. Odds of infection were significantly higher among members living in households which did not own bed nets (aOR = 2.3, 95% CI 1.4 - 3.6) as compared with those who had bed nets. Individuals living in areas with a moderate or high Palmer Drought Severity Index (i.e. in wetter areas) had substantially higher odds of being infected (aOR = 4.0, 95% CI 2.5 – 6.4, and aOR = 2.9, 95% CI 1.7 – 5.0), than those who lived in drier areas. The difference between residents of Unguja and Pemba was not found to be significant. To address locally acquired malaria infections in Zanzibar, targeted efforts should include school-age children, promote the uptake of existing interventions such as bed nets, and encourage housing improvements. Additional studies may be required to identify further high-risk groups based on a detailed analysis of behavioural and occupational patterns.
The surveillance-and-response system currently implemented in Zanzibar appears insufficient to providing all necessary information for a targeted management of malaria, as reflected in ongoing importation of infections from mainland Tanzania and ongoing local transmission. Findings from this thesis and complementary studies suggest that progress in malaria elimination in Zanzibar requires a further strengthening of the surveillance-response system to identify sources of infection and clear asymptomatic carriers, measures to reduce malaria importation into Zanzibar (ideally including transmission reduction in mainland Tanzania), and the promotion and targeting of efforts to identify and interrupt local chains of transmission.