The overall effective coverage of vector control interventions in children under five was found to be extremely high, at 98%, mostly due to high IRS coverage. Maintaining such high coverage is a key driver for keeping the malaria prevalence low and reducing the risk of malaria resurgence. Thus, it improves the prospect of malaria elimination in Zanzibar .
Seventy percent of the under-five children were sleeping under an LLIN. Effective coverage was equitable, ie, equally high in the poorest compared to the least poor socio-economic group. Targeted free mass distribution campaigns were previously found to result in high and equitable under-five coverage in Zanzibar . In the current study, effective ITN coverage among under-fives was found to be 76% after an untargeted mass distribution. This figure is comparable to the effective ITN coverage of 73% in Sierra Leone and 62% in Nigeria [33, 34], following a similar distribution strategy.
IRS effective coverage was much higher; covering 95% of the under-five population, and although it was lower in the poorest compared to the least poor quintile, the difference was not statistically significant. This is also likely due to the fact that IRS is delivered free of charge to all households.
The majority of children (66%) were protected by both LLINs and IRS. Although simultaneous use of LLINs and IRS could benefit from an additive effect, especially if different insecticides are used , Zanzibar had chosen to use pyrethroids for both LLINs and IRS. Nevertheless, the implementation of these two interventions simultaneously elevated the effective coverage by at least one malaria prevention intervention for under-five children.
Given the relatively lower effective coverage of LLINs and the fact that high seasonal usage pattern was detected in the study, as was observed in other studies [13, 28, 29], it will be important to continue IRS efforts until LLIN coverage can be further improved.
Malaria reduction in Zanzibar is well established, as shown by cross-sectional surveys and health facility records [3, 4]. This study indicates that caretakers have noted this reduction, and that low malaria risk perceptions were not found to negatively influence LLIN use, as was previously indicated in qualitative studies in Vanuatu  and Zanzibar .
Bed nets were highly appreciated by caretakers, and were the most commonly spontaneously mentioned method of preventing malaria. In addition to being useful in preventing malaria, they were also perceived as useful in preventing mosquito bites. This added benefit of bed nets as a way of preventing mosquito nuisance has been documented previously [35–38]. However, in this study, perceptions about bed nets were not found to be significantly associated with LLIN effective coverage.
While coverage of IRS was higher than that of bed nets, it was slightly less appreciated for reducing malaria and mosquitoes. This is in line with studies in Mozambique which have shown that acceptance of IRS relied more on sociopolitical factors rather than perceived benefits of malaria and mosquito prevention . Furthermore, one in five caretakers in this study mentioned that IRS also had disadvantages such as itching, which has also been documented previously . Other side effects included insect increase (in particular in bed bugs), which has also been reported from studies where DDT was used for IRS .
This study shows that perceptions of malaria risk, as well as the perceived benefits of vector control interventions, be it malaria prevention or mosquito nuisance prevention, are not significantly associated with the effective coverage of these interventions. Most net distributions are usually accompanied by intensive health information campaigns that focus on improving knowledge and awareness to increase the use of nets against malaria. However, it remains unclear what factors do in fact influence community members to comply with and use these interventions. Factors that were not investigated in this study, and that were shown to influence acceptance of some interventions, include sociopolitical aspects, like those found to influence IRS acceptance in Mozambique . More qualitative research to elucidate these aspects would be beneficial in designing more effective Behaviour Change Communication (BCC) campaigns to accompany these interventions.
While the majority of community members stated an intention to continue using malaria prevention methods as malaria further decreases, community members' intentions to continue adherence to bed nets and IRS will not result in high effective coverage unless high access to these interventions is maintained.
Most questions in the structured questionnaire were dichotomous closed questions with "yes" or "no" answers. These included the questions on perceived risk of malaria, perceived usefulness of bed nets and IRS in preventing malaria and mosquito bites, perceived usefulness of combining preventive measures and continued use of preventive measures after malaria further decreases. Although giving a dichotomous answer to these qualitative-natured questions could cause a bias, it was deemed as an appropriate way to quantify these perceptions. Another option would have been to use other quantitative techniques, such as Likert-type scales, which could potentially have given more nuanced responses. However, following the dichotomous questions, open-ended questions were used, where the participants could further explain or justify their answers.
Use and willingness to use malaria prevention might have been overestimated due to desirability bias. This bias may have been increased due to the fact that the survey was identified with the Zanzibar Malaria Control Program and the interviewers were health professionals. However, an attempt to minimize this risk was made through emphasizing the importance of creating a comfortable environment during the interview.