We used computer-generated simulations to estimate power and found that with a 1:1 allocation ratio, one child per house followed up for 2 years would require 400 houses in each study group to detect a 35% reduction (the minimum we considered to be of public health importance) in the rate of malaria cases with over 80% power at the 5% significance level. For the entomological endpoint, the study was 80% powered to detect a 50% reduction at the 5% level of significance. Methods for sample size calculation have been described previously.18

Analysis was done with Stata version 15.0. Data analysis followed an analysis plan written before study completion. All analyses were done on a modified intention-to-treat basis in which all children with fewer than 50% visits in a year were not included. Data were censored for 4 weeks after a child had a documented malaria infection, and if the child had moved from the study house. If a child was absent more than half the scheduled visits in transmission season, their data were censored for that year. We compared between study groups the incidence rate of clinical malaria over the course of the transmission seasons (June to December) in 2016 and 2017, both separately and combined. Poisson regression with robust standard errors to account for repeat episodes for a child was used to compare the rates between groups, together with Kaplan-Meier plots showing time to first event by study group. An analysis adjusted for age, riverbank, month of year, and ethnicity was also carried out. The rate ratio comparing the mean number of female A gambiae sensu lato per light trap per night groups was calculated using negative binomial regression with a random effect for house, adjusting for ethnicity, riverbank, and burning incense (churai).

Parasite rates and density, anaemia prevalence and haemoglobin concentrations from the cross-sectional data were analysed by either logistic or linear regression according to whether the variable was binary or continuous. The rate of respiratory infection over the course of the two annual transmission seasons was analysed in the same way as clinical malaria.

Entomological inoculation rate (EIR) was estimated in each study group and is defined as the number of infective bites received per person during the transmission season, and was estimated as follows: EIR=HDM × SPR × n (where HDM is the household density of mosquitoes, which is estimated as the mean number of A gambiae sensu lato per light trap per night; SPR is the sporozoite rate; and n is the number of nights during the transmission season, July to December [n=6 × 30=180]).

Houses classified in good condition were those with all window and door screening intact and with the doors having no gaps around the edges when shut. We hypothesised that houses where the doors were kept closed during the day were more likely to have their doors shut at night, compared with houses where the doors were left ajar. Similarly, we considered that houses with more people would have their doors opened more frequently during the night than those with fewer people.

A Data Safety Monitoring Board reviewed the study procedures and results. The trial is registered at ISRCTN02622179.

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.