Table 3.
Impact of cluster-level bednet usage on mosquito mortality. Results from generalised mixed-effect models examining the impact of cluster-level bednet usage on mosquito mortality
| All countries combineda | Beninb | Cameroonc | Indiac | Kenyab | Sudanc | ||
|---|---|---|---|---|---|---|---|
| No. of clusters included (year) | 59 (2012); 87 (2013); 143 (2014); 99 (2015); 80 (2016) | 19 (2015) | 22 (2013); 26 (2014) | 80 (2015); 80 (2016) | 13(2012); 41 (2014) | 46 (2012); 65 (2013); 76 (2014) | |
| Mean net usage (range) (%) | 74.9 (52.5–100) | 67.8 (7.0–100) | 89.9 (60.9–100) | 94.2 (73.7–100) | 78.6 (0–100) | ||
| Effect of cluster-level net usage on mosquito mortality, OR (95% CI) | < 40% | 1 (reference) | – | 1 (reference) | – | – | 1 (reference) |
| 40–80% | 1.03 (0.89–1.19) | 1 (reference) | 1.61 (1.21–2.14) | 1 (reference) | 1 (reference) | 0.69 (0.58–0.83) | |
| > 80% | 0.65 (0.57–0.74) | 1.59 (0.75–3.37) | 1.40 (1.08–1.82) | 0.36 (0.29–0.44) | 2.38 (0.56–10.1) | 0.45 (0.38–0.53) | |
| P-value | <0.001 | 0.225 | <0.001 | <0.001 | 0.241 | <0.001 |
aAdjusted for country
aAdjusted for district
cAdjusted for district, temperature and humidity
Results are presented in terms of change in mortality of mosquitoes for increasing bednet usage category (< 40%; between 40–80%; and above 80%). Bednet usage was calculated for years where cross-sectional survey data was available. Odds ratios are adjusted for locality and temperature and humidity where indicated. The results are shown for each country, as well as all countries combined (with country included as a covariate). Cluster was included as a random effect in all models