Fig 1.
The total follow-up time was 1,592.5 person years at risk (PYAR). The median age was 13 years (Inter-quartile range (IQR) 5, 28; range 6 months-90 years). LLINs ownership was high (71.49%, 2766/3869) and their use varied significantly by month, with July, at the start of the transmission season, having the lowest proportion (52.22%, 1916/3669), and increasing significantly from August (87.05%, 2811/3229) to October (94.04%, 2809/2987), p<0.01 (Fig 2).
Table 1.
Study participant and house structural characteristics in the study sites.
Table 2.
Risk factors of P. falciparum infection at the start of the transmission season.
Fig 2.
Overall prevalence and incidence of malaria infection, clinical malaria, gametocytaemia and proportion of LLIN use by month.
Table 3.
Risk factors of sub-patent P. falciparum infections*.
Table 4.
Incidence rates of P. falciparum infections by month (95%CI).
Table 5.
Incidence rate of clinical malaria by month (95%CI)*.
Fig 3.
Prevalence and incidence of malaria infection and clinical malaria by month and region.
The incidence rates were significantly higher in three regions compared to the WCR; URR-N (P. falciparum infection = 2.82/PYAR, 95% CI: 2.57–3.10; and clinical malaria = 1.0/PYAR, 95%CI: 0.82–1.11, p<0.01) and URR-S (P. falciparum infection = 1.42/PYAR, 95% CI: 1.31–1.54; and clinical malaria = 0.8/PYAR, 95%CI: 0.68–0.85; p<0.01) had the highest incidence, followed by LRR (P. falciparum infection = 0.98/PYAR, 95% CI:0.86–1.12; clinical malaria = 0.36/PYAR, 95%CI: 0.29–0.455; p<0.01). The incidence of P. falciparum infections in the WCR (0.47/PYAR, 95% CI: 0.40–0.56) was 2 times higher than in the CRR (0.29/PYAR, 95%CI: 0.22–0.37) (p<0.01) where not a single clinical malaria case was detected (Table 6 and Fig 3).
Table 6.
Incidence of infection and clinical malaria per person-years, by month and region (95%CI).
Table 7.
Risk factors for ≥ 2 episodes of malaria infection.
Fig 4.
Kaplan Meier survival curves comparing the time to clinical malaria among household members with or without at least a malaria-infected individual at the start of the transmission season.
After adjusting for age, gender, travel, LLIN use and sleeping outdoors, the risk remained significantly higher only for WCR (HR = 4.0, 95% CI: 2.1–7.5, p<0.01) and URR-N (HR 1.5, 95% CI: 1.1–2.1, p = 0.02).
Table 8.
An. gambiae s.l. indoor and outdoor biting estimated by HLC, by region*.
Table 9.
An. gambiae s.l. indoor and outdoor biting by month*.
Fig 5.
An. gambiae s.l. indoor and outdoor biting patterns across study sites.
Fig 6.
Indoor and outdoor biting patterns on An. gambiae s.s, An. melas, An. coluzzii and An.arabiensis during the 2013 transmission season.
Table 10.
Entomological inoculation rates (EIR) and species composition by region.
Fig 7.
Variation An. gambiae s.l species composition across the transmission season.
Vector parity was significantly higher in URR-S and URR-N than in the other 4 regions (Table 11).
Table 11.
Vector parity by region.
Fig 8.
Relationship between parous rate (X axis) and incidence of infection and clinical malaria (Y axis).
This was confirmed by a multivariate analysis (vector parity, EIR, and species variation) in which a percentage increase in vector parity was significantly associated with the incidence of infections (IRR: 7.11; 95%CI: 6.32–8.00) (p<0.01). Sporozoite prevalence was 0.14% (14/10184), with no positive samples in NBR. The EIR was similar in LRR (4.73/person/year), URR-S (3.66/person/year) and URR-N (3.26/person/year), the regions with the highest incidence of malaria infections though CRR, with the lowest incidence of infection, had an EIR of 2.33; EIR was extremely low in WCR (0.14/person/year) and NBR (0/person/year) (Table 10).