Table 5.
Multivariate analysis of mosquito collection methods associated with abundance of three primary vector species by catch site
| An. dirus | An. maculatus | An. minimus | ||||||
|---|---|---|---|---|---|---|---|---|
| bpn (n) | IRRa (95% CI) | bpn (n) | IRRa (95% CI) | p value | bpn (n) | IRR (95% CI) | p value | |
| Suan Oi | ||||||||
| HLC indoor | 0 | – | 0 | – | – | 0.08 (1) | 1.37 (0.19, 9.85) | 0.76 |
| HLC outdoor | 0 | – | 0 | – | – | 0.33 (4) | 0.91 (0.33, 2.50) | 0.86 |
| Pha Man | ||||||||
| HLC indoor | 0 | – | 1.67 (20) | 1 | 5.67 (68) | 1 | ||
| HLC outdoor | 0.17 (2) | – | 2.58 (31) | 1.47 (0.84, 2.58) | 0.18 | 4.25 (51) | 0.71 (0.49, 1.02) | 0.07 |
| Farm huts | ||||||||
| HLC | 0.30 (10) | – | 4.21 (139) | 1.53 (0.96, 2.45) | 0.08 | 5.63 (186) | 0.60 (0.46, 0.80) | < 0.001 |
Multivariate analysis conducted using negative binomial regression and on the average bites per person per night (bpn). Estimates are from HLC data only, bites per night adjusted for difference in frequency of extended hours of collection (17:00–18:00 and 06:00–07:00 collections conducted for 24/33 nights)
aNot enough data to calculate IRR. IRR: estimated incident rate ratio, CI: confidence interval, corresponding p-values based on maximum likelihood estimation of suitable Negative Binomial regression models