Table 4.
Factors associated with S pneumoniae colonisation density in the live attenuated influenza vaccine group
| Change in log10 copies per μL | SE | p value* | |
|---|---|---|---|
| Asymptomatic respiratory virus at day 0 | +0·287 | 0·122 | 0·020 |
| Day 7 (vs day 0) | +0·207 | 0·105 | 0·050 |
| Day 21 (vs day 0) | +0·280 | 0·105 | 0·0082 |
| Recruitment in 2018 (vs 2017) | −0·059 | 0·116 | 0·61 |
| Age, months | −0·019 | 0·006 | 0·0020 |
| Household cooking indoors (vs outdoors)† | −0·290 | 0·185 | 0·12 |
S pneumoniae detection was done by lytA PCR.
p values for factors associated with S pneumoniae density in the live attenuated influenza vaccine group are derived from a generalised linear mixed-effect model, taking into account changes within individuals over time. Backwards model selection was done as described in the appendix before deciding on variables to include in the optimum model. Reference levels for each variable are given within brackets.
Cooking inside (under a roof) compared with cooking using an indoor kitchen. Note, 99% of children lived in households in which wood or charcoal was the primary fuel used to cook regardless of location of cooking.