Table 3. Univariable analysis of the association between directed wind speed covariates (time-changing and time-lagged) and time to infection of premises in the largest cluster (n = 3153), northwest of Sydney, during the 2007 equine influenza outbreak in Australia.
| Meteorological Factor | Time-lag | Term | LRT | df | P-valueb |
| Maximum daily wind | t−1 | nonlinear spline | 3.8 | 4 | 0.430 |
| speed (km hour−1) | t−2 | nonlinear spline | 9.1 | 4 | 0.058 |
| t−3 | nonlinear spline | 16.5 | 4 | 0.002 | |
| directed (k = 1)a | t−4 | nonlinear spline | 6.6 | 4 | 0.159 |
| t−5 | nonlinear spline | 3.4 | 4 | 0.499 | |
| Maximum daily wind | t−1 | nonlinear spline | 14.0 | 4 | 0.007 |
| speed (km hour−1) | t−2 | nonlinear spline | 25.3 | 4 | <0.001 |
| t−3 | nonlinear spline | 34.5 | 4 | <0.001 | |
| directed (k = 2)a | t−4 | nonlinear spline | 8.2 | 4 | 0.083 |
| t−5 | nonlinear spline | 24.6 | 4 | <0.001 | |
| Maximum daily wind | t−1 | nonlinear spline | 41.2 | 4 | <0.001 |
| speed (km hour−1) | t−2 | nonlinear spline | 49.5 | 4 | <0.001 |
| t−3 | nonlinear spline | 75.6 | 4 | <0.001 | |
| directed (k = 3)a | t−4 | nonlinear spline | 38.0 | 4 | <0.001 |
| t−5 | nonlinear spline | 52.3 | 4 | <0.001 |
a
Maximum daily wind speed (‘directed’) based on wind only from within 45° arcs centred on the direction of the k nearest infected premises for k = 1,2,3 (see Figure 2 for details) assuming that premises were infectious for 14 days and one of the nearest k infective premises was the source of infection.
b
P-values derived from likelihood ratio tests (LRT) comparing univariable to null Cox regression models.