Table 1. Maximum correlation coefficient and time lag with time series of influenza-like illness and input variables.
Type | Variables | Correlation coefficient | Time lag (Weeks) |
---|---|---|---|
Total INF | Argentina* | 0.717 | -30 |
Ireland | 0.712 | 1 | |
China | 0.699 | 2 | |
Chile | 0.694 | 25 | |
Spain | 0.692 | 1 | |
Google Trends | Total GT (INF A + INF B)* | 0.901 | -1 |
Keyword (INF A) | 0.825 | -1 | |
Keyword (INF B) | 0.691 | -1 | |
Weather | Average vapor pressure (hPa)* | 0.607 | -27 |
Average temperature (°C) | 0.551 | -28 | |
Relative humidity (%) | 0.436 | -26 | |
Air-pollution | CO (ppm) | 0.426 | -4 |
PM10 (㎍/㎥) | 0.374 | 4 | |
SO2 (ppm) | 0.353 | -2 | |
Reference | Previous seasonal ILI | 0.577 | - |
GT, Google Trends; INF, Influenza; ILI, Influenza-like illness
* Variables with a correlation coefficient of 0.6 or more and time lag less than -1 weeks