Table 1 Parameter estimates of Influenza-Like Illness (ILI) risk in a university hospital by 2 inference methods
Maximum likelihood | Bayesian inference | |||
|---|---|---|---|---|
Mean | 95% confidence interval | Mean | 95% credible interval | |
ILI risk associated with unobserved exposure (per 10,000 patients-days) | ||||
No epidemic in the community | 0.8 | 0.3-1.6 | 0.8 | 0.3-1.6 |
Epidemic in the community | 4.3 | 0.4-11.0 | 4.3 | 0.4-11.0 |
ILI risk associated with observed exposure (per 10,000 patients-days) | 129.1 | 84.5-182.9 | 129.1 | 84.5-182.9 |
Distribution of risk according to lags | ||||
1 day | 82 | 51-95 | 72 | 52-89 |
2 days | 1 | 0-98 | 6 | 0-21 |
3 days | 15 | 0-81 | 12 | 1-32 |
4 to 5 days | 2 | - * | 10 | 0-27 |
Total | 100 | - | 100 | - |
Distribution of risk according to the source’s contagious period | ||||
1 day before the onset of source symptoms | 47 | 24-71 | 41 | 21-62 |
1st day after the onset of source symptoms | 17 | 6-42 | 19 | 4-39 |
2nd day after the onset of source symptoms | 24 | 8-51 | 25 | 9-44 |
3rd to 5th days after the onset of source symptoms | 12 | - * | 14 | 4-29 |
Total | 100 | - | 100 | - |