Table 3. Outcomes by study period.
| Variable | 1997–2001 (n = 285) |
2002–2006 (n = 340) |
2007–2011 (n = 291) |
2012–2016 (n = 204) |
P-value for trend |
|---|---|---|---|---|---|
| 30-day mortality, n (%) | 24 (8) | 16 (5) | 24 (8) | 20 (10) | 0.33 |
| Length of hospital stay, median (IQR), days | 8 (5; 11) | 8 (5; 12.5) | 9 (6; 12) | 8 (6; 13) | 0.070 |
| ICU admission, n (%) | 57 (20) | 78 (23) | 100 (34) | 67 (33) | <0.001 |
| ICU mortality, n (%)a | 15 (26) | 8 (10) | 11 (11) | 7 (10) | 0.023 |
| Length of stay in ICU, median (IQR), daysa | 12 (8; 24) | 14 (9; 27) | 11 (8; 17.5) | 12 (8; 19) | 0.20 |
| Mechanical ventilation, n (%)b | 0.78 | ||||
| Not ventilated | 249 (87) | 292 (89) | 217 (84) | 139 (86) | 0.31 |
| Non-invasive | 0 | 14 (4) | 17 (7) | 11 (7) | <0.001 |
| Invasive | 36 (13) | 23 (7) | 26 (10) | 11 (7) | 0.11 |
Abbreviations: ICU, intensive care unit; IQR, interquartile range. Percentages calculated on non-missing data.
a 57 patients in the 1997–2001 period, 78 patients in the 2002–2006 period, 100 in the 2007–2011 period, and 67 patients in the 2012–2016 period were used to calculate the percentages and the medians (IQR).
b Patients who initially received non-invasive ventilation but subsequently needed intubation were included in the invasive mechanical ventilation group.