Skip to main content
. 2022 May 24;26:151. doi: 10.1186/s13054-022-04016-x

Table 3.

Secondary clinical outcomes

Unadjusted PS matched PS weighted
Erythromycin (n = 235) Controls (n = 470) P value Erythromycin (n = 211) Controls (n = 211) P value Erythromycin Controls P value
30-day mortality 60 (25.9) 117 (25.4) 0.93 54 (26.0) 56 (27.3) 0.91 27.7 25.9 0.67
 ICU length of stay, days 8 [5, 14] 7 [4, 11] < 0.001 8 [5, 13] 8 [5, 13] 0.51 7 [5, 13] 7 [4, 11] 0.06
 Hospital length of stay, days 23 [13, 40] 19 [11, 36] 0.031 22 [12, 37.5] 21 [12, 39] 0.61 21 [10, 37] 20 [11, 37] 0.69
 Duration of mechanical ventilation, days* 6 [4, 11] 5 [3, 9] 0.001 6 [4, 11] 6 [3, 10] 0.48 6 [4, 10] 5 [3, 9] 0.11
 ΔSOFA day 4 −1.5 (2.9) −1.2 (2.6) 0.35 −1.5 (2.9) −1.6 (2.7) 0.79 −1.3 (2.8) −1.4 (2.7) 0.75
 Incidence of ICU-acquired infections 42 (17.9) 51 (10.9) 0.013 35 (16.6) 32 (15.2) 0.79 13.8 12.0 0.51
 Incidence of ICU-acquired AKI 17 (7.2) 24 (5.1) 0.31 16 (7.6) 15 (7.1) > 0.99 6.4 5.5 0.66
 Incidence of ICU-acquired ARDS 9 (3.8) 19 (4.0) > 0.99 8 (3.8) 13 (6.2) 0.38 3.3 4.6 0.43

AKI acute kidney injury, ARDS acute respiratory distress syndrome, ICU intensive care unit, PS propensity score, ΔSOFA change in modified sequential organ failure assessment score (excluding the neurological component) from admission to day 2, 3 or 4

Categorical data are displayed as count (percentage) or and compared using Fisher's exact test (unadjusted) or McNemar's test (after PS matching), or displayed as percentage and compared using a Chi-square test (after PS weighting)

Normally distributed continuous data are displayed as mean (standard deviation) and compared using a t-test (unadjusted), a paired t-test (after PS matching) or a t-test (after PS weighting)

Non-normally distributed continuous data are displayed as median [interquartile range] and compared using Wilcoxon's rank-sum test (unadjusted), Wilcoxon's signed-rank test or Wilcoxon rank-sum test (after PS weighting)

*In those who were mechanically ventilated at ICU admission: 225/235 (96.6%) in the erythromycin group, 419/470 (89.3%) in the control group (in the unadjusted population)

Missing in 21/235 (8.9%) in the erythromycin group and 102/470 (21.7%) in the control group, see Additional file 1: Methods and Additional file 1: Table 3 for details

For weighted samples, as provided in the survey R package