Table 2.

Primary, key secondary outcomes and secondary outcomes according to the use of angiotensin II

Clinical outcomes	Angiotensin II	No angiotensin II	Effect estimate (95% CrI or CI)⁎	Probability of benefit†	Absolute difference (95% CrI)‡
Total number of patients	46	53
Failure to be discharged or death at day 28	28 (60.9%)	31 (58.5%)	0.91 (0.38–2.18)§	58.1%	2.35% (−16.69% to 21.08%)
Hospital mortality at day 28	15 (32.6%)	14 (26.4%)	0.99 (0.61–1.60)’	51.4%	5.98% (−11.57% to 23.54%)
Need for renal replacement therapy at day 28	7 (15.2%)	10 (18.9%)	0.87 (0.25–2.97)§	58.9%	−3.55% (−17.71% to 11.05%)
Median duration of mechanical ventilation at day 28, days (IQR)	13.1 (9.3–24.5)	16.1 (9.1–28.0)
In survivors	16.0 (10.6–28.0)	18.6 (9.9–28.0)	1.07 (0.56–2.07)⁎⁎	0.830††	−2.76 (−8.89 to 3.35)
Median hospital length of stay at day 28, days (IQR)	28.0 (16.2–28.0)	28.0 (20.0–28.0)
In survivors	28.0 (28.0–28.0)	28.0 (28.0–28.0)	0.82 (0.23–2.93)⁎⁎	0.760††	0.92 (−2.13 to 3.97)

CrI = credible interval; ICU = intensive care unit; IQR = interquartile range. Percentages may not total 100 because of rounding.

^⁎For the median duration of mechanical ventilation at day 28 and the median hospital length of stay at day 28, the estimate is reported with the 95% CI; for all others, it is reported with 95% CrI.

^†Defined as hazard ratio (HR) or odds ratio (OR) < 1.00.

^‡Unadjusted.

^§Effect estimate is OR (95% CrI) from a Bayesian model with a Bernoulli distribution adjusted by age, diabetes status, and oxygen saturation measured by pulse oximetry (Spo₂) at admission. Effect estimate is HR (95% CrI) from a Bayesian Cox proportional hazard model adjusted by age, diabetes status, and Spo₂ at admission.

^⁎⁎Effect estimate is the subdistribution HR (95% CrI) from an unadjusted Fine–Gray competing risk model, with death before the event as competing risk adjusted by age, diabetes status, and Spo₂ at admission.

^††P value.