Effect sizes in ongoing randomized controlled critical care trials

Table 2 Prior probability estimates and calculated chances of a true positive result for each trial

Trial name	Prior probability estimates^a (%)		Chance of a true-positive^b (%)
	Median estimate (IQR)	Range of estimates	Median calculated chance (IQR)	Range of calculated chances
A Confirmatory Phase II/III Study Assessing Efficacy, Immunogenicity and Safety of IC43 Recombinant Pseudomonas Vaccine in Intensive Care Patients	5 (0.33–20)	0–82	49 (6–82)	0–99
ADjunctive coRticosteroid trEatment iN criticAlly ilL Patients with Septic Shock	20 (5–50)	0–100	82 (49–95)	0–100
Early Spontaneous Breathing in ARDS	10 (1–25)	0–100	64 (14–84)	0–100
Non-sedation versus Sedation with a Daily Wake-up Trial in Critically Ill Patients Receiving Mechanical Ventilation	20 (5–40)	0–100	80 (46–91)	0–100
Stress Ulcer Prophylaxis in the Intensive Care Unit	10 (0.030–28.75)	0–100	67 (1–88)	0–100
The Augmented versus Routine Approach to Giving Energy Trial	10 (2–40)	0–90	64 (29–91)	0–99
The SuDDICU Cluster RCT of Antibiotic Prophylaxis in Critical Illness	25 (10–50)	0–100	86 (67–95)	0–100
Ticagrelor in Severe CAP	2.5 (0–10)	0–100	29 (0–64)	0–100
Tranexamic acid for the treatment of gastrointestinal haemorrhage: an international randomised, double-blind, placebo-controlled trial	35 (10–57.5)	0–100	91 (67–96)	0–100
Tranexamic acid for the treatment of significant traumatic brain injury: an international, randomised, double-blind, placebo-controlled trial	20 (10–50)	0–100	82 (67–95)	0–100

Abbreviations: APRV Airway pressure release ventilation, ARDS Acute respiratory distress syndrome, CAP Community-acquired pneumonia, GI Gastrointestinal, ICU Intensive care unit, IQR Interquartile range, SDD Selective digestive decontamination, RRT Renal replacement therapy
^aPrior probability was defined as the percentage chance, estimated by clinicians, that a trial would demonstrate a mortality effect equal to or greater than that used by the trials investigators in their sample size calculation
^bCalculated chances of a true-positive result are based on the prior probabilities shown and assume power (β) specified by the trial investigators and an α of 0.05 using the following calculation: True-positive rate = (prior probability) × β/(1 − prior probability) × α) + (prior probability × β)

ISSN: 1364-8535