Skip to main content

Table 5 The probability of flagged trials for the AEs under the same effect for all the AEs with safety issue for scenario II

From: Two-stage Bayesian hierarchical modeling for blinded and unblinded safety monitoring in randomized clinical trials

Example AEs

(True Proportion; Expected Event Rate)

Blinded: Proportion with signal for early termination

Unblinded: Termination Proportion with confirmed by unblinded data

(Pcrit1, Pcrit2)

(0.9, 0.7)

(0.9, 0.8)

(0.9, 0.9)

Pneumothorax

Induced by HBO therapya

(π = 3%; πM = 2%)

Blinded

0.28

0.27

0.28

Unblinded

0.79

0.62

0.33

Overall Rate

0.23

0.17

0.09

Signs of Pulmonary Dysfunctiona

(π = 31%; πM = 25%)

Blinded

0.55

0.54

0.54

Unblinded

0.76

0.63

0.42

Overall Rate

0.42

0.34

0.23

Pneumoniaa

(π = 48%; πM = 40%)

Blinded

0.58

0.59

0.59

Unblinded

0.75

0.65

0.44

Overall Rate

0.44

0.39

0.26

Critical decreased CPPa (< 60 mmHg)

(π = 80%; πM = 75%)

Blinded

0.55

0.55

0.56

Unblinded

0.76

0.60

0.39

Overall Rate

0.42

0.33

0.22

Critical hypotensiona (MAP< 70 mmHg)

(π = 80%; πM = 75%)

Blinded

0.56

0.56

0.54

Unblinded

0.77

0.61

0.39

Overall Rate

0.43

0.34

0.21

Seizures during HBO treatmenta

(π = 1%; πM = 1%)

Blinded

0.23

0.23

0.23

Unblinded

0.79

0.61

0.31

Overall Rate

0.19

0.14

0.07

Hypercarbia during transportationa

(PaCO2 > 45 mmHg)

(π = 13%; πM = 10%)

Blinded

0.44

0.43

0.43

Unblinded

0.77

0.62

0.40

Overall Rate

0.33

0.26

0.17

  1. (Assume the AE with a has a safety issue)