Table 5.
Sensitivity analysis
| Outcome measure | Treatment group | Control group | Difference between groups | ||
|---|---|---|---|---|---|
| Study | N | Events (%) | N | Events (%) | OR [95 % CI]; p-value |
| Pyelonephritis | |||||
| Williams [22] Sensitivity analysis I | 106a | 5 (5) | 105a | 18 (17) | 0.24b [0.19–1.05]b; 0.004c |
| Williams [22] Sensitivity analysis II | 106a | 9 (8)d | 105a | 18 (17) | 0.45b [0.19–1.05]b; 0.066c |
CI confidence interval, CSZ convexity, symmetry, z score, OR odds ratio
aIQWiG’s own calculation: all participants included (211 instead of the 163 actually analysed); both the treatment and control group were of equal size
bIQWiG’s own calculation
cIQWiG’s own calculation, unconditioned exact test (CSZ method as described in [25])
dThe level of significance would not be missed until at least four additional events occurred in the treatment group while no additional events occurred in the control group (assumption)