Page 62 - ESHRE2019

5. Incomplete outcome data
1 “115/120 women randomised were included in analysis”
“Because dropout was small (9%) and balanced, the primary analysis was conducted for those who completed the trial”
All the patients enrolled in the study completed the follow-up,
3 without any adverse events. -
In a trial of metformin vs. diet plan for PCOS. 200 women were
randomised. 50/100 women in the metformin group dropped out
4 due to side effects and another 20/100 dropped out as wanted to
try other treatments. 30/100 women in the diet plan group dropped out as wanted to try other treatments.
“Among women who completed follow-up, live birth occurred in 35% of the intervention group but only 18% of controls”
2
+?
5 6
Sources of bias
?
100 women were randomised, 50 to each group. The pregnancy rate ? in group A was 10% and the pregnancy rate in group B was 25%.
Selection Performance Detection Attrition Reporting
Target population Allocation
Intervention group
Outcome assessment
Control group
Outcome assessment
Selective reporting
Publication of study outcomes
60
57