Xinyue Lang,1 Yanyan Zhao,2 Yingxuan Zhu,2 Lei Song,3 Chuangshi Wang,4 Duoer Wang,2 Chilie Danzeng,2 Yang Wang,2 Wei Li2 

1Department of Pharmacy and Clinical Trial Unit, Beijing Anzhen Hospital, Capital Medical University, Beijing, 100029, People’s Republic of China; 2Medical Research & Biometrics Center, National Center for Cardiovascular Diseases, The National Clinical Research Center for Cardiovascular Diseases, Fuwai Hospital, Chinese Academy of Medical Sciences & Pe-king Union Medical College, Beijing, 102308, People’s Republic of China; 3Department of Cardiology, Fuwai Hospital, National Center for Cardiovascular Diseases, Chinese Academy of Medical Sciences, Beijing, 100037, People’s Republic of China; 4National Clinical Research Center for Cardiovascular Diseases, Fuwai Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, 100037, People’s Republic of China

Correspondence: Yang Wang, Email wangyang@mrbc-nccd.com Wei Li, Email liwei@mrbc-nccd.com

Background: To assess the consequence of different degrees of missing primary endpoint data for randomized controlled trials and to find the influence factors.
Methods: PubMed, Cochrane Library, EMBASE and ClinicalTrials.gov were searched up to Nov 30, 2023. We included trials of the drug-coated balloon/drug-eluted stent with angiographic outcomes as the primary endpoint. The tipping-point analysis was used to deal with the missing data for the primary endpoint. The inconsistency rate, tipping-point standardized effect size (SES) and tipping-point ratio were used to assess the result robustness.
Results: A total of 101 trials were included, which had 109 trial comparisons. Among them, 89 (81.7%) comparisons had superior/non-inferior conclusions (H0 rejected); 85 (78.0%) comparisons had a missing rate of ≥ 10%, and 30 (27.5%) comparisons had a missing rate of ≥ 20%. For H0 rejected comparisons with a missing rate of ≥ 10%, the median of inconsistency rate, tipping-point SES and tipping-point ratio was 32.2% (IQR 19.7%, 45.4%), 0.90 (IQR 0.17, 1.79) and − 1.53 (IQR − 2.43, − 0.39). A higher missing rate and a larger (worse) observed-target SES were associated with a more unreliable result.
Conclusion: A high dropout rate and inflated target effect size could cause an unreliable result. We emphasize a robust evaluation of the results for clinical trials with missing data for the primary endpoint.
Plain Language Summary: Missing data for the primary outcome has a great impact on the interpretation of clinical trials. This study included 101 randomized controlled trials of the drug-coated balloon or drug-eluted stent with an angiographic primary endpoint and used three indicators to assess the result robustness. This study found that 78.0% trial comparisons had a missing rate of ≥ 10%, and 27.5% comparisons had a missing rate of ≥ 20%; The conclusions of some clinical trials may change after dealing with the missing data; A higher missing rate and an inflated target effect size were associated with a more unreliable result.

Keywords: missing data, tipping-point analysis, result robustness