Objective To explore the differences between large and small studies in rare events meta-analysis. Methods Empirical data were collected from The Cochrane Systematic Review Database from January 2003 to May 2018. Meta-analyses with rare events, binary outcomes involving at least 5 studies, and at least 1 large study were screened. Peto and classical ORs were used to compare the magnitude, direction and P-value. Results A total of 214 meta-analyses were included. Among 214 pairs of ORs of large and small studies, 66 pairs (30.84%) were inconsistent in the direction of ORs based on Peto OR (Kappa =0.33), and 69 pairs (32.24%) were inconsistent in the direction of ORs based on classical OR. The Peto ORs resulted in smaller P-values compared to classic ORs in a substantial (83.18%) number of cases. Conclusion There are considerable differences between large and small studies in the results of meta-analysis of rare events.
Causal inference is one of the main goals of medical research. However, due to the lack of an in-depth understanding of the theory of causal inference, researchers tend to blindly use multiple statistical methods to analyse the same question to enhance the credibility of the results, which leads to problems in interpretation of the analysis results. Based on the three basic concepts of potential outcomes, causal effects, and distributive mechanisms of the causal inference counterfactual framework, this paper introduced six main target effects in causal inference and discussed their comparability to help researchers understand the principle of causal inference and correctly interpret and compare research results to avoid misleading conclusions.
Objective To evaluate the robustness of cardiovascular meta-analysis with use of fragility index. Methods By searching PubMed, EMbase, and Web of Science databases from 2018 to 2022, relevant literature on cardiovascular meta-analysis was systematically collected and the fragility indexes were calculated; Spearman correlation analysis was used to explore the relationship between fragility index and sample size, total number of events, effect size and its confidence interval width. Results A total of 212 meta-analyses from 29 articles were included, with a median fragility index of 11 (5, 25), a median sample size of 10301 (3384, 48330), and a median total number of events of 360 (129, 1309). Most meta-analyses chose relative risk as the effect measure (179/212), and chose Mantel-Haenszel method (102/212) and random effects model (153/212). The fragility index was positively correlated with the sample size (rs=0.56, P<0.05) and the total number of events (rs=0.61, P<0.05), and negatively correlated with confidence interval width of the effect size (rs=−0.52, P<0.05). No statistically significant results were obtained in the correlation between the fragility index and effect size. Conclusion The fragility indexes of cardiovascular meta-analyses published in comprehensive journals of high impact factors and professional cardiovascular journals are generally low, and therefore lack robustness. Fragility index is suggested to be reported in medical researches, assisting in explaining the P-value.
Objective To evaluate the reporting quality and influencing factors of patient-reported outcome (PRO) data in lung cancer randomized controlled trials (RCTs) from 2010 to 2024. Methods RCTs of lung cancer with PRO as either primary or secondary endpoints were searched from PubMed, EMbase, Medline, CNKI (China National Knowledge Infrastructure), Wanfang Data Knowledge Service Platform, and VIP Chinese Journal Service Platform between January 1, 2010 and April 20, 2024. Reporting quality of included RCT were assessed based on the CONSORT-PRO extension. Descriptive statistics and bivariate regression analysis were used to describe the reporting quality and analyze the factors influencing the reporting quality. Results A total of 740 articles were retrieved. After screening, 53 eligible lung cancer RCTs with 22 780 patients were included. The patients mainly were non-small cell lung cancer (84.91%), with the median sample size was 364 (160.50, 599.50) patients. The primary PRO tool used was the EORTC QLQ-C30 (60.38%). There were 52 studies (98.11%) whose PRO measured the domain of "symptom management of cough, dyspnea, fatigue, pain, etc.", and 45 studies (84.91%) measured "health-related quality of life." Multicenter studies accounted for 84.91%, and randomized non-blind trials accounted for 62.26%. PRO was used as the primary endpoint in 33.96% of the studies and as secondary endpoints in 66.04%. The reliability and validity of the PRO tools were explicitly mentioned in 11.32% and 7.55% of the studies, respectively. The average completeness of reporting according to the CONSORT-PRO guidelines was 60.00%, ranging from 25% to 93%. The main factors affecting the completeness of CONSORT-PRO reporting included sample size and publication year. For each additional sample size, the completeness of reporting increased by 27.5% (SE=0.000, t=2.04, P=0.046). Additionally, studies published after 2019 had a 67.2% higher completeness of reporting compared to those published in or before 2019 (SE=0.178, t=–3.273, P=0.006). Conclusion The study reveals that the overall reporting quality of PRO in lung cancer RCTs is poor. Particularly, the reporting of patient reported outcome measures reliability and validity, PRO assumptions, applicability, and handling of missing data needs further improvement. Future research should emphasize comprehensive adherence to the CONSORT-PRO guidelines.