PASSReviewer 1· 90% conf
Statistical tests, effect sizes, exact p-values, and data presentation are adequate; the only omission is the name of the statistical software used.
Evidence
direct quote[Methods, Statistical analysis]
“Consequently, Fisher’s exact test was used to estimate the P value and confidence interval of the odds ratio, instead of using a multivariable logistic regression model, as specified in the analysis plan.”direct quote[Table 2, Table 3]
“Odds ratio 0.20, 95% confidence interval 0.06 to 0.70, P=0.006”absence[Methods, Statistical analysis]
No statistical software is named.WARNReviewer 2· 85% conf
Statistical tests are named, exact p-values are mostly reported, and data presentation (Table 3) is clear. However, test assumptions (normality, equal variance) are not verified, effect sizes with CIs are not consistently reported for all secondary outcomes, and the software is not identified.
Evidence
direct quote[Results, Table 2 and Table 3]
“Odds ratio 0.20, 95% confidence interval 0.06 to 0.70, P=0.006”absence[Methods, Statistical analysis]
Assumptions for statistical tests are not verified.direct quote[Methods, Statistical analysis]
“Consequently, Fisher’s exact test was used to estimate the P value and confidence interval of the odds ratio, instead of using a multivariable logistic regression model, as specified in the analysis plan.”PASSReviewer 3· 93% conf
Statistical tests are named, exact p-values are reported, and the primary analysis used Fisher's exact test appropriately. The data presentation is adequate for a clinical trial, and no arithmetic impossibilities were identified.
Evidence
direct quote[Table 2 and Results]
“Odds ratio (95% CI) | 0.20 (0.06 to 0.70) | P value | 0.006”direct quote[Table 3]
“mean difference (95% confidence interval) between groups also reported.”direct quote[Table 3]
“P value | 0.03 ... 0.007 ... <0.001”