Criteria | Items assessed | Quality of reporting | ||
---|---|---|---|---|
Adequate | Inadequate | Not reported | ||
Estimation of sample size | Statistical power; hypothesised effect estimate; effect size; alpha level; prevalence of exposure and probability of the expected outcome | All of these items reported for prospective studies. For retrospective studies, post hoc power estimation or detectable difference | At least one item was not reported | No sample size estimation information provided |
Follow-up time | Start and exit dates and aggregate follow-up time (median/person-time) | Reported all these items | At least one item was not reported | None of the items was reported |
Survival curves | Number of patients at risk at the bottom of the graph; markings to indicate when participants were censored; axes were clearly labelled and used different colors/type of lines to distinguish curves | Reported all these items | At least one item was not reported | No survival curve was plotted |
Comparison of survival curves | Methods for group comparisons and their test results (p-values) | Reported all these items | At least one item was not reported | No comparison was done |
Reporting measures of effect in SAMs | Measures of effect and uncertainty among studies reporting regression analysis | Correct measure of effect and uncertainty reported | Incorrect measure of effect or no measure of uncertainty | No measure of effect and uncertainty reported |
Test of survival regression models underlying assumptions | Survival regression models used; statistical method used to test underlying assumptions and test results | Reported all these items | At least one of the items not reported | All the items not reported |
Analysis of hierarchical clustering | Presence of clustering; methods of investigating heterogeneity and correct method for adjustment | Test of investigating heterogeneity and correct method for adjustment where there was evidence of heterogeneity reported | At least one of the items not reported | No consideration for clustering |