Study
Study design	Source of data (e.g. cohort, case‐control, randomised trial, or registry data)
	Dates
Participants	Participant eligibility and recruitment method (e.g. consecutive participants, location, number of centres, setting, inclusion and exclusion criteria)
	Participant description
	Details of treatment received, if relevant
Outcomes to be predicted	Definition of outcome
	Method of measurement
	Time of outcome occurrence
Prognostic factors (index and comparator)	Type of prognostic factors
	Definition and method of measurement for prognostic factors
	Timing of prognostic factor measurement
	Handling of prognostic factors in the analysis
Sample size	Was a sample size calculation conducted, and if so, how?
	Number of participants
	Number of outcomes
	Number of outcomes in relation to number of candidate prognostic factors (outcomes per variable)
Missing data	Number of participants with missing data for each prognostic factor of interest
	Details of attrition and, for time‐to‐event outcomes, number of censored observations
	Handling of missing data
Analysis	Modelling method of analysis
	How modelling assumptions were checked: in particular, for time‐to‐event outcomes and the analysis of hazard ratios, the method for assessing non‐proportional hazards (non‐constant hazard ratios over time)
	Method for selection of prognostic factors for inclusion in multivariable modelling (e.g. all candidate prognostic factors considered, preselection of established prognostic factors, retain only those significant from univariable analysis)
	Method for selection or exclusion of prognostic factors (including those of interest and those used as adjustment factors) during multivariable modelling (e.g. backward or forward selection, or full model approach including all factors regardless) and criteria used for any selection or exclusion (e.g. P value, Akaike information criterion)
Results	Unadjusted and adjusted prognostic effect estimates (e.g. risk ratios, odds ratios, hazard ratios, mean differences) for each prognostic factor of interest, and the corresponding 95% confidence interval (or variance or standard error)
	For each extracted adjusted prognostic effect estimate of interest, the set of adjustment factors used