. 2012 Nov 16;7(11):e49444. doi: 10.1371/journal.pone.0049444

Table 6. Main Specific Methodological Points to Consider in Planning and Reporting Case-only Studies (to be considered as a complement of the STROBE Statement).

Points to Consider	Recommendation for Planning	Recommendation for Reporting
1. Applicability of case-only designto the study objective (eventand exposure studied)	Study should be planned in order to fulfil the validity assumptions of the design.	Report that setting is valid for the design implementation.In case of unfulfilled assumptions: lack of fulfilment should be stated, impact on results should be discussed, and if possible, comparison with other design(s) performed.
2. Risk and controlperiods’ definition	Risk and control periods definition have to be based onphysiological evidence or hypotheses, referenced orvalidated by an expert group¹². Risk period should beidentical for all subjects. Sensitivity analysis on risk/controlperiods onset, end and duration should be planned.	Report the risk period definition and the justification of their characteristics. Report sensitivity analyses results varying risk/control periods’ characteristics.
	Restriction times should be implemented if appropriate.They should be identical in risk and control periods	Report rationale for introducing restriction times if any.
3. Statistical issues
Statistical model	Conditional logistic regression for CC and Poissonregression for SCCS must be applied as recommendedby the developers.Relevant estimator of risk must be used : OR forCC designs and RR for SCCS designs	Report statistical method and estimator of risk used in the materials and methods section. Report specific estimator with confidence interval in the results section.Use adequate denomination for risk estimators in reports: OR and RR.
Sample size calculation	For CC, sample size calculation should be estimated based oncase-control method²². For SCCS, use the published samplesize formula to calculate the sample size for SCCS¹⁶.	Report the calculate sample size and all the elements necessary to reproduce the calculation
Sensitivity analyses	Sensitivity analyses must be conducted to check therobustness of particular methodology choices, particularlyrisk and control period selection choice.	Report all sensitivity analyses conducted and their results
Results reporting	N/A	Report the risk estimator and its 95% confidence interval.For CC, report count of discordant cases.For SCCS, report count of events in the different time periods.