Table 1.

Items for evaluation of reporting quality

Item	Description
[S1]	Is the study's design indicated with a commonly used term in the title or the abstract? We accepted any term for study designs (such as “cohort study” or “case–control study”) used in typical study classification schemes [18].
[R1]	Is the use of routinely collected data or registry data clearly mentioned in the title or the abstract using common terms? We evaluated whether information in the title or abstract allows a reader or a database search engine to clearly recognize the use of routinely collected or registry data.
[S2]	Are the selection criteria for the analyzed participants clearly described? This was deemed adequate when the study participant selection was reported in a way that it would be clear to whom the results directly apply and for whom they would not be applicable.
[R2]	Is the coding/classification of patients clearly described with sufficient details? We deemed reporting adequate when the description of the coding or classification algorithm was sufficiently clear to allow replication of the analysis.
[S3]	Are all interventions/exposures of interest clearly described? We deemed an exposure or intervention (or risk factor, predictor, effect modifier, and so forth) sufficiently described when the provided details would allow the application of the intervention or the measurement of exposure (or risk factor and so forth) in practice. The reader should know precisely which action (e.g., prescription of a certain dose of a drug) or exposure is being assessed in the study [19].
[R3]	Is the coding/classification of the interventions/exposures clearly described with sufficient details? We deemed reporting adequate when the description of the coding or classification algorithm was sufficiently clear to allow replication of the analysis.
[S4]	Are all outcomes of interest clearly described? The outcome description was deemed adequate if it was equivalent to an outcome description in a planned prospective study designed to specifically investigate the issue (regardless whether such study would be interventional or observational, feasible or not) and if the detail given was sufficient for others to replicate the study. We did not assess if broad or specific outcomes were used, but we assessed if the reporting clearly defined the outcome and how it was measured and defined. For example, we deemed it insufficient when authors reported “we analyzed effects on hypertension” without giving a definition of hypertension (e.g., defined by more than one prescription of an antihypertensive drug within 6 months); or when authors say “we evaluated effects on mortality” without stating whether all-cause or cause-specific mortality has been investigated and without reporting the time-frame (e.g., in-hospital mortality or 30-day-after discharge mortality).
[R4]	Is the coding/classification of the outcomes clearly described with sufficient details? We deemed reporting adequate when the description of the coding or classification algorithm was sufficiently clear to allow replication of the analysis. We deemed it unnecessary for replication that all-cause mortality is operationalized with a specific code because this outcome is typically clear.
[S5]	Are the independent variables in analytic models (1) listed (or are the strategies used to create models reported)? We deemed reporting adequate when all analyzed variables (e.g., age, body weight, smoking) were listed. (2) described in sufficient detail (including categorization) to replicate the study? We deemed reporting adequate when details were provided on how the variables were included in the statistical models (e.g., age and body weight both as continuous variable and smoking as categorical variable such as “never smokers,” “previous smokers,” “smoking 1 to 10 cigarettes daily,” and “smoking more than 11 cigarettes daily”).
[R5]	Are the characteristics of the analyzed data sets clearly described, including (1) covered time period, (2) location, (3) setting, and other potentially important factors? We deemed that reporting was adequate when the covered time period, geographic location, care setting, and other potentially important factors (e.g., essential details about type of data used; decision on a case-by-case basis) were reported.
[R6]	Are the methods of linkage of databases clearly described (if applicable)?
[R7]	Are issues of data sharing clearly addressed, i.e., whether the data set is publicly available (or shared on request)? We accepted any statement regardless of how detailed it was.
[R8]	Is the validation of classification algorithms used for patients, interventions/outcomes/exposures described (if applicable)?