That said, systematic reviewers are usually familiar with their field and a priori aware of potentially eligible clinical trials and/or how the outcome in question is frequently measured. Complete prespecification also could improve efficiency in data abstraction and analysis during a systematic review. We assume that primary outcomes for both clinical trials and systematic reviews are chosen based on perceived clinical importance and/or importance to patients; and that they are usually measured and reported more thoroughly than nonprimary outcomes. Not surprisingly, in our study, primary outcomes were more completely specified than other outcome types. Our estimate of 94.7% protocols pre-specifying a primary outcome is somewhat higher than the 88% that has been reported as pre-specified in clinical trial protocols, and this could be related to the fact that we were examining protocols entered into software that requests the domain names of the pre-specified outcomes. In our study, the most incompletely pre-specified outcome was quality-of-life, a key patient-important outcome. This finding is concordant with other studies that have found that outcome reporting in clinical trials is a bigger problem for patient-important outcomes than other types of outcomes. Further, when patient-important outcomes are not primary outcomes in clinical trials, the likelihood that reporting is complete is further reduced. Our study aimed to evaluate the completeness and comparability of all outcomes, both patient-important and not. Our recommendation is that systematic reviewers should engage in discussion about and strongly consider pre-specifying all five elements of each outcome they wish to examine. When explicit pre-specification of all five elements of a given outcome is not possible, for example when all possible DAPT options for a given outcome element are not known or are too numerous, the systematic reviewers should enumerate all known acceptable options for each element and explicitly state that all options for that element would be accepted, or provide rationale for why it is impossible to completely pre-specify an element. The Preferred Reporting Items for Systematic Reviews and Meta-analyses Protocols is currently under development. We hope that the availability of reporting guidelines will improve the completeness of specification of outcomes. Assuming that the Cochrane Collaboration recognizes the importance of completeness of pre-specification, there are some possible ways to ensure that review authors are aware of the five elements of a completely specified outcome. First, editorial teams at Cochrane Review Groups should make all review authors aware of the five outcome elements early in the process. Second, peer reviewers should be directed to consider whether the outcomes are completely pre-specified and not likely to have been chosen based on the strength and direction of the findings for those outcomes. Third, the Cochrane Handbook and other systematic review guidance materials, in addition to training workshops and other educational avenues, should incorporate explicit descriptions of all five outcome elements. Other organizations producing guidance on systematic review methodology should also incorporate descriptions of the five outcome elements in their guidance materials. Organizations such as the Cochrane Collaboration suggest limiting the number of outcomes examined in a systematic review. However, in order to evaluate whether the effect of an intervention persists over time, an otherwise identical outcome is often measured at a number of time-points. For the purpose of counting the number of outcomes measured.