|Scenario: During the meeting with commissioners, you use a structured conversation guide to identify and prioritize factors which may influence the transferability of the review findings to both the review context (global) and the local context specified in the secondary question (Norway). Together, you develop a number of hypotheses. However, the commissioners and expert group prioritize the following variables as potentially influencing transferability: (1) length of homelessness of participants at baseline; (2) the quality and comprehensiveness of usual housing services in the study context; and (3) climate (weather conditions) in the study context. The group agrees that these factors may influence the transferability of review findings – for example, individuals who have been homeless over longer periods of time are thought to respond less to interventions than those who have short, intermittent periods of homelessness. In addition, they hypothesise that an intervention may have a relatively smaller effect in a setting with high quality and comprehensive usual housing services compared to where the intervention is introduced in a setting with low quality usual services. Finally, they suggest that intervention participants in settings with a cold climate may consistently stay longer in stable housing when it is offered as part of an intervention due to climate rather than the intervention. The relationships between the above mentioned variables and the effect of the intervention are considered by the review authors to be hypotheses and treated as such.|
Following the meeting, you search for any evidence to support the hypotheses that the identified factors may influence transferability of the review findings. Evidence is found to support two of these hypotheses, and the third factor (climate) is included despite a lack of evidence supporting its influence on the effect of housing programmes. These factors are then listed in the protocol as explanatory factors on which subgroup analyses could be undertaken.