Table 1.

Criteria for the matching of an outcome measure with the intervention, study sample, and setting

Criteria		Rating
Intervention	Did the outcome measure match an intervention content?	“Match”	The outcome measure specifically addressed the exercise intervention or an intervention content of the multidisciplinary program (e.g., 6-Meter Walking Test → treadmill walking training; discharge destination → discharge planning).
		“Limited match”	The outcome measure addressed the exercise intervention or an intervention content of the multidisciplinary program only to a limited extent and/or included only single items that specially matched to the intervention (e.g., Barthel Index [transfer, mobility, and stairs items] → strengthening and mobility exercises; physical activity monitoring → weight-bearing exercises)
		“No match”	The outcome measure did not directly address the exercise intervention or an intervention content of the multidisciplinary program or the construct of the outcome measure was not addressed in the intervention (e.g., Lawton IADL scale → no IADL training content or discharge destination → additional exercise intervention).
Study sample	Was the outcome measure feasible in the study sample?	“Match”	The outcome measure showed no floor or ceiling effects (continuous outcomes) or represented no rare event (dichotomous outcomes). Ceiling and floor effects were defined as (1) ≥ 15% of participants reaching a score within the best or worst 15% of the instrument’s rating scale [22] or (2) when the mean score of the sample was within the best or worst 15% of the rating scale. Rare events were defined when the incidence of a dichotomous outcome (e.g., falls, mortality) was ≤15% in the sample.
		“No match”	The outcome measure showed floor or ceiling effects (continuous outcomes) or represented a rare event (dichotomous outcomes).
	How high was the missing data rate for the outcome measure in the study sample?	“Match”	The outcome measure had an acceptable missing data rate. Missing data included any outcome data that (1) could not be collected for reasons other than death or study withdrawal or (2) were collected but not presented. A missing data rate of < 15% was considered as acceptable.
		“No match”	The outcome measure did not have an acceptable missing data rate (≥ 15%).
Setting	Did the outcome measure match the acute care hospital setting?	“Match”	The outcome measure addressed a construct or activities that can be appropriately assessed within the acute care hospital setting (e.g., hospital costs or Barthel Index).
		“Limited match”	The outcome measure addressed a construct or activities that can be appropriately assessed only to a limited extent within the acute care hospital setting and/or included only single items or contents that were appropriate for use within the acute care hospital setting (e.g., combined ADL-IADL measures).
		“No match”	The outcome measure addressed a construct or activities that cannot be appropriately assessed within the acute care hospital setting (e.g., IADL measures).