Outcome variables
Percentage attendance at the exercise groups was calculated
for each home and for each individual participant.
Percentage attendance
¼ Number of groups actually attended
Number of possible groups available to attend 100
This calculation was used because a different number
of exercise groups ran in each home and therefore,
percentage attendance allowed a standardised approach to
capture attendance.
Data on maximum possible attendance was calculated
for each home by multiplying the number of study participants
by the number of groups delivered. However, in
practice, actual ‘exposure’ to group exercise varied on an
individual level because of deaths, transfers to other institutions
or inability to attend due to illness/other reason.
Data were recorded prospectively to allow accurate
estimation of individual attendance.