This observational study was designed to assess whether routine measur
ement of outcomes using standard instruments is possible as part of cl
inical routine in care of the elderly, and to establish if such instru
ments are responsive to clinical change. Indices of functional status,
cognitive function and subjective health status were collected routin
ely on admission and discharge in 540 inpatients and 340 patients atte
nding a day hospital. Data collection became integrated into clinical
routine. Response rates were generally good and yielded acceptably com
plete data. For inpatients, outcome was reflected by measurement of su
rvival, physical function (Barthel index) and social status, each of t
hese indicators showing significant change between admission and disch
arge. For day hospital patients, neither these, nor the Nottingham ADL
scale, nor a health status indicator proved sufficiently responsive t
o clinical change to merit recommendation as outcome indicators for ro
utine use in older patients. In ambulatory care in older patients, suc
h as those attending a day hospital, new approaches are needed to meas
ure clinical outcomes.