Current methods of contracting for patient treatment in the health ser
vice are not suitable for elderly patients with multiple pathology and
functional impairment. This study examines Resource Utilization Group
s (RUGs) for use in geriatric medicine in England and Wales, based on
1675 patients from 26 hospitals in eight health districts in England a
nd Wales. Nurses completed a questionnaire on the clinical characteris
tics of patients required to allocate them to one of 44 RUG groups. Nu
rse/patient contact times were recorded over a 24-hour period. Therapi
st/patient contact times were recorded for a week. Data were analysed
using analysis of variance with time as independent and RUG group as d
ependent variable. Variance explanation in excess of 45% was achieved
in acute and rehabilitation wards. In long-stay wards, variance explan
ation (23%) was no better than ADL score alone. RUG-III could form the
basis of a casemix system for geriatrics in England and Wales in acut
e and rehabilitation settings.