This study evaluated the effectiveness of a new rehabilitation service, com
pared with existing services, for carers of people with traumatic brain inj
ury (TBI). Subjects were 96 adult carers of people consecutively admitted t
o two local hospitals. They were assigned to one of three groups: Early new
service (pre-discharge); Late new service (post-discharge); or a Control c
ondition (existing services only). Individual randomization was not possibl
e and randomization by hospital site was rejected because of demographic an
d clinical differences between sites. Group assignment was determined by a
pre-specified timetable which alternated between hospitals. Two outcomes we
re compared at 6 months post-injury: carers' emotional distress and how wel
l-informed they felt about TBI and available resources. Analyses adjusting
for potential confounding factors confirmed a clinically plausible superior
outcome for both intervention groups compared to the control group. Howeve
r, these differences did not obtain statistical significance (p greater tha
n or equal to 0.01). Several reasons for these findings are discussed, incl
uding the novelty of the new service and methodological issues such as stat
istical power. Longer term follow-up studies are required as psychological
sequelae and, therefore, the need for intervention is known to increase wit
h time and may not be most apparent during the first 6 months.