In this paper the difficulties of conducting a controlled evaluation o
f counselling (brief psychotherapy) in general practice are discussed.
Results of a pilot study indicate that patients referred by family do
ctors to counsellors are often seriously emotionally distressed and re
covery is slow. Counsellors come from different backgrounds and use a
variety of therapies. Although the results show that controlled resear
ch is feasible, in a definitive trial patients should be randomized in
a stratified manner, according to severity, by the researcher after i
nitial assessments have been made. Counsellors should have a recognize
d accreditation and preferably be employed for the trial to ensure uni
formity of approach and avoid long waiting lists. Blind assessments of
outcome are desirable but are not always feasible and reliance on pat
ient self-report is important. Within the limitations of current knowl
edge, only controlled evaluations will provide a greater understanding
of the efficacy of counselling in general practice.