Our aim was to evaluate the efficacy of a revised tuberculosis (TB) co
ntact tracing procedure in South Glamorgan whereby routine annual radi
ological surveillance was abandoned and contacts were either discharge
d or referred to chest clinic following their initial screening. We re
viewed and evaluated data from the TB contact tracing clinic, the Publ
ic Health Service Mycobacterium Reference Unit, Cardiff and the Consul
tant in Communicable Diseases Control, South Glamorgan Health Authorit
y and compared these results with those of our previous study. One hun
dred and three index cases and 732 contacts were identified. Seven hun
dred and seven contacts, 526 close and 181 casual, were screened, of w
hom 102 casuals should not have been. One hundred and sixty-one contac
ts were given BCG vaccination. Fifty-four contacts were referred to th
e chest clinic. Seven cases of TB were detected, all in young, unvacci
nated, close contacts of pulmonary disease. Twenty-one contacts were g
iven chemoprophylaxis, 20 of whom were close contacts of pulmonary TB
and one of extrapulmonary disease. Five contacts who were screened and
initially discharged developed TB later: in two the protocol had not
been followed and three presented with extrapulmonary TB. Compared wit
h the results of the previous protocol fewer contacts were unnecessari
ly screened and referrals to the chest clinic increased, as did the nu
mber given chemoprophylaxis. The case finding rate is similar to that
found prior to the revision of the protocol. The yield from tracing ca
sual contacts continues to be nil. It is very low in contacts of extra
pulmonary disease. When the protocol was followed no case of pulmonary
TB was missed. The revised protocol seems to be as effective as the p
revious, more complex protocol. In our area, one of low incidence of T
B, screening of casual contacts and of contacts of extrapulmonary TB i
s not cost-effective. We will concentrate even more on screening close
contacts of pulmonary TB.