The purpose of this study was to establish the extent of undernotifica
tion of tuberculosis in AIDS patients resident in 2 inner London local
authorities. For residents of the 2 authorities, statutory notificati
ons of tuberculosis between 1986 and 1992 were compared, using sounder
codes of surnames, sex and year of birth, with AIDS cases reported to
the Public Health Laboratory Service (PHLS) AIDS Centre during the sa
me period where TB had been recorded on the AIDS report form. In 36 of
613 AIDS cases reported as residents of the 2 authorities tuberculosi
s was recorded on the AIDS report form. Matching revealed that only 2
(6%) of these cases had been notified to the local authority. These re
sults highlight the need to resolve the dilemma between concerns about
patient confidentiality and the statutory requirement to notify tuber
culosis so that clinical management of contacts can be undertaken and
the true impact of HIV infection on the incidence of tuberculosis in t
he UK can be elucidated.