Of the 411 patients notified with tuberculosis in Leeds during 1986 th
rough 1991, case notes were available on 406 (99%) and the diagnosis o
f mycobacterial disease was accepted in 397 (98%). Twenty-two patients
had infection with atypical mycobacteria and were excluded from analy
sis, as were children under 16 years. Comparison between the age group
s was further restricted to white patients because of a skew distribut
ion of Asian patients towards the younger age range. Ninety-six elderl
y patients (aged 65 years and older) were compared with 127 younger pa
tients. The distribution of pulmonary and extrathoracic disease was si
milar as was the incidence of positive bacteriology. Elderly people ha
d more frequent lower-zone and more frequent miliary shadowing, otherw
ise radiographic features were similar. Elderly people were nearly thr
ee times more likely to have reactions to antituberculous drugs, six t
imes more likely to die from tuberculosis and over twenty times more l
ikely than younger patients to have the diagnosis made at autopsy rath
er than during life.