Aim. To ascertain the reliability of notification of tuberculosis in O
tago; to establish a more accurate incidence rate for cases of tubercu
losis in Otago.Methods. Official notification information 1985-93 was
cross checked against other record systems. Results. In the 1985-90 pe
riod the official notifications were underestimated by 33% (28/84) and
during 1991-2 by 48.5% (17/35). The crude annual incidence rates for
all cases were 7.3 per 100 000 for the 1985-90 period and 9.1 per 100
000 for the 1991-3. Conclusion. The average annual crude incidence rat
es for tuberculosis were considerably higher than the official rates.
Tuberculosis remains an important clinical and public health issue. Im
provements in notification locally and nationally are required if the
impact of the projected increase in tuberculosis is to be minimised.