The association between hospitalisation for pneumonia and the diagnosis of lung cancer.

Pneumonia may mask or be the first sign of lung cancer. The literature is, however, sparse. We compared the incidence of lung cancer among patients hospitalised for pneumonia with a general population. The patients (index group, N=7044) were hospitalised for pneumonia from 1988 to 2002 at Akershus University Hospital. The reference group was recruited from a general population survey (N=81,373) performed during 1984-1986 and 1995-1997. We had detailed information on smoking habits in both groups. Data on lung cancer incidence, emigration, and death were obtained from the Cancer Registry of Norway. Data were analyzed using multivariate Poisson regression. There were 177 cases of lung cancer in the index group and 608 cases in the reference group. Current smoking modified the association between pneumonia and lung cancer. Among current smokers and non-smokers the incidence density ratio (IDR) was 6.0 (95% confidence interval) (4.6-7.8) and 12.0 (8.6-16.9) in the index group compared with the references, respectively. The crude IDR decreased from 78.4 (43.8-146) 0-3 months after the inclusion to 2.4 (1.4-3.7) 6 years after inclusion. In the index group the IDR was only slightly higher for smokers than non-smokers. Patients hospitalised for pneumonia had an increased incidence of lung cancer compared with the general population. This risk persisted for 6 years after the onset of the lung infection.