UICC World Cancer Congress 2006

Objective: To evaluate the performance of low-dose helical CT (LDCT) screening for lung cancer in an asymptomatic population.

Methods: This is a systematic review of published research on the screening performance of LDCT from January 1990 to September 2005. Main outcome measures of the study are the performance characteristics of screening test (sensitivity, specificity, likelihood ratios, receiver operator characteristic curve).

Results: Seven studies were included with a total of 18,356 subjects. All were prospective cohort studies. Most screened populations were at increased risk of lung cancer. In the pooled analysis of prevalence screening, LDCT had a sensitivity of 81.4%(95%CI: 74.2%, 87.0%), with a specificity of 80.2%(95%CI: 79.6%, 80.8); positive likelihood ratio of 4.64 (95%CI: 2.97, 7.27) and negative likelihood ratio of 0.21 (95%CI: 0.10, 0.40); diagnostic odds ratio of 25.59 (95%CI: 16.38, 39.98). The area under the SROC curve was 0.91 (95%CI: 0.84, 0.95). There were 82.9% (95%CI: 75%-88%) of lung cancers at stage I. Similar results were seen in the repeat screening. LDCT is effective for detecting lung cancer, although this is based on an assumption of no over-diagnosis bias existing in these studies. There was significant methodological and statistical heterogeneity between studies.