Role of low-dose spiral CT scan in early lung cancer screeing of high risk population: a systematic review of the literature with a Meta-analysis

Objective: To assess the overall diagnostic value of low-dose spiral computed tomography (LDSCT) screening trials for early lung cancer of high risk population. Methods: A search in PubMed, Cancerlit, Embase bank, Cochrane Library, CBMdisc and CNKI was performed to identify relevant English and Chinese-language articles from January 1, 1995 to June 1,2005. Quality evaluation and heterogeneity test were performed to the included researches. Consistency analysis was executed. Pooled weighted sensitivity and specificity were calculated. Receiver operating characteristic (ROC) curve was performed and area under the curve (AUC) was calculated. Test of odds ratio(OR)was performed to these results of base screening and annual follow-up screening at least two years. Finally, sensitivity analysis was performed. Results: Fifteen articles were included, with 9 independent researches and a total of 8168 subjects. Heterogeneity existed. Base screening was undertaken in high risk population in the grade of BO (discovering nodule), B5 (discovering nodule of diameter more than 5 mm), B10 (discovering nodule of diameter more than 10 mm) and corresponding Kappa value was 0.061, 0.134 and 0.460, pooled weighted sensitivity was 0.686, 0.721 and 0.640, pooled weighted specificity was 0.708, 0.815 and 0.959. Area under the ROC was 0.778, 0.949 and 0.539 respectively. Statistically significant difference did not exist between the AUC of BO and B5 (P>0.05), but existed between B0 and B10, B5 and B10 (P<0.05). With the fixed effect model, the pooled OR in the grade of B5 and B10 were 1.11 and 1.66 respectively. Conclusion: The articles entered into this Meta-analysis were of high quality. The pooled weighted sensitivity and specificity of LDSCT screening for early lung cancer of high risk population was moderate. The method was affected by chance factors easily. It was relatively reasonable to define the positive standard as the discovered nodule's diameter was more than 5 mm. Close flow-up could raise the accuracy of screening test. (authors)