A meta-analysis of colorectal cancer screening with fecal occult blood

Tags: Oral
Fardy JM, MacIntosh DG

Introduction/Objective: The main objective of this study was to determine if the literature provides evidence that fecal occult blood testing reduces mortality from colorectal cancer. A secondary objective was to perform an economic analysis based on the data provided by the meta-analysis.

Methods: A meta-analysis was performed using randomized controlled trials in which screening for colorectal cancer with Hemoocult tests was compared to no screening. The outcome of interest was mortality from colorectal cancer. Secondary outcomes included the pathologic stage of identified colorectal cancer cases. Data was abstracted independently by the two investigators. Odds ratios (OR) were calculated by the Mantel-Haenszel technique.

Results: Three fully published randomised trials were identified. Pooling the identified trials yielded an odds ratio of death from colorectal cancer in the screened group compared to the placebo group of 0.83 (95% CL 0.76, 0.92) in favour of the screened group. The odds ratio of early pathologic staging (Duke's A) versus unresectable disease (Duke's C or D) was 1.73 (95% CL 1.48, 2.01) in favour of the screened group. Subgroup analysis of trials using non-dehydrated slides yielded a pooled OR of 0.84 (95% CL 0.75, 0.95) compared to 0.81 (95% CL 0.67, 0.98) for trials using rehydrated slides. Despite similar efficacy, trials using non-rehydrated slides had a much lower rate of colonoscopy (25/1000 screened) versus trials using rehydrated slides (400/1000 screened).

Discussion: Given the equivalent efficacy and lower rate of colonoscopy, it appears that hemoccult screening with non-hydrated slides may be feasible. Economic analysis suggests this may be as cost-effective as other accepted screening programs.