Performance Characteristics of Fecal Immunochemical Tests for Colorectal Cancer and Advanced Adenomatous Polyps: A Systematic Review and Meta-analysis

Background Studies report inconsistent performance of fecal immunochemical tests (FITs) for colorectal cancer (CRC) and advanced adenomas. Purpose To summarize performance characteristics of FITs for CRC and advanced adenomas in average-risk persons undergoing screening colonoscopy (reference standard) and to identify factors affecting these characteristics. Data sources Ovid MEDLINE, PubMed, Embase, and the Cochrane Library from inception through October 2018; reference lists of studies and reviews. Study selection Two reviewers independently screened records to identify published English-language prospective or retrospective observational studies that evaluated FIT sensitivity and specificity for colonoscopic findings in asymptomatic, average-risk adults. Data extraction Two authors independently extracted data and evaluated study quality. Data synthesis Thirty-one studies (120 255 participants; 18 FITs) were included; all were judged to have low to moderate risk of bias. Performance characteristics depended on the threshold for a positive result. A threshold of 10 µg/g resulted in sensitivity of 0.91 (95% CI, 0.84 to 0.95) and a negative likelihood ratio of 0.10 (CI, 0.06 to 0.19) for CRC, whereas a threshold of greater than 20 µg/g resulted in specificity of 0.95 (CI, 0.94 to 0.96) and a positive likelihood ratio of 15.49 (CI, 9.82 to 22.39). For advanced adenomas, sensitivity was 0.40 (CI, 0.33 to 0.47) and the negative likelihood ratio was 0.67 (CI, 0.57 to 0.78) at 10 µg/g, and specificity was 0.95 (CI, 0.94 to 0.96) and the positive likelihood ratio was 5.86 (CI, 3.77 to 8.97) at greater than 20 µg/g. Studies had low to high heterogeneity, depending on the threshold. Although several FITs had adequate performance, sensitivity and specificity for CRC for 1 qualitative FIT were 0.90 and 0.91, respectively, at its single threshold of 10 µg/g; positive and negative likelihood ratios were 10.13 and 0.11, respectively. Comparison of 3 FITs at 3 thresholds was inconclusive: CIs overlapped, and the comparisons were across rather than within studies. Limitations Only English-language studies were included. Incomplete reporting limited quality assessment of some evidence. Performance characteristics are for 1-time rather than serial testing. Conclusion Single-application FITs have moderate to high sensitivity and specificity for CRC, depending on the positivity threshold. Sensitivity of 1-time testing for advanced adenomas is low, regardless of the threshold. Primary funding source Department of Medicine, Indiana University School of Medicine.