Anton Gies1, Laura Fiona Gruner2, Petra Schrotz-King3, Hermann Brenner4. 1. Division of Preventive Oncology, German Cancer Research Center, and National Center for Tumor Diseases, Heidelberg, Germany; Medical Faculty Heidelberg, University of Heidelberg, Heidelberg, Germany. 2. Medical Faculty Heidelberg, University of Heidelberg, Heidelberg, Germany; Division of Clinical Epidemiology and Aging Research, German Cancer Research Center, and National Center for Tumor Diseases, Heidelberg, Germany. 3. Division of Preventive Oncology, German Cancer Research Center, and National Center for Tumor Diseases, Heidelberg, Germany. 4. Division of Preventive Oncology, German Cancer Research Center, and National Center for Tumor Diseases, Heidelberg, Germany; Division of Clinical Epidemiology and Aging Research, German Cancer Research Center, and National Center for Tumor Diseases, Heidelberg, Germany; German Cancer Consortium, German Cancer Research Center, and National Center for Tumor Diseases, Heidelberg, Germany. Electronic address: h.brenner@dkfz.de.
Abstract
BACKGROUND & AIMS: Fecal immunochemical tests (FITs) measure hemoglobin in stool to identify individuals at risk for colorectal cancer (CRC). However, there are many different FITs, with different instructions for fecal sample collection. In routine practice, participants do not always follow these instructions exactly. We assessed the effects of violations of fecal sampling instructions on the diagnostic performance of 9 quantitative FITs. METHODS: We obtained stool samples from 76 patients with CRC scheduled for surgery at 4 hospitals in Germany and 100 participants without advanced neoplasms who participated in a prospective colonoscopy screening program. We filled fecal sample tubes according to the manufacturers' instructions or with 3 violations that are likely to occur in routine practice. The diagnostic performance was assessed for a total of 6336 FIT samples (176 participants × 9 FITs × 4 sampling methods). RESULTS: Sample collection instructions varied widely among FITs but included 3 key components: multiple insertions of the sampling rod into stool, a visual check of rod for complete filling with stool, and once-only insertion of the stool-filled rod into the tube. Violation of the first 2 components (inserting the rod into the stool sample only 1 time or not visually checking the rod for complete filling) reduced levels of hemoglobin measured. However, the effect on diagnostic performance was generally small. Violation of the third component (insertion of more stool into the tube than recommended) increased levels of hemoglobin measured in samples and identified more patients with CRC (increase of median sensitivity by almost 10% units) at a small loss of specificity (decrease of median specificity by 2% units), and produced the highest area under the curve for detection of CRC cases for 6 FITs. CONCLUSIONS: Violations of fecal sampling instructions can lead to non-negligible variations in fecal hemoglobin measurements. The limited adverse effects on diagnostic performance indicate the robustness of FITs. The potential for increasing diagnostic performance by collecting more stool material should be followed up in further research.
BACKGROUND & AIMS: Fecal immunochemical tests (FITs) measure hemoglobin in stool to identify individuals at risk for colorectal cancer (CRC). However, there are many different FITs, with different instructions for fecal sample collection. In routine practice, participants do not always follow these instructions exactly. We assessed the effects of violations of fecal sampling instructions on the diagnostic performance of 9 quantitative FITs. METHODS: We obtained stool samples from 76 patients with CRC scheduled for surgery at 4 hospitals in Germany and 100 participants without advanced neoplasms who participated in a prospective colonoscopy screening program. We filled fecal sample tubes according to the manufacturers' instructions or with 3 violations that are likely to occur in routine practice. The diagnostic performance was assessed for a total of 6336 FIT samples (176 participants × 9 FITs × 4 sampling methods). RESULTS: Sample collection instructions varied widely among FITs but included 3 key components: multiple insertions of the sampling rod into stool, a visual check of rod for complete filling with stool, and once-only insertion of the stool-filled rod into the tube. Violation of the first 2 components (inserting the rod into the stool sample only 1 time or not visually checking the rod for complete filling) reduced levels of hemoglobin measured. However, the effect on diagnostic performance was generally small. Violation of the third component (insertion of more stool into the tube than recommended) increased levels of hemoglobin measured in samples and identified more patients with CRC (increase of median sensitivity by almost 10% units) at a small loss of specificity (decrease of median specificity by 2% units), and produced the highest area under the curve for detection of CRC cases for 6 FITs. CONCLUSIONS: Violations of fecal sampling instructions can lead to non-negligible variations in fecal hemoglobin measurements. The limited adverse effects on diagnostic performance indicate the robustness of FITs. The potential for increasing diagnostic performance by collecting more stool material should be followed up in further research.
Authors: Julia Butt; William J Blot; Kala Visvanathan; Loïc Le Marchand; Lynne R Wilkens; Yu Chen; Howard D Sesso; Lauren Teras; Marc D Ryser; Terry Hyslop; Sylvia Wassertheil-Smoller; Lesley F Tinker; John D Potter; Mingyang Song; Sonja I Berndt; Tim Waterboer; Michael Pawlita; Meira Epplein Journal: Cancer Epidemiol Biomarkers Prev Date: 2020-09-24 Impact factor: 4.254
Authors: Anton Gies; Tobias Niedermaier; Laura Fiona Gruner; Thomas Heisser; Petra Schrotz-King; Hermann Brenner Journal: Cancers (Basel) Date: 2021-02-05 Impact factor: 6.639