Diagnostic accuracy for colorectal cancer of a quantitative faecal immunochemical test in symptomatic primary care patients: a study protocol

Anna Lööv; Cecilia Högberg; Mikael Lilja; Elvar Theodorsson; Per Hellström; Alexandra Metsini; Louise Olsson

doi:10.1186/s41512-022-00129-7

Diagnostic accuracy for colorectal cancer of a quantitative faecal immunochemical test in symptomatic primary care patients: a study protocol

Diagn Progn Res. 2022 Aug 18;6(1):16. doi: 10.1186/s41512-022-00129-7.

Authors

Anna Lööv^{1

2}, Cecilia Högberg³, Mikael Lilja³, Elvar Theodorsson⁴, Per Hellström⁵, Alexandra Metsini^{6

7}, Louise Olsson^{6

8}

Affiliations

¹ Department of Medical Sciences, Örebro University, Örebro, Sweden. anna.loov@oru.se.
² Skebäck Primary Care Centre, Region Örebro län, Örebro, Sweden. anna.loov@oru.se.
³ Department of Public Health and Clinical Medicine, Unit of Research, Education and Development Östersund, Umeå University, Umeå, Sweden.
⁴ Department of Biomedical and Clinical Science; Clinical Chemistry, Linköping University, Linköping, Sweden.
⁵ Department of Medical Sciences, Uppsala University, Uppsala, Sweden.
⁶ Department of Medical Sciences, Örebro University, Örebro, Sweden.
⁷ Department of Knowledge Management and Patient Safety Unit, Region Värmland, Karlstad, Sweden.
⁸ Centre for Assessment of Medical Technology, Örebro University Hospital, Örebro, Sweden.

Abstract

Background: There is increasing evidence supporting the use of faecal immunochemical tests (FIT) in patients reporting symptoms associated with colorectal cancer (CRC), but most studies until now have focused on selected subjects already referred for investigation. We therefore set out to determine the accuracy and predictive values of FIT in a primary care population.

Method: A prospective, multicentre, single-gated comparative diagnostic study on quantitative FIT in patients aged 40 years and above presenting in primary care with symptoms associated with CRC will be conducted. Patients representing the whole spectrum of severity of such symptoms met with in primary care will be eligible and identified by GPs. Participants will answer a short form on symptoms during the last month. They will provide two faecal samples from two separate days. Analyses will be performed within 5 days (QuikRead go®, Aidian Oy). The analytical working range is 10-200 μg Hb/g faeces. Reference test will be linked to the Swedish Colorectal Cancer Registry up to 2 years after inclusion. Accuracy, area under ROC curves, and predictive values will be calculated for one FIT compared to the highest value of two FIT and at cutoff < 10, 10-14.9, 15-19.9 and ≥ 20 μg Hb/g faeces. Subgroup analyses will be conducted for patients with anaemia and those reporting rectal bleeding. A model-based cost-effectiveness analysis based on the clinical accuracy study will be performed. Based on previous literature, we hypothesized that the sensitivity of the highest value of two FIT at cutoff 10 μg Hb/g faeces will be 95% (95% CI + / - 15%). The prevalence of CRC in the study population was estimated to be 2%, and the rate of non-responders to be 1/6. In all, 3000 patients will be invited at 30 primary care centres.

Discussion: This study will generate important clinical real-life structured data on accuracy and predictive values of FIT in the most critical population for work-up of CRC, i.e. patients presenting with at times ambiguous symptoms in primary care. It will help establish the role of FIT in this large group.

Trial registration: NCT05156307 . Registered on 14 December 2021-retrospectively registered.

Keywords: Colorectal cancer; Diagnostic accuracy study; Primary care; Quantitative faecal immunochemical test; Sensitivity.

Associated data

ClinicalTrials.gov/NCT05156307