Performance Value of Research of Occult Gastrointestinal Bleeding by Immunoassay in the Diagnostic Process of Iron Deficiency Anemia in Patients Over 75 Years

Overview

This study aims to characterize the diagnostic performance of immunological testing of occult gastrointestinal bleeding in stool in the population aged over 75 years with iron deficiency anemia. As secondary objectives, the study aims to: * determine a threshold of positivity optimizing the immunoassay performance for the study population, in accordance with the probabilities of error (false positives, false negatives) and weights (defined by expert consensus) allocated to these errors. * Assess the benefit of a double measure of bleeding (two stools) by immunoassay compared to a single measure.

Immunological testing will be performed systematically for each patient. Research of blood in fecal sample will be based on 2 methods: one based on a qualitative immunochromatographic method (such as HemSign4 technique, Servibio, France; or all other techniques used at AP-HP) and another quantitative immunoturbidimetric method (OC-Sensor, Eiken, Japan). They will be performed systematically with two fresh fecal samples for each patient, before coloscopy examination. Treating physicians will not be provided with the testing result. The diagnosis of reference will be established by experts based on the result of explorations (colonoscopy, coloscanner) and the patient's progression, blinded to the result of immunoassay. The intrinsic characteristics of the immunoassay (sensitivity, specificity...) will be determined with reference to this reference diagnosis, using a positivity threshold defined using expert consensus on the weights to be allocated to potential errors. 2 modeling scenarios (without vs with immunological testing) using decision trees based on prevalence and diagnostic performance observed and stochastic simulations.

Conditions

Eligibility