Falling prevalence of gastrointestinal malignancy in adults with iron deficiency anaemia, and identification of additional predictive clinical variables: the IDIOM V study

Objective (1) To determine whether the prevalence of gastrointestinal (GI) cancer in adults presenting with iron deficiency anaemia (IDA) is continuing to fall, and (2) to identify clinical variables with the potential to improve the accuracy of an established system for predicting cancer risk in IDA—the Iron Deficiency as an Indicator Of Malignancy (IDIOM) score. Method (1) An audit of 1253 consecutive adults to the IDA service of a single general hospital in 2022/2023, with analysis of the outcome of investigation for GI cancer, and the relationship between this outcome and a series of eight clinical variables. (2) Refinement of the IDIOM score using an imputation method to combine the development and new predictor data sets. Results The prevalence of GI cancer in those undergoing full investigation for IDA was 5.6% (52/934). On univariate logistic regression analysis, there were strong negative associations between GI cancer risk and (1) previous colonic imaging within the last 5 years (OR 0.31, 95% CI 0.08 to 0.87), and (2) long-term exposure to proton pump inhibitor (PPI) therapy (OR 0.18, 95% CI 0.09 to 0.34). Model extension analysis suggested that adding PPI exposure status could improve the accuracy of the IDIOM score: c-statistic—0.84 vs 0.77, calibration slope—1.01 vs 1.05. Conclusion The prevalence of GI cancer in adults presenting with IDA continues to fall, strengthening the case for risk stratification and targeting of invasive investigation at those at higher risk. Adding PPI exposure status may improve the value of the IDIOM score, though this requires confirmation in prospective studies.