Diagnostic yield of gastrointestinal endoscopy for investigation of iron deficiency anemia, iron deficiency without anemia, and non-iron deficiency anemia.

Abstract

Background and Aim: Guidelines recommend routine gastrointestinal endoscopy to investigate the cause of iron deficiency anemia (IDA). However, gastrointestinal endoscopy is also often performed to investigate iron deficiency in the absence of anemia (ID) and non-iron deficiency anemia (non-ID anemia), with unclear recommendations from expert bodies and few studies comparing the diagnostic yield for these indications. Furthermore, demand for endoscopy continues to increase, along with challenges in completing the procedures within the timeframe of their urgency category. The aim of this study was to compare the diagnostic yield of gastrointestinal endoscopy for IDA, ID, and non-ID anemia found during elective endoscopy and to identify factors contributing to cancer or other significant findings, to assist with risk stratification. Method(s): A 12-month retrospective single-center cohort study was undertaken in 2019 (before the COVID-19 era) at a large Australian quaternary health service. The study included all adults undergoing elective gastroscopy and colonoscopy for investigation of iron deficiency and/or anemia, with pathology tests performed within 12 months before endoscopy. Anemia was defined according to the World Health Organization as hemoglobin (Hb) level < 130 g/L for men and < 120 g/L for women, and iron deficiency was defined as a ferritin level < 45 mug/L. For descriptive statistics, continuous variables were expressed as median and IQR, and categorical variables as absolute frequencies. Analysis of factors associated with cancer or other significant endoscopic findings was performed using the Mann-Whitney U test, Fisher's exact test, or univariate regression. Result(s): A total of 296 patients were included in the analysis. Of these, 142 (48.0%) had IDA, 98 (33.1%) had ID, and 56 (18.9%) had non-ID anemia. There was a greater proportion of women in the ID group (62.2%), compared with the IDA (46.5%, P = 0.018) and non-ID anemia (42.9%, P = 0.028) groups. The ID group were also significantly younger (median age, 61 years; IQR, 48-71) compared with the IDA group (median, 67 years; IQR, 51-77 years; P = 0.014) but not when compared with the non-ID anemia group (median, 64 years; IQR, 51-73 years; P = 0.313). Chronic kidney disease was most prevalent in the non-ID anemia cohort (48.2% vs 21.2% for IDA and 17.6% for ID; P < 0.001) and a significant cause of anemia. The non-ID anemia group were less likely to be completely asymptomatic (23.2%), compared with the IDA (41.5%, P = 0.022) and ID (43.9%, P = 0.014) groups. There were 11 cancers (3.7%) detected overall, with a trend towards a greater proportion of cancers in the IDA than the ID group (5.6% vs 1.0%, P = 0.061). There were no statistically significant differences between the IDA, ID, and non-ID anemia groups for significant findings on gastroscopy (37.3%, 37.8%, and 35.7%, respectively; P = 0.178), colonoscopy (43.0%, 45.9%, and 53.6%, respectively; P = 0.402) or overall during endoscopy (62.0%, 63.3%, and 66.1%, respectively; P = 0.865). Increasing age was a statistically significant factor associated with cancer (odds ratio [OR], 1.07; 95% CI, 1.01-1.12; P = 0.020) and with a significant colonoscopy finding (OR, 12.62; 95% CI, 1.59-99.89; P = 0.016). Male sex was a statistically significant factor for a significant gastroscopy finding (OR, 1.80; 95% CI, 1.12-2.91; P = 0.015). There was no association between cancer or significant endoscopic findings based on the degree of anemia or iron deficiency. Conclusion(s): Gastrointestinal endoscopy identifies significant endoscopic findings in more than 65% of patients undergoing investigation for iron deficiency and/or anemia. The majority of these findings are non-serious, and the prevalence appears similar in patients with IDA, ID, and non-ID anemia. However, there is a trend towards reduced cancer prevalence in patients with ID compared with IDA, although the overall numbers of cancers were small in this study population. Increasing age and male sex were factors associated with significant endoscopic findings and can perhaps be used to risk-stratify endoscopic investigation in the ID and non-ID anemia groups.