Audit of investigations in patients with iron deficiency anaemia

Patients with iron deficiency anaemia (IDA) are commonly referred to the gastroenterologists for exclusion of gastrointestinal (GI) pathologies. The British Society of Gastroenterology (BSG) has published a guideline for management of IDA. As recommended by this guideline, all patients should have examinations of both upper and lower GI tract with the exception of pre-menopausal women younger than 45 years old. The primary aim of this audit was to determine how thoroughly patients referred to our unit at the Singapore General Hospital had been investigated. The secondary aim was to determine the yield rate of gastroscopy and lower gastrointestinal investigations (colonoscopy or barium enema) in our patients.

We reviewed the case notes of patients who underwent gastroscopy for indication of anaemia between the period from April to December 1999. We classified iron deficiency anaemia as having haemoglobin of lower than reference range with either low ferritin level or iron/ TIBC ratio of less than 10%.

Of a total of 326 patients reviewed, 172 patients (65 men, 109 women) met the inclusion criteria of IDA. The median age for the group was 59 (Range 16 to 88) years old and there were 107 (62.2%) women. There were 31 female patients younger than 45 years old (group A) and 141 patients in group B which included all the males and females older than 45 years old. Due to the method of data retrieval, all of the patients in group A fulfilled the standard set by the BSG guideline in that they all underwent gastroscopy. In this group, few gastrointestinal pathologies were found (one patient with peptic ulcer and two patients with colonic cancer). In terms of finding potential gastrointestinal causes of IDA, the yield rate of gastroscopy and colonoscopy were 3.2% and 13.6% respectively. In group B, 96 patients (68%) underwent upper and lower gastrointestinal investigations. Evaluation with gastroscopy showed peptic ulcer disease in 31 patients (21.7%) and gastric cancer in 9 patients (6.3%). Only 96 patients (68.1%) in group B had lower gastrointestinal investigations. Twenty-six patients were found to have gastrointestinal disorders (12 patients with colon cancer, 10 colonic polyps, one Crohn's disease, one colonic tuberculosis and two haemorrhoids). The yield rate of gastroscopy and colonoscopy were 31.2% and 25% respectively.

In this study of patients with confirmed IDA, we found much higher incidence of gastrointestinal disorders in the group of male and postmenopausal women than in the group with premenopausal women. Furthermore, over 30% of patients in this group did not undergo lower gastrointestinal tract investigations as recommended by the BSG guideline. We also found two cases of colonic cancer in the latter group. We would recommend bidirectional endoscopy for postmenopausal women and men with IDA. For premenopausal women, we would recommend similar approach unless there are clinical pointers to dietary iron deficiency or menorrhagia.