Iron deficiency anaemia (IDA) is a common clinical problem presenting for endoscopic investigation. There are no data from New Zealand to guide clinicians in this country. Guidelines for the investigation of IDA have been published but these are frequently not strictly adhered to.1–4 It is commonly accepted that there is a high incidence of dual pathology and it is generally accepted that both upper and lower gastrointestinal (GI) investigations are necessary for a full GI workup in the investigation of IDA unless malignancy is clearly demonstrated on the initial endoscopic investigation.

Thus we performed a study to investigate the spectrum of disease found in patients who undergo a full GI workup for IDA in New Zealand. We were also interested in knowing whether on the basis of spectrum of disease, presenting symptoms or degree of iron deficiency it could be predicted which test might be the most appropriate initial endoscopic investigation.

A computerised Endoscopy Database (EndoscribeTM) was used to look for patients who had IDA as an indication for upper and lower GI endoscopy at Hutt Hospital from May 1998 to January 2004. Males and females over the 50 years of age were included in the study. Patients were only included if both upper and lower endoscopies were performed within a four month period of each other.

Eighty-five patients were entered into the study; 46 were female. The mean age was 72 years. Significant GI lesions were found in 46 (54%) of the patients. In 22 patients (26%), lesions likely to cause IDA were found during upper endoscopy. These were all benign. Twenty-two patients (26%) were found to have colorectal cancer. Nine patients (11%) had lesions in both the upper and lower GI tracts that could be causing IDA.

Thirty-seven patients (44%) underwent duodenal biopsy and in all of these the histology showed normal duodenal mucosa. Helicobacter pylori was checked in 61 patients using either a CLO test or histology. Six of these patients (10%) were found to be H. pylori-positive.

Faecal occult blood (FOB) was checked in 36 patients. Of the 11 patients with positive FOB, only 4 had abnormal colonoscopies. Of the 26 patients with a negative FOB, 6 had abnormal colonoscopies. This gives a negative predictive value of 77% and a positive predictive value of 36% for a positive FOB.

The presence or absence of upper or lower GI symptoms was not predictive of the nature nor site of pathology found. Neither was the degree of anaemia.

As in studies from other Western countries, significant lower GI pathology is a common finding in investigation for IDA.5–8 Dual pathology is common. Due to the high incidence of lower GI malignancy in this study and the low known incidence of upper GI malignancy as a cause of IDA it would seem prudent to perform colonoscopy as the initial endoscopic procedure in the investigation of IDA in New Zealand.