Evaluation of Capsule Endoscopy Referrals for Anemia: Is the Diagnosis of Iron Deficiency Anemia Correctly Made?: ACG/AstraZeneca Fellow Award

Abstract

Purpose: Patients with iron deficiency anemia (IDA) due to obscure gastrointestinal bleeding (GIB) is a recognized indication for small bowel capsule endoscopy (SBCE). Our primary study goal was to determine the frequency that patients with anemia not due to iron deficiency were referred by gastroenterologists inappropriately for capsule endoscopy. We also wanted to determine the financial impact of these inappropriate referrals. Methods: A retrospective review of SBCE studies from January 2004 through December 2007 was performed on patients referred for IDA after negative preceding EGD and colonoscopy. Hemoglobin, MCV, MCHC, RDW, serum iron, total iron binding capacity, iron saturation, and ferritin levels were obtained from the patients' records when possible. Patients were excluded if they were referred for active GIB with anemia. Patients were classified as definite IDA, no evidence of IDA, or equivocal. Patients with only a CBC available were defined as suggestive of IDA, no suggestion of IDA, or equivocal for IDA. Criteria for diagnosing definite iron deficiency: serum iron <60 ucg/dL, TIBC >350 mcg/dL, iron saturation <15%, and ferritin <20 ng/mL. Supportive CBC indices consistent with IDA were MCHC <32 g/dL, MCV <80 fL, hemoglobin <12 g/dL in women and <13 g/dL in men, and RDW >15%. Normal CBC and iron studies were classified as no evidence of IDA. Laboratory values were considered equivocal if the ferritin levels were between 20-50 ng/mL and the majority of the laboratory pattern fulfilled our definition of IDA. Results: A total of 135 patients were included. Pre-procedural laboratories were available in 107 (79%). Seventy-eight (58%) of the 135 patients had both CBC and iron study data available. However, in only 32 (41%) were the laboratory values supportive of IDA. Twenty-eight patients (36%) had no evidence of IDA and 18 (23%) had equivocal laboratories. Of the total 135 patients, 29 (21%) only had CBC and no iron studies. Five patients (17%) had CBC indices suggestive of IDA, 15 (52%) had indices not supportive of IDA, and 9 (31%) had equivocal laboratories. Twenty-one percent of the total patients had no available laboratories. Conclusion: These findings indicate that a large number of patients were inappropriately diagnosed as IDA and referred for SBCE. At our facility, the all inclusive cost of SBCE is $3695.00. The cost of doing inappropriate SBCE was $262,345 during the study period. Reasons for inappropriate referrals are misinterpretation of laboratory results and incomplete work-up of anemia. Additional costs include EGD/colonoscopy for non-IDA patients.