Sensitivity and Specificity of the Suspected Blood Identification System in Video Capsule Enteroscopy

Background and Study Aims: Capsule enteroscopy has become a standard tool for the evaluation of obscure gastrointestinal bleeding. Reviewing the video recordings of capsule examinations is time-consuming and requires prolonged attention. Recently, software that can recognize the frames containing “red spots”, the Suspected Blood Identification system (SBIS), has been developed with the aim of assisting in the analysis of video recordings. We assessed the sensitivity and specificity of the SBIS in patients undergoing capsule enteroscopy.
Patients and Methods: 100 consecutive patients underwent capsule enteroscopy at our tertiary referral center, for the following indications: obscure gastrointestinal bleeding (75 patients), metastatic carcinoid (7 patients) known or suspected Crohn’s disease (5 patients), miscellaneous (13 patients). Capsule endoscopy was carried out by the standard method. Four gastroenterologists, experienced in capsule endoscopy, reviewed the recordings.
Results: Small-bowel visualization was obtained in 95 cases. The physicians identified 209 “red spots”, 54 of which (25.8 %) were also identified by the SBIS. The overall sensitivity, specificity, and positive and negative predictive values of the SBIS, calculated on the number of true-positive, true-negative, false-positive and false-negative results, were 40.9 %, 70.7 %, 69.2 % and 42.6 %, respectively. Sensitivity was higher for the identification of red blood (60.9 %) than for nonbleeding ”red” lesions such as arteriovenous malformations (25.8 %).
Conclusions: The SBIS has low sensitivity and specificity. It can be used as a complementary and rapid screening tool, but complete review of the recordings is still necessary.

References

• 1 Iddan G, Meron G, Glukhovsky A, Swain P. Wireless capsule endoscopy.  Nature. 2000;  405 417
• 2 Gong F, Swain P, Mills T. Wireless endoscopy.  Gastrointest Endosc. 2000;  51 725-729
• 3 Lewis B S, Swain P. Capsule endoscopy in the evaluation of patients with suspected small intestinal bleeding: results of a pilot study.  Gastrointest Endosc. 2002;  56 349-353
• 4 Ell C, Remke S, Helou L. et al . The first prospective controlled trial comparing wireless capsule endoscopy with push enteroscopy in chronic gastrointestinal bleeding.  Endoscopy. 2002;  34 685-689
• 5 Saurin J C, Delvaux M, Gaudin J L. et al . Diagnostic value of endoscopic capsule in patients with obscure digestive bleeding: blinded comparison with push enteroscopy.  Endoscopy. 2003;  35 576-584
• 6 Pennazio M, Santucci R, Rondonotti E. et al . Outcome of patients with obscure gastrointestinal bleeding after capsule endoscopy: report of 100 consecutive cases.  Gastroenterology. 2004;  126 643-653
• 7 Mylonaki M, Fritscher-Ravens A, Swain P. Wireless capsule endoscopy: a comparison with push enteroscopy in patients with gastroscopy and colonoscopy negative gastrointestinal bleeding.  Gut. 2003;  52 1122-1126
• 8 Costamagna G, Shah S K, Riccioni M E. et al . A prospective trial comparing small bowel radiographs and video capsule endoscopy for suspected small bowel disease.  Gastroenterology. 2002;  123 1385-1388
• 9 Liangpunsakul S, Mays L, Rex D K. Performance of Given suspected blood indicator.  Am J Gastroenterol. 2003;  98 2676-2678
• 10 Zanati S A, Tang S J, Dubcenco E. et al . Value of the suspected blood indicator in wireless capsule endoscopy.  Gastrointest Endosc. 2004;  59 M1785
• 11 Michalet V, Roman S, Radenne S. et al . Automatic detection of “red spots” during videocapsule study of the small intestine for unexplained GI bleeding: is it useful?.  Gastrointest Endosc. 2004;  59 M1786