The role of capsule endoscopy combined with double-balloon enteroscopy in diagnosis of small bowel diseases

Background The diagnosis of small bowel diseases remains relatively inefficient using traditional imaging techniques. Capsule endoscopy （CE） and double-balloon enteroscopy （DBE） are two novel methods of enteroscopy for examining the entire small bowel. The aim of this study was to evaluate the detection rate and diagnostic accuracy of CE and DBE in patients with suspected small bowel diseases and to investigate the clinical significance of combined use of these two novel modalities. Methods Two hundred and eighteen patients were evaluated for suspected small bowel disease, including 116 with obscure gastrointestinal bleeding and 102 with obscure abdominal pain or chronic diarrhea. One hundred and sixty-five out of these patients underwent CE first and 53 patients underwent DBE （under anesthesia with propofol） first. DBE was recommended after negative or equivocal evaluation on CE and vise versa. Introduction of the endoscope during DBE was either orally or anally and the patients were referred for a second procedure using the opposite route several days later when no abnormalities were found on the first procedure. The detection rates, diagnostic accuracy, tolerance and frequency of adverse events of these two modalities were then analyzed. Results Failure of the procedure was seen in one patient with CE and in two patients with DBE. Sixty-four DBE procedures were carded out in 51 patients; by the oral route in 34 cases, the anal route in 4 and both routes in 13 cases. The overall detection rate of small bowel diseases using CE （72.0%, 118/164） was superior to that with DBE （41.2%, 21/51）; X2=16.1218, P〈0.0001. The diagnostic rate （51.8%, 85/164） was also higher than that with the latter procedure （39.2%, 20/51）, but was not significantly different （X^2=2.4771, P〉0.05）. Furthermore, the detection rate of small bowel diseases in patients with obscure gastrointestinal bleeding using CE （88.0%, 88/100） was superior to that of DBE （60.0%, 9/15）; X2=7.7457, P=0.0054. Lesions were detected by DBE in 1 out of 4 patients in whom CE had a negative result. Suspected findings by CE were confirmed by DBE combined with biopsy in 12 out of 15 patients. On the other hand, small bowel lesions were identified by CE in all 3 patients after negative evaluations by DBE. There were no severe complications during or after either of the two procedures. Conclusions The detection rate of small bowel diseases by CE is very high. CE should be selected for the initial diagnosis in pa