Background: The is currently a heavy burden on endoscopy services worldwide and although guidelines for the appropriate use of esophagogastroduodenoscopy (EGD) have been well studied, there are few such studies with respect to colonoscopy and none for the Asia-Pacific region. This study aimed, firstly, to determine the 'appropriateness of colonoscopy' for procedures performed in the endoscopy unit of a large Asian hospital using the American Society of Gastrointestinal Endoscopy (ASGE) 2000 guidelines, and secondly, to determine predictive factors including 'appropriateness of colonoscopy' for positive findings and colorectal cancer (CRC).
Methods: A prospective cross-sectional study was carried out on consecutive colonoscopies performed in the University of Malaya Medical Center. The unit has an open-access endoscopy policy for doctors who work in the hospital. Referrals were from endoscopists (gastroenterologists and surgeon-endoscopists), primary care physicians and other specialists. The indication of a procedure referral was recorded and judged 'appropriate' or 'inappropriate' using the ASGE criteria. The colonoscopic findings were recorded and classified as positive (endoscopies showing any pathology that had direct therapeutic or prognostic consequences) or negative findings (endoscopies showing no pathology or minor pathologies). Predictive factors for positive findings and CRC were determined using multivariate analysis.
Results: Of 380 patients referred for colonoscopy, 220 (57.9%) were classified as appropriate according to the ASGE guidelines, and 49 (12.9%) as inappropriate. The remaining 111 patients (29.2%) presented with complaints and conditions that could not be categorized. The rate of appropriate referral was similar for all three categories of physician (endoscopists: 59.8%, primary care physicians: 48.1%, others: 51.1%). When referrals by endoscopists were substratified according to gastroenterologists and surgeon-endoscopists, the rate of appropriate referral among gastroenterologists (78.4%) was significantly higher than that of surgeon-endoscopists (56.1%) (P = 0.049), primary care physicians (P = 0.013) and 'others' (P = 0.009). The most common appropriate indications were unexplained Rectalbleeding (79 cases, 20.8%) followed by CRC surveillance (45 cases, 11.8%). The most common inappropriate indication was inappropriately timed colonic cancer surveillance (32 cases, 8.4%). Chronic constipation in 36 cases (9.5%) was the most common 'unlisted' indication. A positive colonoscopic finding was detected in 131 (34.5%) examinations and CRC was found in 36 patients (9.5%). Appropriateness of indication was not a predictive factor for positive findings or CRC and there was no difference in the proportion of cases with positive findings or CRC in the three 'appropriateness categories'. Multivariate analysis revealed that only Rectalbleeding and smoking were significant independent positive predictive factors for positive findings and CRC.
Conclusion: The appropriateness of colonoscopy was not high among the different sources of referrals except for the subgroup of 'gastroenterologist'. Furthermore, the rates of positive findings and CRC among the cases with appropriate, inappropriate and unlisted indications did not differ. The ASGE guidelines will need to be modified for Asia to be of practical use.