Endoscopists who took about a minute longer during screening esophagogastroduodenoscopy (EGD) detected significantly more upper gastrointestinal neoplasms than their quicker colleagues, in a retrospective study reported in the August issue of Gastroenterology ( doi: 10.1053/j.gastro.2017.05.009 ).

Detection rates were 0.28% and 0.20%, respectively (P = .005), and this difference remained statistically significant after the investigators controlled for numerous factors that also can affect the chances of detecting gastric neoplasms during upper endoscopy, reported Jae Myung Park, MD, and his colleagues at Seoul (Korea) St. Mary’s Hospital.

Gastric cancer is the second leading cause of cancer mortality worldwide, they noted. Although EGD has been practiced globally for decades, there is no consensus on which performance metrics optimize diagnostic yield. Both Korea and Japan have adopted screening EGD to help diagnose gastric cancer, but interval cancers and false-positive results remain concerns, and the field lacks rigorous analyses comparing EGD time with gastric cancer detection rates. Therefore, the researchers studied 111,962 consecutive patients who underwent EGD at one hospital in Seoul from 2009 through 2015. Fourteen board-certified endoscopists performed these procedures at a single endoscopy unit.A total of 262 patients (0.23%) were diagnosed with gastric neoplasia or dysplasia, duodenal neoplasia, or esophageal carcinoma. This finding closely resembled a prior report from Korea , the investigators noted. Detection rates among individual endoscopists ranged from 0.14% to 0.32%, and longer EGD time was associated with a higher detection rate with an R value of 0.54 (P = .046). During the first year of the study, EGDs without biopsies averaged 2 minutes, 53 seconds. Based on this finding, the researchers set a cutoff time of 3 minutes and classified eight endoscopists as fast (mean EGD duration, 2 minutes and 38 seconds; standard deviation, 21 seconds) and six endoscopists as slow (mean duration, 3 minutes and 25 seconds; SD, 19 seconds). Slower endoscopists were significantly more likely to detect gastric adenomas or carcinomas, even after the investigators controlled for the biopsy rates and experience level of these endoscopists and patients’ age, sex, smoking status, alcohol use, and family history of cancer (odds ratio, 1.52; 95% confidence interval, 1.17-1.97; P = .0018).

Older age, male sex, and smoking also were significant correlates of cancer detection in the multivariable model. Biopsy rates among endoscopists ranged from 7% to 28%, and taking a biopsy was an even stronger predictor of detecting an upper GI neoplasm than was EGD duration, with an R value of 0.76 (P = .002). Although this study was observational and retrospective and lacked data on prior EGD examinations, surveillance intervals, and rates of false-negative results, it “supports the hypothesis that examination duration affects the diagnostic accuracy of EGD,” Dr. Park and his coinvestigators reported. “Prolonging the examination time will allow more careful examination by the endoscopist, which will enable the detection of more subtle lesions. Based on our data, we believe that these indicators can improve the quality of EGD.”

Funders included the Ministry of Education, Science, and Technology; the Ministry of Science, ICT, and Future Planning; and the Ministry of Health & Welfare, Republic of Korea. The investigators reported having no conflicts of interest.