CHICAGO – Only a small portion of patients with nondysplastic Barrett’s esophagus received appropriately timed endoscopic surveillance, a large database study showed.

But rather than being neglected, patients were more likely to be overassessed, with follow-up endoscopy performed more frequently than the recommended 3- to 5-year intervals, Anna Tavakkoli, MD, reported at the annual Digestive Disease Week.

“Very few patients entered our surveillance program with appropriate surveillance intervals,” said Dr. Tavakkoli, a gastroenterology fellow at the University of Michigan, Ann Arbor.

The review of 1,602 patients found that a close link with primary care significantly mitigated the risk of inappropriate follow-up, Dr. Tavakkoli said in an interview.

“We don’t have a formal program at the University of Michigan that drives coordination of care, but we do have great communication here between our primary care providers and our specialists. Our electronic medical records system makes quick messaging between providers easy, and primary care is very good about incorporating diagnoses into patients’ problem lists.”

Malignant transformation of nondysplastic Barrett’s is uncommon, with rates of no more than 4% per year. This understanding led three major societies – the American Gastroenterology Association, the American Society of Gastrointestinal Endoscopy, and the American College of Gastroenterology – to amend their surveillance recommendations in 2011 and 2012. All three societies now recommend a surveillance endoscopy every 3-5 years after the initial diagnosis of nondysplastic Barrett’s esophagus. In fact, the AGA has incorporated this suggestion into its five “Choosing Wisely” recommendations aimed at decreasing overutilization of testing and procedures.

Dr. Tavakkoli’s study examined surveillance timing in a cohort of 1,602 patients with nondysplastic Barrett’s who entered the University of Michigan Barrett’s Esophagus Registry from 1994 to 2016. All of these patents had at least three endoscopies or at least 5 years of follow-up data since their last endoscopy. The primary outcome was identification of trends in the appropriateness of surveillance of patients with nondysplastic Barrett’s esophagus at the University of Michigan. In her analysis, oversurveillance was defined as less than 3 years between the second and third endoscopy; undersurveillance was defined as more than 5 years between them. Dr. Tavakkoli and her colleagues also looked at patients who were lost to follow-up, defined as never receiving a second endoscopy after their initial diagnosis of nondysplastic Barrett’s esophagus and patients who were never surveilled, defined as never receiving their third endoscopy. All patients were compared with those who underwent appropriate surveillance, defined as 3-5 years between their second and third procedure.

The majority of patients were male, and the mean age was 59 years; 30% had long-segment Barrett’s, and 41% had a primary care provider in the university health care system. Most (90%) had their second endoscopy before 2012, when two of the three major societies issued their updated surveillance recommendations.

Of the entire cohort, 40% were lost to follow-up; 17% were never surveilled, and 3% were undersurveilled. Almost a third (31%) were oversurveilled, while just 8% had the appropriate surveillance, Dr. Tavakkoli said.

She then looked at several demographic and clinical factors associated with surveillance in each group, including sex, age, race, and income, comorbidities, length of Barrett’s, family history of esophageal cancer, and whether the patient had a University of Michigan primary care provider.

Having long-segment Barrett’s was associated with a 2.5-times increased risk of receiving a third endoscopy earlier than 3 years, which may be driven by studies that have shown that the risk of malignant transformation increases with Barrett’s length, she said.

The presence of a primary care physician significantly reduced the risk of inappropriate follow-up in every group, except patients who were undersurveilled, she said. The presence of a primary care physician at the University of Michigan decreased the risk of oversurveillance by 56%.

The positive influence of an in-system primary care physician was an important finding in this study, Dr Tavakkoli said. “The oncology data have shown us that poor coordination of care between oncologists and primary care providers contributes to avoidable patient morbidity and mortality, fragmented care, and increased costs. In 2005, the Institute of Medicine published a report emphasizing that coordination between specialists and primary care providers is one of the four key components to cancer survivorship care. There have been a number of GI studies looking at how primary care’s involvement in colorectal screening improves the rates of patients who undergo screening, but among Barrett’s patients, there have not been data showing that having a primary care physician at the center where endoscopic surveillance is done improves utilization patterns.”

Dr. Tavakkoli had no financial disclosures.

msullivan@frontlinemedcom.com

On Twitter @alz_gal

Primary care involvement improves chance of meeting recommended surveillance intervals