WASHINGTON—With a number of conflicting studies over the last five years looking at the effect of delays between positive colon cancer screenings and the subsequent colonoscopy that could confirm a diagnosis of colorectal cancer (CRC), VA researchers have struggled to determine exactly how a holdup of care effects outcomes.
Now, they are turning to technology for the answers.
At the Michael DeBakey VAMC in Houston, researchers are working on creating an electronic health record-based trigger system to detect potential delays in colorectal cancer diagnoses, giving VA facilities a way of tracking delays and giving physicians help in speeding up a diagnosis.
In 2007, VA declared that no colonoscopy should be delayed more than 60 days after certain clinical indicators had been uncovered, namely iron deficiency anemia (IDA) or a positive fecal occult blood test (FOBT).
Since then, Houston researchers have designed and validated a series of electronic screening algorithms that identified patients who had a positive screening but failed to have a colonoscopy during that 60-day window. They then conducted a study to determine whether using IDA and FOBT positive screenings and triggers would accurately identify patients most in need of a colonoscopy.
Between March 2009 and March 2010, 337 veterans were screened at high-risk for CRC using a FOBT and 258 at risk because of IDA. Those triggers were found to be relatively reliable. Of the positive FOBT screenings, 73% of patients were found to have cancer, while 76.7% of those with IDA triggers were found to have cancer.
The researchers followed a random selection of 120 of those patients who had positive screenings. Of them, 90 would eventually be found to have cancer—a 75% accuracy rate. After 30 days, one-third of those patients did not have a colonoscopy ordered, while one-third had one ordered, but never had it done.
“We saw most number of colonoscopies happening between 60 and 90 days,” said Archana Laxmisan, MD, a Baylor University instructor and one of the Houston VAMC researchers. Laxmisan, who detailed the study at the annual VA Health Services R&D conference, added, “We do think our positive predictive value is superior to other methods of detection being used out there for diagnostic delays. It’s definitely more efficient than manual chart review.”
By installing this kind of electronic algorhythm into the electronic health record, it can keep facilities abreast of delays and, with the addition of a reminder function, can help physicians follow up with patients. “Our methodology can be used by facilities to proactively identify delays in diagnosis,” Laxmisan said.
Laxmisan noted that, of the 30 patients in the randomly selected group who had positive screenings but were found not to have cancer, 18 of them used an outside physician, and 19 patients went for a colonoscopy outside the VA system. “We’re also looking at why patients don’t get these kinds of things in the VA system,” Laxmisan said.