Some patients are getting colonoscopies too often

A colonoscopy is one of the most effective tests available to protect against a major cancer — a message heard by growing numbers of adults. But a recently published study revealed that some patients get the exam too often, exposing them to unneeded risks and raising costs.

The research, which looked at doctors and their patients in the Boston area, is part of a significant shift in how medical care is studied. Until recently, computerized medical records have been used mostly to study whether patients are failing to get important tests, such as mammograms to detect breast cancer and blood tests to check for high cholesterol.

But increasingly, providers are building more sophisticated electronic medical record systems that allow researchers to study when patients receive too many, rather than too few, medical procedures and tests.

Eliminating unnecessary care has become a major focus in American medicine. Since 2012 a consumer campaign called Choosing Wisely has created lists of 350 things doctors and patients should question, such as antibiotics for ear infections, CT scans for simple headaches, and too-frequent colonoscopies.
Patients get unneeded care for many reasons, including when doctors are not up-to-date on clinical research, providers have a financial incentive to give more treatment, and patients themselves push for certain tests.
The Choosing Wisely project helped prompt Dr. Thomas Sequist, chief quality and safety officer at Partners HealthCare, to look at whether patients at Harvard Vanguard Medical Associates, a large Massachusetts physicians’ group, were getting colonoscopies more often than necessary.
Harvard Vanguard significantly increased colon cancer screening rates during the 2000s — last year it screened 86 percent of privately insured patients between ages 50 and 75, one of the higher rates in the state. Sequist wondered whether this dramatic improvement would encourage unnecessary use of the screening.
Sequist, who formerly oversaw the research program at Harvard Vanguard, and his colleagues identified 1,429 adults ages 50 to 65 who underwent their first screening colonoscopy between 2001 and 2010. The patients’ medical records included detailed information, such as whether growths found during the procedure were low risk or high risk for cancer, and whether patients had a family history of colon cancer, allowing researchers to more precisely determine when patients should have had follow-up procedures under national guidelines.

They discovered that most patients received follow-up procedures far earlier than the recommended 10 years for patients who were cancer free and had no problematic findings on their colonoscopies.

Patients who did not have growths, called polyps, in their colon received follow-up colonoscopies 6.9 years later on average, according to the study published last month online in the Journal of General Internal Medicine. Patients who had a noncancerous growth called a hyperplastic polyp, underwent a follow-up procedure even sooner, at 5.7 years.

“Those patients were getting colonoscopies repeated as though that finding was worrisome,’’ Sequist said.
National guidelines recommend that patients with adenomas, which are considered pre-cancerous, have a follow-up colonoscopy sooner, within 3 to 5 years. Nearly half of these follow-up exams for higher-risk patients were also performed too early. Colon polyps grow very slowly.

Most patients who had early colonoscopies did so because their endoscopist, usually a gastroenterologist, recommended it.

“We want physicians and patients to be thinking about what will they gain from this test,’’ said Daniel Wolfson, executive vice president of the American Board of Internal Medicine Foundation, which runs the Choosing Wisely project. “The question is not why didn’t I do this test but why did I do this test? That cultural shift is the most important thing.’’

Sequist’s study did not examine why doctors recommended colonoscopies early, but Harvard Vanguard decided to examine the issue several years ago on its own. Dr. Richard Lopez, chief medical officer at Atrius Health, the parent organization of Harvard Vanguard, said doctors felt they were being more conservative by requiring a follow-up exam in seven years, and developed a pattern of practice that their colleagues reinforced, even though it was not based on scientific evidence.
“We have thought 10 years is a scary interval,’’ said Dr. Steven Shields, the organization’s chief of gastroenterology.
Harvard Vanguard hired outside experts to review the scientific evidence with doctors and the organization developed its own guidelines that mirror national recommendations. Leaders invited malpractice insurers to advise doctors that they could be more easily defended in a lawsuit if they had been following the guidelines.
The organization changed the timing of a follow-up procedure for patients with no evidence of cancer to 10 years, and doctors now comply with the guidelines, Shields said.
Dr. Lawrence Kosinski, an Illinois gastroenterologist who is on the board of the American Gastroenterological Association, does not believe US doctors “are trying to manipulate the system and make more money. We have many patients whose primary care doctor tells them it’s time to have another one. Or a relative or friend gets colon cancer and they come in afraid,’’ he said.
In other cases, patients do not prepare properly for the procedure and the gastroenterologist is not confident about getting a good view of the entire colon.
But Kosinski said that in some cases, doctors bring back patients early because there is no penalty for doing so. They think “I might as well do it,’’ he said.
Colonoscopies are generally low risk — about 1 in 200 patients who have polyps removed will experience complications. For patients who do not have growths removed, the complication rate is 1 in 1,000. But when accidental tears in the colon do occur, they can be serious.
There is also the cost to the health care system if colonoscopies are done too often, which Sequist estimated at $1.3 million for the 1,051 Harvard Vanguard patients screened early.

Kosinski said he does not want the study to send a message that too many people are getting colonoscopies. Nearly 40 percent of Americans over the age of 50 have not been screened for colon cancer by any method. In Kosinski’s practice, the average age for a first colonoscopy is 58.

“The greater issue is the number of people not undergoing colonoscopies,’’ agreed Dr. David Lichtenstein, director of endoscopy at Boston Medical Center.