This patient’s medical record said she’d given birth twice — in fact, she’d never been pregnant
Like many patients with a chronic illness, Gleason sees a lot of doctors. So she and her mother, Amy, who works at a health IT company called CareSync, always make it a point to request her medical records after a visit so they can store a copy of all her records in one place.
Two years ago, Gleason requested her records after visiting a women’s health clinic in Florida. To her surprise, she found a note in the record saying she’d had two children. One was apparently still living and the other had died shortly after she gave birth.
According to the chart, which she provided to CNBC, Gleason would had to have given birth to the first child at age 13 for the dates to line up.
In fact, Gleason had never been pregnant.
It wasn’t the first mistake. In a different record, Gleason had spotted a diagnosis of diabetes shouldn’t have been there. She had looked into it after a doctor had gone off script by peppering her with questions about her blood sugar.
Gleason’s case might seem extreme, but it is far from unusual.
Ross Koppel, a leading academic in the world of health IT, estimates that about 70 percent of records have wrong information.
“There are millions of records out there and screw-ups happen,” he said.
Most errors are irrelevant to health outcomes — for instance, a record might state that a patient hurt her finger on Thursday, when it was in fact, Friday.
But mistakes can be harmful or even fatal if they lead to a misdiagnosis, a failure to recognize an allergy or a wrong dose of a medication. A Johns Hopkins study found that more than 250,000 people in the U.S. die every year from medical mistakes, making it the third-leading cause of death after heart disease and cancer. Other studies have found the numbers to be higher than 400,000.
But instead of fixing it, the assistant on the other end of the line insisted that Gleason was wrong. “If you hadn’t told us this, there’s no way this could have been in your chart,” Gleason was told. She ended up having to file a formal document requesting a correction of her record, before any changes were made.
Koppel, a sociologist and bio-informaticist at the University of Pennsylvania and Penn Medical School, was not surprised that Gleason had such a hard time getting the correction. Doctors don’t like to admit wrongdoing for fear of being sued, he explained, but there’s also an institutionalized distrust of patients. Some patients do lie or hide certain embarrassing facts.
That needs to change, suggests Koppel, because doctors are human. It’s easy to imagine how mistakes might happen if a busy resident has multiple records open on a laptop, and they’re using copy and paste. In other instances, they might simply mistype or mishear something: A common example of that is “hypo” versus “hyper.”
There are also cases of mistaken identity. In the United States, unlike in other developed nations, such as the United Kingdom and many of the Nordic countries, there’s no unique patient identifier number linked to a single centralized record. That makes it a challenge for doctors to ensure that they’ve pulled up the record of the right patient if they happen to have a common name.
To rectify that, Koppel always encourages nurses and doctors to take photos of their patients and clip them to the record so they can at least ensure that they’re talking to the right person.
Gleason is extremely concerned that these errors will continue to follow her, given the broad array of medical record vendors. It’s already been hours of work for her to try to fix them across multiple systems.
For that reason, she has a few words of caution for new technology players like Amazon and Apple as they get deeper into health care with a particular focus on medical records.
“I hope that companies in tech don’t start looking at the text in physician notes and making determinations without a human or someone who knows my medical history very well,” she said. “I’m worried about more errors.”
