Skip to main content

5 Tips for Avoiding Diagnostic Errors

 |  By cclark@healthleadersmedia.com  
   September 08, 2010

Now that the quality movement seems to be chugging along on a well-conceived, evidence-based track, some thoughtful person has come along tooting a horn to warn us to "wait here just a darn minute. Aren't you all forgetting something that's really, really important?"

Or words to that effect.

The man who made such a stir this week, and got my attention, is hospitalist and patient safety expert Robert Wachter MD, an author and associate chair of the University of California San Francisco Department of Medicine.

Measurements of quality—checklists, process measures, and pay for performance score cards—are all fine as ways to reduce medical errors, Wachter says. But they neglect to force us to repair an enormous defect that now exists in the way we care for patients.

Far too often we're getting the diagnosis wrong.

Turns out, the healthcare system performs rather terribly in this earliest stage of the healthcare delivery system. And hospitals, physician groups and quality experts together need to start paying attention, he says.

Wachter gave some examples in his article published this week in the journal Health Affairs "Why Diagnostic Errors Don't Get Any Respect—And What Can Be Done About Them."

First, "approximately one in 10 autopsies uncovers some disease or condition that—had its existence been known when the patient was alive—would have altered his or her care or changed the prognosis," he wrote.

"That number has remained shockingly stable for 30 or 40 years in studies of autopsies," he said in an interview, adding that the error rate continues today, despite the fact that "It's hard to die now without getting a CT or MRI of some part of the body."

According to a 2004 report published in the Archives of Internal Medicine, autopsies revealed incorrect diagnoses among patients hospitalized in the intensive care unit in this order: 21 cancers, 12 strokes, 11 myocardial infarctions, 10 pulmonary emboli and nine endocarditis.

Second, the game-changing report published by the Institute of Medicine nearly 11 years ago, To Err Is Human, estimated that medical errors cause between 44,000 and 98,000 deaths annually, which it drew from another report, the Harvard Medical Practice Study.

But what the IOM failed to mention was that the Harvard study it drew from also noted that of all medical adverse events, 17% were due to diagnostic errors, "far more than medication errors," which are the focus of so much preventive spending and effort today, Wachter says. "If you review the IOM report, the term 'medication errors' is mentioned 70 times, while 'diagnostic errors' occurs twice," he wrote.

Third, diagnostic errors account for twice as many malpractice lawsuits as other types of medical errors.

What all this means is that doctors are missing an enormous opportunity to save lives, prolong or improve lives, avoid waste in the healthcare system and save money.

Somewhere the patient safety movement has drawn a trajectory for improvement, but it has left the challenges of improving patient diagnoses far behind, he says. Neither the National Quality Forum's list of 'never events' and the Agency for Healthcare Research and Quality's Patient Safety Indicators includes a diagnostic error.

Wow, I thought. This is huge!

Of course there are many understandable reasons why the patient safety movement now focuses on these other preventable medical mishaps, such as wrong-site surgery, forgotten foreign surgical objects, medication errors, hospital acquired infections, and a variety of other hospital-associated mishaps resulting in shocks, burns, or falls.

Wachter provided a few.

In improving patient safety, he says, "thoughtful people look for low-hanging fruit, things that are easy to measure, and then that becomes the scope of a problem," he told me this week. "These are not trivial errors. But the conspiracy that occurs is the conspiracy of measurement. It's so much easier to measure whether you gave the patient the right medication than it is to measure whether you made the right diagnosis.

"Likewise, it is easier to measure whether surgeons called a "time out" before surgery, or whether a series of processes was performed to prevent central line infecitons.

"Diagnostic errors mostly reflect cognitive issues, such as failing to adequately consider alternative diagnoses," Wachter wrote. "No comparable series of processes (or structures) has been identified to prevent them."

Diagnostic errors, Wachter says, "don't elicit the visceral dread that accompanies wrong-site surgery." And, he says, "none of the examples of medical errors that produced an uproar in the media has involved a diagnostic error."

Wachter suggests five ways to start addressing this system of neglect:

  1. Improve board certification standards by requiring more frequent reviews and perhaps annual maintenance of certification documentation, with hospitals making that a requirement for staff privileges for certain specialties. Already, he says, efforts are underway to make this process more rigorous and remove some of the grandfather privileges that exempt doctors trained decades ago if they are still practicing.
  2. Encourage research on diagnostic errors to better understand how and when they happen, and whether computerized decision support tools reduce them. The AHRQ has provided some seed funding for such research.
  3. See what sorts of training are associated with improved diagnostic performance, and hospitals should be required to offer them or ensure that their medical staffs participate in them.
  4. Use technology, perhaps some of the $20 billion in federal support from the stimulus bill, to find health information technology strategies that reduce diagnostic errors.
  5. Improve medical teaching by having the Accreditation Council for Graduate Medical Education ensure that residencies and medical schools train students in diagnostic reasoning, including more creative use of simulations and model patients.

Above all, Wachter believes that doctors all too often miss the right diagnosis because they engage in "premature closure," settling on one diagnosis without considering alternatives. They should ask themselves "what else could this be?" or "What is the worst thing that could be going on?" perhaps by imagining that if the patient were to be admitted in the hospital with a different diagnosis a few days later, what would the reason be?

In the end, Wachter says, while our post-IOM focus on systems improvement has been critical, we also need to continue turning out "the really smart doctor: someone who keeps up on the literature, goes home at night scratching his head wondering what this could be, who can make the really tough diagnosis." 

He worries that, "unless we specifically value diagnostic excellence, these skills may end up going by the wayside." 
"We will look good on that quality and patient safety report card," he says, "but get the diagnosis wrong."

Tagged Under:


Get the latest on healthcare leadership in your inbox.