New study puts hospital error death rate at twice IOM's total
By Don Long
Source: The Worldwide Biotechnology News and Information Source
Published: July 28, 2004
Medical Device Daily Managing Editor
People frequently die in hospitals because of medication errors and other mistakes. And when discussing this unfortunate reality, most put this yearly death total at 98,000 - a figure provided by the 1999 report "To Err is Human," by the Institute of Medicine (IOM; Washington). But that figure may be too low.
Way too low, according to HealthGrades (Lakewood, Colorado), a company that specializes in tracking patient outcomes and giving awards to hospitals that they assess as performing the best.
HealthGrades yesterday released a report setting the figure at double the IOM's estimates - a total of 195,000 - the result of what the organization terms "potentially preventable, in-hospital medical errors."
Worse, the actual figure may be even higher, according to Dr. Samantha Collier, vice president of medical affairs for HealthGrades.
"We're relying on data that hospitals submit, and that might be a reason to under-document" the actual number of errors and resultant in-hospital deaths, Collier told Medical Device Daily.
HealthGrades says the "Patient Safety in American Hospitals" study is the first to look at the mortality and economic impact of medical errors and injuries that occurred during Medicare hospital admissions nationwide from 2000 to 2002.
The HealthGrades study applies the mortality and economic impact models developed by Dr. Chunliu Zhan and Dr. Marlene Miller in a study published in the Journal of the American Medical Association (JAMA) in October 2003. The Zhan/Miller study supported the IOM's figure of 98,000 unnecessary in-hospital deaths annually, a figure which translates, in terms of deaths, to a national epidemic.
But Collier says that HealthGrades' total is probably much more accurate than these previous studies because its study is more comprehensive.
The IOM study, she noted, used extrapolated figures from three states, and the Zhan/Miller study looked at 7.5 million patient records from 28 states over one year. By contrast, HealthGrades analyzed three years of Medicare data in all 50 states and the District of Columbia, or about 45% of all hospital admissions (excluding obstetric patients) in the U.S. from 2000 to 2002, extrapolating them to the entire nation
This method "makes the extrapolation more accurate, but I [still] don't know if it's the exact right number," Collier told MDD. "And we were only looking at in-hospital errors," she said, suggesting that errors in the outpatient setting would balloon the death total even higher.
While the IOM report resulted in considerable "outrage" and clearly supported the need for "a complete paradigm shift," she says there has been little progress in taking the actions necessary to stem the tide of deaths.
Putting the problem in the most graphic terms possible, she says the 195,000 figure translates to "the equivalent of 390 jumbo jets full of people . . . dying each year due to likely preventable, in-hospital medical errors, making this one of the leading killers in the U.S." And she adds: "If the Center for Disease Control and Prevention's [Atlanta, Georgia] annual list of leading causes of death included medical errors, it would show up as No. 6, ahead of diabetes, pneumonia, Alzheimer's and renal disease."
In its study, HealthGrades examined 16 of the 20 patient-safety indicators defined by the Agency for Healthcare Research and Quality (AHRQ; Washington) - from bedsores to post-operative sepsis - omitting four obstetrics-related incidents not represented in the Medicare data used in the study. Of these 16, the mortality associated with two - failure to rescue and death in low-risk hospital admissions - accounted for the majority of deaths that were associated with these patient safety incidents.
Those two categories of patients were not evaluated in the IOM or JAMA analyses, accounting for the variation in the number of annual deaths attributable to medical errors, according to HealthGrades.
"If we could focus our efforts on just four key areas - failure to rescue, bed sores, postoperative sepsis and postoperative pulmonary embolism - and reduce these incidents by just 20%, we could save 39,000 people from dying every year," says Collier.
She acknowledges that the study was not designed to uncover a set of solutions, since the particular solutions "will vary from hospital to hospital." But she cites two important avenues hospitals need to pursue.
The first, Collier says, is "transparency" concerning medical errors, while adding that this is "easier said than done, given our legal situation and [what is happening in] the malpractice crisis states." But, "we will never get to improvement [in medical errors] without being transparent about the problem," she says. "We need to put pressure on hospitals, doctors and nurses to find out `What is the error rate at your hospital, and what are you doing about it?' At minimum, [hospitals] should make this transparent within their own organization."
Put still another way, she says that a solution is unlikely without knowing the extent and nature of the problem.
A second approach, she says, is for hospitals to develop a team or "task force" assigned to tackle patient safety issues. "We need to understand the processes that occur and bring back solutions in a team approach. That's what everybody's advocating." The approach, she says, would be similar to that used by airlines - specifically "a whole team, all on the same page, before we take off and land."
Collier says that HealthGrades is attempting to promote this goal by providing information to consumers concerning the best hospitals, defined in terms of the number of errors they report. It has just honored 88 hospitals in 23 states with its Distinguished Hospital Award for Patient Safety.
"Our mission is really to improve the quality of healthcare by giving consumers, or current or future patients, information [about hospitals]," she says. "Yes, they should understand that medical errors do happen. But they can have some control over that by asking the right questions. They can empower themselves with information."
For more results from the HealthGrades study, see box, p. 5.
The complete study, including the list of AHRQ patient-safety indicators, can be found at http://www.healthgrades.com.