Hospitals' focus on patient safety hasn't eliminated preventable deaths
Some years ago, I got a call at 3 a.m. from the hospital because a patient of mine had spiked a high fever. Suspecting an infection, I called in some antibiotics. A few hours later, a frantic nurse called to say my patient had turned red and was wheezing, likely from an allergic reaction.
I rushed to the hospital and saw my own note in the patient's chart stating that the patient was allergic to the antibiotic I had ordered. I had made a preventable medical error; so had the nurse who had missed the allergy wristband on the patient and the pharmacist who had neglected to check the patient's allergy drug list.
Last weekend when I was on call and seeing my sixth consult, I inadvertently ordered a penicillin-family drug for a patient with a penicillin allergy. This time the pharmacy's software flagged the potential error.
Despite the best of intentions, errors are common in hospitals even with new safeguards. A decade ago the Institute of Medicine published its landmark report "To Err Is Human: Building a Safer Health System," which estimated that 44,000 to 98,000 deaths occur annually because of preventable medical errors in U.S. hospitals.
Over the past decade as a hospital-based physician, I have noticed great efforts to improve patient safety. The software at my hospital's pharmacy is just one example of progress.
So I was surprised when a report on hospitalized Medicare patients released last month found that one in seven of the patients experienced an adverse event such as excessive bleeding, a hospital-acquired infection or aspiration pneumonia. Those events, both preventable and not preventable, led to about 180,000 deaths a year.
A study published last month in the New England Journal of Medicine of 10 North Carolina hospitals found that "harms remain common, with little evidence of widespread improvement" from 2002 to 2007.
For me, these findings raise two critical questions: Why do adverse events happen in the first place, and why are we not able to lower the number of adverse events in the hospitals after a decade of effort?
An analogy may help. Adverse events in the hospital are much like car accidents. The problem is complex, and the solutions require massive system redesign and significant behavioral changes by individuals.
To reduce car accidents, system changes might include installing reflectors between lanes and redesigning dangerous intersections. Similarly at hospitals, checklists and electronic medical records are system changes designed to reduce adverse events.
On an individual level, drivers reduce road fatalities when they do not text or drink while driving. Similarly, in hospitals, doctors and nurses reduce adverse events when they dutifully perform hand hygiene and watch for drug interactions.
The major reason so many deaths occur on the roads and in hospitals, of course, is that both are inherently hazardous places. I am disturbed that 1.5 percent of hospitalized Medicare beneficiaries who experience an adverse event die each year. But I also recognize that many elderly patients are so ill that they might well die if not hospitalized.
At times, I see the adverse events at hospitals as collateral damage. As one director of quality improvement told me about hospital care, "There is no other situation where people from so many diverse backgrounds and skill levels come together every day to carry out a common mission with so much stress, so many high-tech processes, relying heavily on low-tech manual interventions, which are regularly influenced by the human factors that manifest mistakes. Unless, of course, you count war."
That said, why have the reports not shown any reduction in adverse events in the hospitals over the past decade?
I called up Carolyn M. Clancy, the no-nonsense internist who has been the director of the federal Agency for Healthcare Research and Quality since 2003. I asked her if our hospitals are any safer now than they were 10 years ago.
She acknowledged the challenge of reducing the high numbers of adverse events but added that some of the more recent numbers are the result of better tracking and expanded definitions of which events are preventable. For example, a decade ago certain hospital-acquired infections were considered side effects to usual care, but now we know that nearly all can be eliminated. "We are safer," Clancy said.
I see what she means.
A detailed look at the North Carolina hospital study shows that although the total number of harm cases remained the same, there was a decreasing trend over five years in the number of preventable harms.
And in a paper I published in 2006 with Donald Berwick, who is now head of the federal Centers for Medicare and Medicaid Services, we showed how we reduced hospital-acquired infections at my hospital by 50 percent. (This was in collaboration with the nonprofit Institute for Healthcare Improvement, then led by Berwick.)
Take another example: invasive hospital-acquired MRSA infections.
MRSA is a staph bacteria highly resistant to antibiotics that spreads through hospitals and nursing homes. The Centers for Disease Control and Prevention recently reported that since 2005 the rate of such infections in hospitals has declined, on average, by 10 percent per year.
Yet another study, conducted between October 2007 and September 2008, showed that the use of checklists led to a decline in surgical complications.
We are seeing pockets of success in arenas of patient safety where there is intense, focused activity, such as MRSA and surgical infections prevention.
At the same time, patient safety activity has broadened with new tools and policy. Clancy told me that her agency has developed "Common Formats," which provide uniform reporting standards for hospitals that tally adverse events. We can improve only what we measure.
Also, many states now require hospitals to publicly report certain adverse events, such as hospital-acquired infections. This gets the attention of the consumer advocacy groups. There is also pressure to prevent drug companies from giving new drugs names that are confusingly close to drugs that are already on the market, another common source of error.
And lastly, as of September 2008, the Centers for Medicare and Medicaid Services has stopped paying hospitals for "never" events - adverse events that should never occur, such as surgery on the wrong limb or infections from urinary-tract catheters. This has many heads of hospitals following rates of bloodstream and urinary-tract infections as closely as they follow their budget line items.
But, to make hospitals really safe, we need to take a leap. We need to build a culture in which patient safety is the priority not just for the quality improvement director but also for every nurse, doctor, administrator, aide, housekeeper, dietary worker and hospital board member.
Drivers still make mistakes with dreadful consequences, but over the past decades we've created a culture that is very well aware that traffic safety is imperative. Likewise, a culture - not just a strategy - of patient safety is a must.
Jain is an infectious-disease specialist in Memphis and an adjunct assistant professor at the Rollins School of Public Health at Emory University in Atlanta.