Preventing Hospital Errors Part II by Howard C. Berkowitz
Given the medical community’s reluctance to step up and admit to mistakes, Medicare has decided to get tough, saying that it will stop reimbursing hospitals for the thirteen adverse advents listed below. Before discussing the list, let me suggest that not all of these events are within a hospital’s control. I’ve rated the mishaps on the list from 1 to 4, with “1” indicating something that, I agree, should never happen, and “4” referring to something that, in my experience, a hospital may not be able to prevent.
The 13 Things That Should Never Happen in a Hospital
1. Catheter-associated urinary tract infection [2]
2. Bed sores [1]
3. Objects left in [THE PATIENT”S BODY] after surgery [1]
4. Air embolism, or bubbles, in bloodstream from injection [1]
5. Patients given incompatible blood type [1]
6. Bloodstream Staphylococcus (staph) infection [2]
7. Ventilator-associated pneumonia [2]
8. Vascular-catheter-associated infection [2]
9. Clostridium difficile-associated disease (gastrointestinal infections) [3]
10. Drug-resistant staph infection [3]
11. Surgical site infections [3]
12. Wrong surgery [1]
13. Falls [4]
IF YOU’D LIKE TO
While the events labeled “1” are potentially fatal mistakes that should never happen, I’m still not sure that financial penalties are the answer. If Medicare refuses to reimburse for badly infected bedsores, air bubbles, a blood mix-up, or the occasional missing scalpel, won’t that only encourage hospital personnel to cover up their blunders?
If the oversights are never reported, procedures won’t change to avoid serious errors. Here, hospitals need “no-fault” reporting ( like the no-fault reporting in the airline industry that I described in my first post on hospital errors)—and laws that address malpractice concerns. If hospital staff slipped up or neglected the patient, the hospital must pay. But voluntary reporting should not be used by malpractice lawyers to go on a fishing expedition. And there should be some reward for the physician or hospital willing to confess to the error, up front. After all, such openness can save everyone (including the plaintiff) a fortune in lawyer’s fees.
In the meantime, healthcare technology could address some of these problems. To avoid operating on the wrong patients, for example, hospitals should give patients electronic ID tags that can be matched to a corresponding market on the treatment.
When it comes to giving a patient the wrong blood type, the consensus is that this can be prevented by matching a patient’s electronic label with a label on the unit of blood (or blood product). I first worked on developing prototype equipment in 1970, when we just weren’t able to do it.
Today, we have commercial equipment that is user-friendly and reliable, but implementing it hospital-wide is a significant investment. Mass General did it, but not every hospital can afford the technology.
In cases where a non-profit hospital is operating in the red, who pays the up-front costs? Perhaps private insurers should contribute to funding since, in the long run, they would reap the savings. Liability insurers might also chip in, since systems that reduce errors would cut their costs. Finally, of course, hospitals that are operating in the black need to begin making capital investments in technology that will keep patients safe. Too often, hospitals invest any surpluses in cosmetic improvements designed to attract well-heeled patients (and their insurance.) But while waterfalls and newly decorated waiting rooms are nice, safety is clearly more important.
As for missing instruments, someday we may develop a system that provides 100 percent electronic counts of surgical instruments before sewing the patient up
Automated systems also can page a nurse or aide to turn the patient and check for bedsores every so many minutes, on a schedule to prevent sores. The system can confirm that the caregiver went into the patient room, and, if one doesn’t show up, escalate the care request. It would also log a complete failure. These tools aren’t cheap, because they requires caregiver sensors in every room and both pagers and “beacons” on every caregiver, plus the centralized system and the equipment at nursing stations.
I’ve lumped together certain infections, plus ventilator-associated pneumonia and given them a “2” because it is not always possible to prove that these problems were caused by hospital error. If a catheter or ventilator tubing can be cultured and the organism found on it, that’s a clear error. But do we really want to start incurring the substantial cost of saving and culturing all catheters?
When it comes to the infections that I have labeled “3, it’s quite possible that the patient may have been carrying the germs when he or she was admitted—even though the symptoms did not become apparent until a day or two later.
Finally I’ve given patient falls a “4” because when a patient winds up on the floor, it may or may not be the hospital’s fault. Thinking about things I’ve done as a patient, I can’t always blame the hospital. Assuming the nurses put up the bed rails, a patient who is quite sick can’t release them. I’ve had times, though, when I felt much better, couldn’t get a nurse to answer a page, was able to unlock the rails, and then, trailing my IV pole, go to the bathroom. I didn’t have a problem. Other patients might. Of course having adequate staff (which many hospitals don’t have) could address this problem.
In my last post on hospital errors, I discussed what we might learn from the airline industry. Let me conclude this post by suggesting what Medicare might learn from Germany about preventing errors.
Medicare proposes penalizing hospitals for mistakes on a per patient/per error basis. By contrast, while Germany has a tightly regulated health care system, it does not attempt to use financial penalties to micromanage hospital care. Instead, the German National Institute for Quality Management in Health Care (BQS) creates a representative list of over 20 diagnoses and procedures, and has an expert panel establish 10-15 quality metrics for these interventions. They collect data, quarterly, from hospitals at the regional and national levels. (BQS pays the hospital a fee for the required documentation, and penalizes the hospital if it fails to produce the documentation.)
Regional groups of experts evaluate the hospital results, although they do not know which hospital is associated with which group. Hospitals that are especially good or bad statistically are asked for the reasons. The best hospitals’ practices are shared in a positive manner, while the low-scoring hospitals participate in a quality improvement program, which is done, not in a punitive but in a “collegial” manner.
Hospitals publish their statistics on the Internet every 2 years, and the good ones use these results for marketing.
Thus hospitals have an incentive to improve, and rather than simply punishing them for mistakes, the system helps them make progress. The U.S. government also has some programs that offer a carrot (rather than a stick) for quality improvement, but none as comprehensive as Germany’s.