Monday, October 4, 2010

Medical Errors

To err is human, but to forgive is divine. That is the common Alexander Pope saying. It essentially means that it should not be surprising if you screw up, but expect to be crucified if you do. This is no more true than in medicine. Medical errors accounted for about 195,000 deaths a year nationwide in 2002. That number seems to be holding steady even in recent years. There are numerous ways in which a medical error may result or contribute to a patient’s death, including incorrect medication selection and dosing, incorrect surgical site and retained surgical instruments, missed medications, and errors or complications during procedures. Now 190,000 is not a tremendous number when you consider the number of people who pass through hospitals each year, including admissions, ER visits, same day surgery procedures, and infusion room visits. But who wants to be one of the 190,000?

So every year someone comes up with another bright idea to eradicate the medical error problem. It would seem logical that you should have a hospital that has no medical errors, right? So you simply have to remove all the factors that make errors more likely to occur.

First were the abbreviations. Since there is a lot of Latin in medicine, many dosing frequencies are abbreviated using the Latin instruction—“qd” for daily, “qod” for every other day, “ or “qid” for four times a day. Also, non Latin abbreviations, such as, “u” for units, “MSO4” for morphine sulfate, “MgSO4” for magnesium sulfate, and so on. There were also issues with numbers and where the decimal point and places after the decimal point were concerned, such as, “.1” instead of “0.1”, or “1.0” instead of “1”. You can see how each of these instances could result in a medication error. So they were done away with, and everything had to be written out. There was little aftermath from this change because the electronic medical record and computerized physician order entry came later, and described below.

Second was the fatigue factor. Researchers took volunteer subjects and kept them awake for thirty plus hours and then asked them questions to see how many they would get incorrect. When compared to subjects that were not sleep deprived, the sleep deprived subject made more mistakes. So it then had to follow that if you had a resident on call, he or she would have been up that morning, up that night, and up the next morning until at least 5pm. The propensity for medical errors should be enormous. So the edict was sent out that no resident on call should work no more than 80 hours a week, *and* no shift should exceed 30 hours, *and* there should be 10 hours off between shifts. And while it was state that this would be a voluntary recommendation, no program would be allowed to remain accredited unless it complied.

There are, however, some things to take into account. Since work hour limitations have been implemented, there has yet to be shown a clear decrease in hospital mortality rates. The other issue that will be difficult to determine is that if you reduce a medical trainee’s work hours, you reduce his or her exposure to the medical field. Without an increase in duration of training, you will be left with more undertrained physicians over time. This coupled with the fact that patients are sicker and carry more medical comordities these days results in a precarious healthcare scenario.

Third it was the handwritten medical record. It used to be that everything was written—notes, prescriptions, orders, medication records. But with the classic joke of a doctor’s poor penmanship, this was an easy target. The medication records were the first to be digitized for the pharmacists and nurses. Then physician orders were digitized. Now the remainder of the medical record is set to be digitized. The only problem with the electronic medical record (EMR), is that there is no standard. As a result, innumerable software companies have developed their own EMR system to hawk to healthcare facilities everywhere. None of the EMRs are compatible with each other, so in order to import information from one system to another you end up with a PDF file of the printout from the other EMR. So while some errors from poor penmanship are avoided, you now are left with lost information that may otherwise impact care.

Fourth, the Centers for Medicare and Medicaid Services (CMMS) require hospitals to report medical errors. These reports are reviewed and compensation is linked to the error rate of a hospital. Not a bad idea, right? What is not advertised is that CMMS *expects* a minimum error rate, and if the hospital is below that rate CMMS assumes there is underreporting. Of course, it could not be because the hospital actually has a lower rate of medical error. The hypocrisy of Medicare continues. And the fact that there is a minimum error rate tells you that it is there only to look for potential underreporters of medical errors.

So to summarize, medical errors account for about 200,000 deaths a year, and costs about $8 billion a year. Look at the time and money we have thrown into medical errors, with legislation, lawsuits, and wasted money on nonstandard software. On the other hand, tobacco use results in 440,000 deaths a year and costs about $97 billion a year.

But then again, smokers make up 21% of eligible voters, and physicians only 0.1% of eligible voters. And do not think I will not make a Venn diagram to account for physicians that smoke, because I will.