How Do You Put a Yardstick on Safety?


Making patient safety one of the elements of the Best Hospitals methodology has been a goal for us and our resourceful and skilled contractor, RTI International, for several years. We're working on it and getting closer. But whatever data go into the mix have to meet reasonable standards, and so far the existing safety data are ungettable or skimpy. Or they fail the "face validity" test, as statisticians put it, meaning that if you give the numbers a common-sense once-over, they fall apart.

An example of mostly ungettable information is the rate of hospital-acquired infections that patients pick up during their stay. Many are transmitted from patient to patient by caregivers who don't clean their hands or whose gleaming white coats are filthy with bacteria, as Betsy McCaughey's horrifying article lays out in the new edition of Best Hospitals. Every hospital's infection rates, broken down by type of bug and by source (after surgery or from a central intravenous line, for instance), should be publicly available. Why aren't they, other than in Pennsylvania, where lawmakers stood up to hospitals' attempts to kill the disclosure requirement before it became law?

As for flunking the common-sense test, I'd like to introduce you to decubitus ulcers, more familiar to us as bedsores or pressure sores because they are most common in those who are bedridden or in a wheelchair and are in one position for many hours. The continued pressure on a particular patch of skin stops blood from flowing, and the skin breaks down. A bedsore can work all the way down to the bone, destroying tissue and giving bacteria easy access to the bloodstream. A hospital should be able to prevent bedsores by turning patients frequently and using special beds, mattresses, and cushions.

Now, a few years ago, an arm of the federal government called the Agency for Healthcare Research and Quality, which has the mission of making the unwieldy blob we call healthcare safer and better, put out a list of Patient Safety Indicators. These PSIs are things everyone agrees shouldn't happen in hospitals, such as leaving sponges inside the body after surgery and allowing complications to occur from anesthesia. The current version is 27 items long.

One of the 27 on the list is decubitus ulcers. The problem is that many patients susceptible to bedsores, such as the elderly and infirm, were admitted from long stays at home or in extended-care facilities and may have developed the beginnings of a bedsore before they came in. If it wasn't caught when they were first examined—the first stage isn't always easy to see—then anything after that becomes the hospital's fault. That doesn't wash.

That is one of many reasons that we are still sifting through the PSIs and other indicators, looking for those that are both telling and fair. It shouldn't be too long before you see some results.

Next: A hospital killed a patient. Why is it still in the rankings?