A commentary in the just-out issue of the Journal of the American Medical Association caught my eye today. I'll come back to the subject—the logic of withholding Medicare payments for treating certain preventable hospital-acquired complications—in a future post. I'm mentioning it now because of an observation by the authors about what happens in healthcare when well-meaning people make changes: "Unintended consequences are the norm rather than the exception."

Earlier in the day I'd run into an interesting example in long post in Paul Levy's "Running a Hospital" blog. Last spring, Beth Israel Deaconess Medical Center, where Levy is president and CEO, introduced a new approach to rescuing patients who might be on the verge of a sudden decline. Signs that all is not well aren't obvious. It's more like hints—they may be breathing faster, or their blood oxygen or urine production may have dropped. The hospital's "Triggers" program authorized a nurse to bring in help if any of a set of clinical triggers (such as a heart rate higher than 130 beats per minute) goes awry, or even if the nurse simply has an uneasy feeling about a patient—"marked nursing concern" is what the program calls it.

In its first year, the Triggers program has had remarkable success. "The number of 'codes' on our floors," writes Levy, "has gone down so dramatically that residents now need to practice emergency resuscitation mainly in the simulation center because so few actual patients need it." Hmm. There's an unintended consequence. But, as he says, what a lovely problem to have....

I noted in my story on medical travel online and in this week's issue of U.S. News that more Americans are boarding flights to have bypasses, joint replacements, and other needed procedures performed at hospitals in India, Thailand, and other far-flung destinations. How many, as I also wrote, is several notches below uncertain. Figures as high as 500,000 U.S. patients per year have been tossed around.

Yesterday at a major medical travel conference in Las Vegas, consultant McKinsey & Co. reported that the number is more like 60,000 to 85,000. That's more like travel writer Josef Woodman's rough guess of 50,000, which appeared in my article.

Turn up the volume, right? For years, experts have believed that in general the more often a hospital does a procedure, the better the outcome. For some procedures, specific cut points have become accepted—a recommended minimum of 500 a year for coronary artery bypass surgery, for instance. Now comes a study in this week's Archives of Surgery claiming that over the years, lower-volume hospitals have gotten to be just about as good at heart bypass surgery as the big bypass factories.

The researchers looked at a large sample of the 1.1 million bypass surgeries performed between 1988 and 2003. They found that by the end of the period, differences in death rates at high-volume behemoths that perform 500 or more bypasses per year, at somewhat less busy hospitals that do from 250 to 500, and at the lesser lights that perform 249 or fewer were trivial. Death rates at the low-volume centers, in fact, had declined more sharply than at hospitals in the two higher-volume groups—from 5.9 percent to 3.5 percent.

Last week I turned this space over for a day to Karen Madsen, an advanced-practice nurse and nursing instructor. As "Sister Nurse," a regular columnist for the Stressed Out Nurses Web page, she had written about the good, bad, and ugly of her 15-year-old daughter's recent hospitalization with acute appendicitis. Madsen described the insensitivity of the front-desk ER nurses. She also told of her own failure to pick up evident signs of Grace's illness, the warmth and skill of the clinical nurses, and her demanding, sometimes over-the-top behavior as her daughter's parent and advocate. I was struck by the honesty and tone and posted a slightly compressed and edited version.

It's drawn quite a reaction, mostly from other nurses (as well as some 70 comments on allnurses.com, an online gathering place for the nursing community). Many who weighed in here echoed Madsen's observations. Others were biting. Her reply:

Sunshine not only fights infection—sunshine laws push data out of the shadows into the public arena, where it belongs. So let's salute yesterday's release of the second annual "Hospital-Acquired Infections in Pennsylvania." It's a 98-page report on how well in 2006 the state's 165 acute-care hospitals, each one named, kept patients from being infected while being treated within hospital walls. It is illuminating, if you'll excuse one more sunshine reference.

• More than 3,700 patients with hospital-acquired infections died in 2006. Hospital-infected patients died at a rate nearly six times that of uninfected patients—about 1 in 8, compared with 1 in 50.

Who better than a nurse, especially a nursing instructor, to judge the compassion and skill a hospital brings to patient care? And if the nurse is the mother of a teenage girl brought to the ER in agonizing pain, what she has to say is definitely worth reading.

As "Sister Nurse," Karen Madsen writes a monthly column for Stressed Out Nurses, an arm of hospital consultant HCPro. She is an assistant professor at Cox College of Nursing and Health Sciences in Springfield, Mo., and she is the mother of 15-year-old Grace. I was informed and moved by Madsen's account and offer it to you (edited slightly to be shorter than her two-part opus) as something more than just another cautionary tale:

I'm a little surprised more hospitals that excelled in last Friday's initial release of the government's surveys of hospital patient satisfaction, which I described Friday afternoon here and here, haven't fired up their marketing bandwagons. "Customer relationship management," the buzz term for building consumer loyalty, has become just as important to hospitals as it is to other corporations of any kind or size. But the Society for Healthcare Strategy & Market Development, an arm of the American Hospital Association, doesn't even headline the survey on its home page, and my quick Google search using combinations of "hospital," "patient," "satisfaction," and "survey" didn't produce a gush of announcements.

My confessed bias is to limit the weight these survey results are given when selecting a hospital, so maybe the absence of puffery about happy patients is a good thing.

Incidentally, I owe thanks to Suze, who questioned my claim that there was 100 percent participation from all of the hospitals in Baltimore and San Francisco and within 25 miles of our Washington, D.C., offices. She found some in all three places that did not submit patient satisfaction data, and she is correct on all counts—having taken another look, I can report that there are six opt-outers in the D.C. area alone. That's what can happen when my fingers move faster than my thoughts. Embarrassing.