“No problem,” I told the intern over the phone. I had done over a hundred of them previously without a single complication.

The patient turned out to be slender, with purplish skin, fine threadlike smoker’s wrinkles around his mouth and cheeks and a chest that looked as wide and rounded as a wine barrel. He had emphysema; and as I stood agape at his doorway, all I could imagine were a pair of enormous hyperinflated lungs billowing dangerously around my vein of interest, the subclavian. There was a good chance that I would prick his lung with my needle and collapse it while trying to access his vein. But I couldn’t refuse; the man desperately needed the central line for his medications.

I pushed the thoughts out of my head and stepped forward to prepare for the procedure. As I slid the large needle under his skin, I remembered how I had done exactly the same so many times before. I found my spot on his chest below his shoulder, pushed the needle through until it touched the collarbone, “walked” the needle down the bone, then slid it slowly underneath while drawing back on the attached syringe, waiting for the dark red blood to rush back indicating that I had hit my target.

But the blood did not appear. Instead, I felt something else: a small, nearly imperceptible poof of air.

I had pricked his lung.

Devastated, I withdrew my needle and quickly took steps to confirm, then care for, his punctured lung. But a few days later in the I.C.U. when one of the heart surgeons asked me to place a central line in another patient, I couldn’t help but hesitate. He repeated himself and then I confessed. I had lost my nerve with this once seemingly straightforward procedure.

“These things happen to the best of us,” the surgeon said, smiling. He looked around the I.C.U., his blue eyes resting on several of his own patients, individuals others had refused to operate on but who were now recovering and expected to go home thanks to him. “Remember this, Pauline,” he said turning back to look at me again. “If you play with fire, you’re going to get burnt.”

Over the course of my training, I would repeat and recall that senior surgeon’s advice more than a few times. But three years ago, when officials at the Centers for Medicare and Medicaid Services (C.M.S.) made the decision to deny payment for certain complications, I thought of those words again.

In August 2007, officials at the agency announced that Medicare and Medicaid would no longer pay for the treatment costs of “conditions that could reasonably have been prevented” in an effort to improve patient safety and rein in health care costs. Instead, hospitals and physicians would have to take responsibility for these errors and cover their own costs. The group’s decision was based in part on work previously done by a national coalition of health care safety and quality experts. These experts had assembled a list of complications so egregious that they called them “never events,” as in they should never occur. “Never events” included complications stemming from operating on the wrong side of the body to leaving instruments in a patient after a procedure.

While the C.M.S.’s list has always included several of the original “never events,” it has since grown to include complications that were “never” mentioned by the health care experts. Over time, too, other insurers have begun adopting similar policies of refusing to pay for treatment costs of certain complications, cobbling together their own versions and definitions of “never events.”

While the original list is horrifying to read, these newer lists from C.M.S. and other payers fail to strike the same chord. The original complications derived their emotional power because they were truly “never events,” both devastating and eminently preventable. But in the newest lists compiled by C.M.S. and other insurers, the emotions are gone. Instead one is left wondering: are these complications, as the payers assert, really preventable?

The answer, according to a study published earlier this year in The Archives of Surgery, is no. Researchers from the health care think tank Michael Pine & Associates in Chicago analyzed data from almost 900,000 cases in over 1,000 hospitals and found that patient characteristics could have a marked impact on the rate of several of these payer-specific “never event” complications. Pre-existing conditions like diabetes, chronic lung disease or malnutrition rendered patients more susceptible to certain complications no matter how much the physician intervened. In fact, with mathematical modeling the researchers were even able to predict the degree of susceptibility.

“There are many things we can be doing to improve quality,” said Dr. Donald E. Fry, lead author of the study and executive vice president for Michael Pine & Associates. “But there has to be the understanding that high-risk patients can affect the rate of certain complications.”

Dr. Fry and his co-investigators found, for example, that patients with malnutrition or kidney failure were more likely to develop severe wound infections after heart bypass operations. This likelihood leaves surgeons who initially evaluate these patients for surgery in an ethical Catch-22. If they decide to operate and care for them through any complications, they may not get reimbursed for the effort. But if they don’t operate, they knowingly turn away patients who cannot survive for long without some type of operative intervention. “We don’t object to the idea that physicians and hospitals need to be accountable for what happens,” Dr. Fry said. “But regardless of how good you and the technology are, there are adverse events that relate to the vulnerability of the patient.”

As a result, physicians may be tempted to “cherry-pick” patients, opting to work with only the healthiest individuals. “I have had the experience that safety net hospitals sometimes get patients who are so far down the road because a certain selection process has gone on,” said Dr. Fry, who practiced surgery for several decades and was previously chairman of surgery at the University of New Mexico in Albuquerque. “I have grave concerns that with an aging population, more patients being enfranchised and a health care infrastructure and workforce that are on the edge, you may very well see cherry picking and denial of care.”

What Dr. Fry and his co-investigators propose is that C.M.S. and other insurers adopt a warranty policy of payment. The warranty would be an all-inclusive upfront payment that would take into account the probability of having a complication, as well as the patient’s prior condition; and it would be standardized according to what the best quality hospitals pay for the same diagnosis, procedure and complications. “We have the mathematical ability to look at large populations and predict the rate of these complications,” Dr. Fry noted. “Our payment systems need to recognize that.” When doctors and hospitals rack up more costs than are covered by the payment, they lose money. “Physicians and hospitals will then have a real incentive for quality.”

Dr. Fry added: “I find it offensive that someone wants to pay you for what is right. We’re supposed to be doing what is right all the time. We just can’t be accountable for things that are beyond the scope of what we can do at the present time.

“What all of us want is for quality and efficiency to dictate expectations, not politics and policy decisions.”