I’ve been getting emails about the New York Times piece and my quotation that the penalties for readmissions are “crazy.”  It’s worth thinking about why the ACA gets hospital penalties on readmissions wrong, what we might do to fix it—and where our priorities should be.

A year ago, on a Saturday morning, I saw Mr. “Johnson,” who was in the hospital with a pneumonia.  He was still breathing hard but tried to convince me that he was “better” and ready to go home.  I looked at his oxygenation level, which was borderline, and suggested he needed another couple of days in the hospital.  He looked crestfallen.  After a little prodding, he told me why he was anxious to go home: his son, who had been serving in the Army in Afghanistan, was visiting for the weekend.  He hadn’t seen his son in a year and probably wouldn’t again for another year.  Mr. Johnson wanted to spend the weekend with his kid.

I remember sitting at his bedside, worrying that if we sent him home, there was a good chance he would need to come back.  Despite my worries, I knew I needed to do what was right by him.  I made clear that although he was not ready to go home, I was willing to send him home if we could make a deal.  He would have to call me multiple times over the weekend and be seen by someone on Monday.  Because it was Saturday, it was hard to arrange all the services he needed, but I got him a tank of oxygen to go home with, changed his antibiotics so he could be on an oral regimen (as opposed to IV) and arranged a Monday morning follow-up.  I also gave him my cell number and told him to call me regularly.

Much of the weekend went smoothly.  When I talked to him on Sunday morning, he reported having slept poorly but had a joyful tone in his voice that I never heard in the hospital.  He was planning on having a few beers with his son and watching the Patriots game.  I told him to take it easy on the beers.

Sunday afternoon, I caught him during half-time and he assured me everything was fine.

On Monday morning, I got a call that Mr. Johnson was back in the hospital. I rushed to his room to see him lying in bed, looking sad.  He told me that his breathing had gotten worse overnight and at 3 a.m. his son drove him to the hospital.  His vital signs looked fine, although his oxygenation was a little worse than Saturday.  He screwed up, he said, and told me I’d been right.  He should not have gone home.  I asked if he had enjoyed the weekend, and his face lit up.  He had loved it.  Let’s be clear: he had been right to go home. There was no screwup.  We had gotten him a weekend at home with his son, who would soon be heading back to Afghanistan.

In 2012, more than 100,000 Americans will die in U.S. hospitals because of medical errors such as preventable infections, receiving the wrong drug, or having the wrong surgery.  Even more Americans will likely die because they failed to get simple therapies like the right antibiotic for their pneumonia.  Millions of people will report suffering in the hospital from undertreated pain or the indignities of not being always treated with respect.  Yet the Affordable Care Act says that my “mistreatment” of Mr. Johnson—sending him home and having him come back—was far more egregious and deserves the biggest penalties.  While the ACA is extremely important in improving access to millions of Americans, several of the provisions to improve the “delivery system” are not quite right.  The notion that readmitting people to the hospital is worse than killing them due to medical errors?  Sorry, but that is crazy.

The Leapfrog Group will soon be putting out another report of patient safety in U.S. hospitals (I’m on their advisory panel).  It will provide letter grades on the state of safety of every hospital.  The grading system is not perfect—primarily because hospitals are not required to report their rates of medical errors. Yet Leapfrog soldiers on, trying to make their best assessment.  I wish Medicare would make patient safety half as much of a priority as reducing readmissions.  Oh, and by the way?  Reducing medical errors can likely save us a lot more money than reducing readmissions—so even if we do it for the money, that should be our target.

So—should we penalize hospitals for readmissions?  I think it’s probably fine (although we should know that we will primarily end up penalizing hospitals that care for the sickest and poorest patients).  But by putting so much energy on readmissions and so little on patient safety, we have made our priorities clear, and I think they are the wrong priorities.

If my hospital had made my readmission rate part of my performance evaluation, would I have sent Mr. Johnson home that weekend? Maybe not.  I could have easily strong-armed him into staying, and he would have listened.  He was, what we call, a “compliant” patient.  But if we had kept him in the hospital, he would have lost the chance to watch the Pats game with his son.  His son and family would have lost having the weekend with their dad and husband.  But I would have “won,” coming across as a better doctor for having a lower readmission rate.

Policies have consequences.  They set up subtle, often perverse incentives.  Before we decide that readmissions are the biggest priority for cleaning up American hospitals, we should ask whether Mr. Johnson should have been sent home that weekend.

Ashish Jha is an Associate Professor of Health Policy and Management, Harvard School of Public Health.  He blogs at An Ounce of Evidence and can be found on Twitter @ashishkjha.