Comment by tokenadult | Hacker News Reader

Comment by tokenadult | Hacker News Reader

From the submitted article:

"Nothing makes hospitals shape up more quickly than this kind of public reporting. In 1989, the first year that New York's hospitals were required to report heart-surgery death rates, the death rate by hospital ranged from 1% to 18%—a huge gap. Consumers were finally armed with useful data. They could ask: "Why have a coronary artery bypass graft operation at a place where you have a 1-in-6 chance of dying compared with a hospital with a 1-in-100 chance of dying?"

"Instantly, New York heart hospitals with high mortality rates scrambled to improve; death rates declined by 83% in six years. Management at these hospitals finally asked staff what they had to do to make care safer. At some hospitals, the surgeons said they needed anesthesiologists who specialized in heart surgery; at others, nurse practitioners were brought in. At one hospital, the staff reported that a particular surgeon simply wasn't fit to be operating. His mortality rate was so high that it was skewing the hospital's average. Administrators ordered him to stop doing heart surgery. Goodbye, Dr. Hodad."

So we have a known model that improves results. Let the patients who decide which hospitals to visit know beforehand how hospitals compare in getting good results for patients. Give patients power to shop. That provides incentives for hospitals to do better, and nudges hospital managers to do what is necessary to win the trust of informed patients.

This kind of reform would be very good to apply to schools as well. Already, one state in the United States that is conspicuous among all fifty states in educational achievement (the state I live in) allows all residents of the state to enroll in the public schools of ANY school district anywhere in the state. (The school district I live in has inbound open-enrollment students from the geographical territories of forty-one other school districts.) In general, the power to shop is the incentive factor to bolster by government regulation, letting each consumer decide what trade-offs are important. That works considerably better in driving improvement than regulating outcome goals, and better still than regulating inputs into the provision of the important service being regulated.

Real data and informed consumers are a very valid way to promote improvement through "competition".

Often times though, that's not enough as is the case in LA right now where patients are being prevented from using the best doctors:

Two of the most prestigious names in Southern California healthcare — Cedars-Sinai and UCLA — are getting shut out of a major insurance plan for being too expensive.

In a bold cost-cutting move, Anthem Blue Cross has eliminated doctors affiliated with the hospitals from a health plan offered to about 60,000 employees and dependents at the cash-strapped city of Los Angeles.

The city opted for Anthem's plan because it will save $7.6 million in annual premiums next year by excluding physicians from the two institutions known for tending to the Southland's rich and famous. About 2,200 city workers and family members are expected to lose access to their doctors under the plan.

This dramatic step shows that even some of the most-respected names in medicine can't get by on reputation alone at a time when the U.S. is grappling with a $2.6-trillion healthcare bill annually. Major hospitals and medical groups face growing pressure to justify their charges. And employers increasingly are willing to risk the ire of workers by cutting popular providers to clamp down on costs.

- http://www.latimes.com/business/la-fi-hospital-costs-2012092...

Reporting death rates does give hospitals an opportunity to reduce them, yes, and one way they might that is to get rid of unsafe surgeons and fix unsafe practices. Unfortunately that can only reduce death rates so far - they're also affected by the number of high-risk cases the hospital takes on. So it turns out that reporting these statistics gives hospitals a financial incentive to turn down, drive away, or otherwise prevent and discourage high-risk patients from obtaining medical help.

That's how home births get their low mortality rate: Midwives (luckily) just send all the high risk cases to hospital, and even a cancelled home birth where the actual birthing happens in the hospital would count towards the hospital's mortality rate.

the problem is this method encourages hospitals and doctors to avoid taking the difficult cases.

lets say you have a complicated problem which only has a 25% success rate. but due to these statistical models, nobody will do the procedure for you, since it will skew their results. what do you do?

Can you not simply modify the way in which you calculate the statistics?

Since doctors tend, as I understand it, to state the expected success rate to patients before surgery - this could be used as a factor in the reporting.

A more confident surgeon might be willing to quote a higher success rate, which would attract more patients (or patients willing to pay more) but would be more punishing were he shown to be incorrect.

Obviously this is very vague, and whilst I work with statistics I don't deal in medicine. But the difference between a surgeon/hospitals success rate and the expected success rate seems like a good measure of competence?

Reply on news.ycombinator.com