The GiveWell Blog

Good vs. better

I recently read Better by Atul Gawande, and found myself particularly struck – and reminded of our own situation – by his analysis of hospital care.

According to Dr. Gawande, conventional wisdom has long been that the vast majority of hospitals provides top-notch, quality care, and only a fraction treats their patients incompetently. This implies that the most important thing a patient can do is weed out incompetent hospitals – but worrying about “average” vs. “exceptional” isn’t worth it. But a systematic study found otherwise.

The Cystic Fibrosis Foundation had long monitored the well-being of CF patients at hospitals around the country. When they evaluated the data, they found that quality of care and patient outcomes, such as life expectancy and quality of life, varied more than they had expected. A relatively small group of hospitals provided low-quality care, but another group provided top-notch care. Their patients lived longer and lived better. (The foundation also found that the vast majority of hospitals fell in the giant, indistinguishable middle, providing average care that was neither incompetent nor excellent.)

The differences in impact weren’t just academic. The evaluation found that average life expectancy for someone diagnosed with CF was 30 yrs, but at top hospitals, life expectancy was significantly longer, averaging 46 yrs. While alive, patients at average hospitals had lower quality of life than those without CF because they had lower lung function and consequently couldn’t participate in a host of normal activities. At top hospitals, patients’ lung function was equivalent to those without the illness. The differences were real and they were stunning.

Recognizing the significance of the results, the Cystic Fibrosis Foundation did the only thing they could: they made all the information public. Patients should know which hospitals provide best care, shouldn’t they? Doctors should know which methods work best, shouldn’t they?

Critics feared that demand for care at the best hospitals would outstrip the hospitals’ ability to provide it, but instead something else happened. Care improved across the board as hospitals’ staff met and implemented the practices of those at the top.

The reported conventional wisdom on hospitals reminds me of the conventional wisdom we constantly hear expressed about charities: that the vast majority of them do great work, and that it’s important to weed out the “frauds” but that distinguishing between “legitimate” charities amounts to nitpicking. Our instinct, however, is that charities are like hospitals, companies, and any other set of highly complex organizations that vary in their people and approach to difficult problems. Our instinct is that in charity as in most other things, the difference between “best” and “average” is at least as important as the difference between “legitimate” and “illegitimate” – and that starting to examine the differences, publicize them, and push for the best may lead to improvement across the board.


  • Mal Content on May 22, 2008 at 5:53 pm said:

    Elie, you’ve confused association with causation. Longer survival at one hospital does not mean that hospital’s care is better. My guess is that you did not go to graduate school yet to learn the difference.

  • Ryan Walsh on May 22, 2008 at 7:20 pm said:

    I really liked this post.

  • Grant on May 23, 2008 at 1:23 am said:

    Great post, Elie!

  • rob s on May 24, 2008 at 3:19 pm said:

    without a detailed reading, it seems quite possible that the CF patient study lacks validity to prove causality somewhat satisfactorily.

    however… its up to the critics to HELP identify other variables that might be influencers and test for them.

    the SAME should be done for Charity outcomes. as for requiring grad school to understand causality, i’ll refrain from comment.

  • Ullysses on May 24, 2008 at 10:27 pm said:

    Gosh. Maybe your BEST post ever, Elie.

  • Erich Riesenberg on May 29, 2008 at 4:42 pm said:

    You guys still suck.

    This post is absurd. Especially this: “The reported conventional wisdom on hospitals reminds me of the conventional wisdom we constantly hear expressed about charities: that the vast majority of them do great work, and that it’s important to weed out the “frauds” but that distinguishing between “legitimate” charities amounts to nitpicking.”

    I don’t ever recall reading anywhere that it is not worth the time to try and identify the best non profits. The comment you linked to criticized you for the specific questions you asked of a non profit, that you were asking stupid questions, not for the goal of tryng to make an informed choice.

    I do recall people, including myself, ridiculing you for claiming that weeding out poorly run non profits is not worth the effort.

    You people really should put a cork in it and get on with your lives.

  • Sean Stannard-Stockton on May 29, 2008 at 8:32 pm said:

    Elie, I’d like to post further about your thoughts later, but this is simply one of the best posts either of you have ever written. I have Better on my bookcase, but haven’t read it yet. You’ve just forced me try and find time to read one more thing!

    Seriously, the shift from from weeding out fraudulent charities to identifying high-impact charities is one of the most important trends in philanthropy. I’d never seen the concept expressed so well before. Thanks for keeping at blogging even with so many critical comments.

  • S Puppet on May 30, 2008 at 4:52 am said:

    Great essay, Elie. A+


  • John J. on May 30, 2008 at 7:04 am said:

    Great post, huh? This gang again misunderstood criticisms made of it, and the post is just a summary of someone else’s good book.

  • Erich Riesenberg on June 4, 2008 at 11:36 am said:

    Sean is such a tool. Seriously.

    Divert and deflect, is that what you have taught each other?

Comments are closed.