Monday, March 26, 2012

Get the data and then use it right

Andy Oram from O'Reilly Communications offers some thoughtful observations in this post, entitled, "Five tough lessons I had to learn about health care."

Let's start with a headline version of the five lessons (for you can read the full explanation on Andy's site):

1. The health care field will not adopt a Silicon Valley mentality;
2. Regulations can't force change;
3. The insurance companies are not the locus of cost and treatment problems;
4. Doctors don't want to be care managers;
5. Patients don't want to be care managers.

Moving on to the possibilities, Andy focuses on the reams of data that exist in the field.  Imagine if they could be put to work:

Although doctors eschew case management, there are plenty of other professionals who can help them with it, and forming Accountable Care Organizations gives the treatment staff access to such help. Tons of potential savings lie in the data that clinicians could collect and aggregate. . . .  ACOs and other large institutions can hire people who love to crunch big data . . . to create systems that slide seamlessly into clinical decision support and provide guidelines for better treatment, as well as handle the clinic's logistics better. So what we need to do is train a lot more experts in big data to understand the health care field and crunch its numbers.

This is a good point.  Unless health care organizations understand enough about their patients to create decision support mechanisms, there will be a tendency towards anecdotal biases in diagnosing and treating patients.  I would go further than Andy, though, and put his recommendation in the category of necessary but not sufficient.

As Brent James, Lucien Leape, and others have made clear, physicians need to be trained in applying the scientific method to the field of care delivery -- just as they now apply it to basic research.  Generally accepted protocols must be adopted, against which experiments in care improvement can be measured.  Doctors must be trained in this expertise, so that the data analysis capability Andy Oram persuasively explains to be within reach can be applied.

No comments: