Editor’s Note: This post concludes a series on health information technology (IT). It follows posts by Mark Leavitt and Nancy Davenport-Ennis. This blog series appears in tandem with new papers on the Health Affairs Web site [2-week free access], including a lead article on why we need to expand beyond narrowly focused standard setting to unlock the potential of health IT by the Markle Foundation’s Carol Diamond and New York University’s Clay Shirky; a Perspective by Robert Kolodner, the national coordinator for health IT, and coauthors; and a Perspective by David Kibbe of the American Academy of Family Physicians and Curtis McLaughlin of the University of North Carolina.

[Disclosure note: As indicated below (with asterisks and at the end of this post), I’m an investor in many of the companies I mention. Please take this not as sleazy promotion, but as evidence of my sincere conviction in what I write. Probably not all of these companies will be successful, but they are all evidence of healthy new growth amidst the lumbering forests of today’s health care system. And you may find this discussion overly commercial, but it is branded commercial services — that consumers actually want to use and possibly even pay for — that will drive progress.]

Our goal in health care should not be the adoption of information technology (IT); it should be health. But I think the use of IT, especially in the hands of individuals (not just patients), can be key to that broader outcome.

But government standards efforts (or magical thinking) won’t make it happen. Rather, I think it will pretty much happen by itself — or rather by the decentralized efforts of millions of people and the slightly more centralized or at least clustered efforts of hundreds of companies, mostly start-ups but eventually some larger ones, too. Ultimately, the people will demand a new incentive system and the government will respond, by putting in place an insurance regime that will reward health care providers for the difference between predicted and actual health incomes for a covered population, rather than for the costs they expend in providing care. (And people with poor health prospects will have their insurance rates subsidized.)

Of course, this won’t all come about smoothly and without glitches, but it will be pretty good compared to the alternative, which is continuation of the current reality.

Here’s a few of the things that will happen. I leave it as an exercise to the reader to knit them together into a coherent story.

Consumers managing their own information

First of all, many users will start to generate and manage their own medical information, just as they create and manage information about the books they buy, the money they spend, and the trips they take. Some of this information will be by-products of online services: You can track your spending at your bank or credit-card company; you can track your travels at TripIt* or Dopplr*; you can manage your virtual bookshelf and compare it to those of friends on Facebook. And now, at PatientsLikeMe,* you can track your symptoms, treatments, and everything else if you have ALS or Parkinson’s disease. A friend of mine has a pedometer; healthy people sometimes record their weight; those with diabetes learn to monitor and record their blood sugar.

And vendors are responding. Years ago, Intuit’s Quicken let people manage a checkbook online and print out checks. As the world moved online, Intuit built interfaces to the online banking services of certain major banks so that users could upload and download their financial information. Pretty soon after that, the less-popular banks built their own interfaces to Quicken in order to capture or regain market share. None of them really wanted to let users mingle data from one source with data from another . . . but of course it happened. And now, users can use Mint or Wesabe* not just to mingle data but to analyze it, to compare their own spending patterns with others’ and (here’s the business model) to receive “offers” from other financial institutions based on those spending patterns.

I expect exactly the same thing to happen with health information. And, just as banks compete (mostly with success) to keep customers’ money and data safe, so will the organizations that handle users’ medical information. This doesn’t actually require any legislation or standards to happen. Companies are already emerging that will provide interfaces to the IT systems of market-leading health care and insurance providers; eventually, the laggards will adapt to the market standards set by the leaders.

In varying stages of development, I know of companies that help users monitor drug interactions (DoubleCheck MD), persistent pain (ReliefInsite*), medical bills (Simo Health, acquired by Revolution Health), and their overall health data (Microsoft Health Vault and Google Health). 23andMe*, Navigenics, and DecodeMe let consumers explore their own genomes; 23andMe lets them compare themselves with others (by mutual consent).

The point is that this is already starting to happen. The first few vendors to do it right will set the standards, and the others will follow.

Personal health records

You may wonder where the huge numbers of “personal health record” (PHR) companies are in all this. Many of them are still around; some are gaining traction, especially those with a large provider, insurer or employer promoting their adoption. And Microsoft and Google are joining the fray. However, most consumers (other than expectant or new mothers) aren’t that interested in a health record per se. They are interested in dealing with medical bills, or a particular condition, or a particular threat. So much adoption, I believe, will be driven by the kinds of specific applications I mention above. Over time, the PHRs will be able to accept the specific data types generated by the applications, and new applications will generate data that they know the PHRs can handle. (Meanwhile, as Carol Diamond and Clay Shirky point out, it’s more important that the data can move than that it can be parsed on arrival.)

User demand is going to drive this, and eventually the institutions will respond.

The financial side

Ultimately, we’ll end up with a wealth of data. Then the issues around data aggregation and data mining will come to the fore. If we have done things right, we’ll have systems in place to go back and ask consumers for permission to use their data in studies — of everything from their diets and behavior to their genomes, their medical records, and their pharmaceutical purchases. A lot of this is already available, but it is just starting to be aggregated. One of the leaders is Active Health, a start-up acquired by Aetna, which monitors prescription transactions and alerts doctors to possible conflicts.

But so much more is possible. The first phase is simply collecting data about an individual for that individual’s benefit. The next phase is aggregating lots of data, in order to assess different approaches to prevention and care, among other things. That should improve health care outcomes, but it should also improve health. This data will also be increasingly visible to individuals: They’ll be able to see not just their own data, but also their own, personally predicted health: “Eat like this, and you’ll look like that. Exercise more, and you could look like that, even when you’re 70!” “Use this drug, and you’ll get the best relief.”

As any businessperson knows, measurement is the first step toward behavior change.

Where information meets money

And then the third phase is where it gets interesting financially and institutionally. One place the whole system is tender to the touch is where information meets money: People are afraid of letting insurers or others know too much about their health because they are afraid they won’t get insurance — or that their conditions won’t be covered even if they have insurance.

But the spread of information is inevitable, as are ineffective laws trying to protect people from insurers who are in turn trying to protect themselves (and their widow-and-orphan shareholders) from insurance fraud or just from the costs of caring for the sickest people.

This will all come to a head. Too much knowledge will ultimately make the insurance business untenable as it is currently structured: It is ignorance about predictable outcomes that enables most people to be insured. As that ignorance dissipates, our system will have to pay health care providers for improvements over the predicted outcomes — and will have to subsidize those whose predicted outcomes are worst. How we get there will be complex, and it will involve politics, just as the first phases I outlined will involve business, marketing, and even (heavens!) advertising.

But it’s the messy, amazing way most things happen from the bottom up.

[Disclosure: Direct investor in 23andMe, Dopplr, PatientsLikeMe, ReliefInsite, Indirect investor in TripIt and Wesabe. Board member of 23andMe]