Incentives Matter

Incentives Matter

Within the healthcare IT community, it's widely known that Health Information Exchanges (HIEs) have had a sustainability problem since their inception. Everyone wants to reap the benefits of exchanging information, but no one wants to pay for it. In economic terms, this is a classical "public goods" problem. A widely used example of a public good is roads. Other key examples include clean air/water, and parks. Everyone wants awesome highways, pristine air and beautiful parks. Someone has to pay for them.

Given that the HIE sustainability problem closely mirrors a problem that we've been solving for hundreds of years, why don't we just implement the same solution that we have been for so long? 


We could tax hospitals, clinics, labs, pharmacists, and everyone else that would access the HIE data to generate sustainability revenue, but that seems politically unlikely because all the parties involved have a vested interest in keeping their own taxes as low possible. With no tax revenue and no sustainable business model in sight, HIEs will continue to struggle.

However, HIEs have an interesting opportunity within large healthcare delivery systems. It's common knowledge these days that healthcare enterprise are growing, mostly through mergers and acquisitions. Hospitals are buying up clinics left and right, and many hospitals and health systems are merging. They're trying to reduce overhead through increased overhead synergies, and to negotiate better more favorable contracts with payers.

So we have a situation where large healthcare organizations are buying up lots of entities with disparate EHR, PM, and patient portal systems. Some hospitals are offering their newly owned clinics an opportunity to adopt Epic or Cerner for their practice at a discount. But many of the doctors are hesitating to jump into the new system after having just purchased their own EHR recently. The small doctor practices are rightly worried that adopting big brother's system will give the hospital new insight and ultimately control over their practice. They know that if the hospital can't see into their practice, the hospital can't control their practice.

The environment described above is perfect to foster the growth of private integrated HIEs. Lets pretend to be the CMIO of a major health system that now operates a dozen hospitals and 200 clinics with a total of 500 disparate IT systems, of which perhaps only a few dozen talk to each other. We have 3 fundamentally different options:

1) Coerce thousands of people onto our Epic/Cerner database, and open ourselves up to enormous HIPAA-risk (a significant number of the comments in that HIStalk post regard clinic access to the hospital's database) and vendor-risk.

2) Maintain the status quo, and work with an inadequately funded, small 3rd party HIE to enable interoperability within parts of our health system and parts of other nearby health systems. We know that this HIE will split its resources among our organization and others, creating more opportunities for delays and failure, and less value for us.

3) Develop and manage our own HIE organization (an entity, not software; that can be easily purchased), and mandate it among all of the participants within our health system. Given that we currently eat the cost of re-collecting patient information 20 times within our own health system, we have a strong incentive to make it easier for our providers to get the information they need from other providers instead of re-collecting it 20x over. This option also reduces the substantial vendor-risk associated with putting all of our eggs in Epic's or Cerner's basket.

Although HIEs haven't had a great track record so far, I'm optimistic about their future. I don't think they will manifest as many were hoping they would - as independent entities that connect separate healthcare delivery systems - because of the fundamental incentive problem. But within a large health system, the there are incentives to build, manage, and maintain a robust HIE.

Moral of the story: People will do what you pay them to do, not what you tell them to do. Incentives matter. Make them want to do it, and they will.