The Difficult Path to Electronic Health Records

They will produce better outcomes. But building this 'network of networks' will be complicated.
September 29, 2009 AT 3:00 AM
By Stephen Goldsmith  |  Contributor
Professor of practice at the Harvard Kennedy School and director of the Innovations in American Government Program

Last week the National Governors Association convened its State Alliance for e-Health to discuss state approaches to developing electronic health records systems. By earmarking billions of dollars for health information technology spending in the American Recovery and Reinvestment Act, this administration has made digitization of all Americans' health care records a top priority. The effort clearly affects a complex array of public, nonprofit and for-profit health care providers, but it also merits careful study by anyone dealing with complicated public-private delivery systems.

Discussions about EHR systems involve networks of people in various agencies, companies and nonprofits. When we hear "systems" and "networks," we need to think about how individuals can accomplish their public goals through better access to other public servants and practitioners and their information. As we discussed in Governing by Network , technology is the key to binding the network together -- the technology network supports the human network.

Whether one is discussing health care, education, housing, or any of the other myriad and complex public challenges, the path to transformative breakthroughs starts with electronic records. Unfortunately even the process of creating electronic records presents enormously difficult issues. In my work as a prosecutor and mayor, I found that every entity can see how easier access to data from other organizations will facilitate better decision-making -- but inevitably, each entity questions whether the other organizations can be trusted with its information.

To move toward an integrated network of digital information requires that participants agree on terms, governance, protocols and a process to mediate conflicts. Designing these data networks is complicated because you're actually creating a network of networks, all of which have their own issues. A hospital will manage a network of providers and interested parties -- including, for example, its clinics and its doctors. Yet this network will also be part of the statewide electronic network. To the governor's health secretary, he or she is at the network's center, but to the hospital president, or medical group administrator, or community health center, they are the center, and their efforts and interests take precedence. Harmonizing these interests requires careful mapping and understanding of the consequences.

That is the easy stuff. The electronic record is a step towards better outcomes, not an end in itself. Who, for instance, gets to define the outcomes, especially when an outcome that is central to one party is not so central to others? Huge amounts of accessible data can now be managed to provide insights that will dramatically improve public interventions, but that management requires careful consideration about how it happens, who supports it and how the data gets to the field worker. Too often these systems serve the person at the top but not the person really doing the work. We have seen this in, for instance, large federally supported child welfare systems that view the job of the caseworker as less important than supporting oversight in Washington.

I invite our other management authors to discuss this issue, which is too complex for one column. Here I suggest some questions for them, and our readers, to consider:

o How do we set up a process -- carefully, from the beginning -- to define public and private value in a network?

o How does one design this network? Which players should be included, and what is the process of identifying assets?

o How can we set up a dynamic governance process that accommodates daily tensions as priorities shift and discoveries are made?

o How does government develop the new set of leadership skills needed in these situations -- mediating, negotiating and facilitating, not commanding?

o How does a network operate when government is really not a partner but the 800-pound gorilla? How do you make the gorilla use its strength in a constructive fashion?

State governments have already seen breakthroughs in the medical world as they use newly acquired information: to conduct long-term health care audits for purposes of reducing or delaying nursing home visits; for better handling of emergency room admissions; and to develop targeted prevention programs for the chronically ill. Pennsylvania led the way with an early digital network called the PA-NEDSS. Developed a few years ago by the Pennsylvania Department of Health, the goal of the network was to increase sources of information on outbreaks and reduce the reporting cycle time from three weeks to 24 hours. The better, faster information resulted in quicker interventions.

Electronic records systems will produce dramatically better outcomes in the future, but the path to those better outcomes requires careful analysis and approach. And public officials not involved with health issues would nonetheless do well to observe the ongoing EHR efforts with care.