In the Conversation with Dr. Jonathan Perlin in this issue, Mark Neuenschwander commented that most hospitals have bar coding — in the gift shop! Bedside bar coding for dispensing medication has been demonstrated to reduce hospital errors and save lives, but few hospitals have implemented it. Why is it so difficult to bring new technology to health care? One of the reasons is economic. To argue this case, I will first look at two other industries where technology brings great advantage, and look at the underlying economic model governing this deployment.
For many years, I was involved in the development of technology for The Boeing Company. We weren’t developing new computers or mobile devices. Rather, using tools that were available from the market, we built applications that could create an advantage for the company.
Here are four examples.
- Using early personal digital assistants, we created a mobile tool for the airline mechanic, providing access to maintenance information and diagnostic tools. By tracking what the mechanic did, we were also able to automate the documentation of his work.
- Much earlier, we developed mathematical algorithms to allow Boeing to simulate very large scale circuit design, long before any commercial tools were available.
- We built software to diagnose problems in the supply chain in the assembly of aircraft, enabling the company to overcome the bottlenecks, which had begun to arise when trying to greatly increase the production of aircraft.
- We built the first software to enable the visualization of an airplane design, freeing Boeing from building a physical mockup in order to test that each part in the design would fit together without gaps or interference.
To my knowledge, though these tools were created between the late 1970s and the late 1990s, all are in use today in the company in some form or another.
The first example became the portable maintenance aid, a product Boeing offers to the airlines. The second enabled the company to win competitive research and development projects because of the technical edge it created. The third created documented savings for the company measured in very large numbers as Boeing avoided line shutdowns and costly delays in aircraft delivery. And the fourth was at the heart of the design of the first all-digital airplane, the 777, creating measurable design time savings, and performance improvement on the airplane.
As different as these tools were, there was a common theme. Every project had to earn its way. The measure of success was not just the technical accomplishment and the acceptance from the technical community (measured by patents or publications), but the payoff for the company. In other words, Boeing could measure the return on investment in technology, and all investments were channeled through this lens. There was no “technology for technology’s sake,” and everyone in the technology organization was fully aware of the yardstick by which we were ultimately measured.
By contrast, in the late 1970s, our organization broadened its horizons beyond the Boeing Company for a short period. We looked at applying our research talents to other industries with two objectives: we could get paid for the research and we could develop technology of value to Boeing. This took us first to the energy sector, starting with the electric utility industry. The generation, distribution, and cost-effective delivery of electricity provided a great research base and at the lowest level had many problems in common with Boeing. After the major East Coast blackout of 1977, there was research money available to address these very large and challenging problems. For a few years we were deeply involved in this industry.
But then something changed. Since the utilities were still regulated, and since the memory of the blackout began to fade, there was little economic incentive to address these problems, which remain to this day. The rate structure and the rate commissions took away the incentive for doing the research to make the generation and distribution of power more effective, efficient, and reliable. They also took away incentive to implement available new technology. The money for alternatives such as solar power also soon dried up. The problems were still there, but the economic incentives to solve them were not.
I suspect with deregulation and the concern for the environment, the picture is somewhat different today than in the late 1980s, though likely not much. But the example is clear: Market forces can provide a powerful incentive for addressing complex problems through technology. And when they are not present, it becomes much more difficult to attract the research dollars or the implementation dollars for new technology.
This brings me to the use of technology in health care. I believe the challenge of applying technology to address health care issues is more like the electric utility industry than the design and production of airplanes. The health care system has incredibly challenging problems where technology can play a significant role.
But the linkage to market forces is weak and indirect. For the people who have access to health care through insurance, those making the choices about care (the doctors and patients) are not the ones who pay for care. Making it worse, doctors sometimes benefit from more care in a “fee for service” environment, and patients feel better about it as well, even when it brings no advantage toward ultimate outcomes. This, too, is like the electric utility problem: More efficient usage lowers the income for the electric utility.
Luke McGuinness, CEO of the Dupage County Health Care System in suburban Chicago said,
“It’s unrealistic to think that the person who consumes health care is going to be able to deal with price at the same time. I don’t think we can depend on them to care about the cost.”
Insurance companies try to rule from the side, denying coverage or limiting access. Hospitals try to control costs. But this is far more indirect and less systematic than traditional market forces.
Why would bedside bar coding be slower to implement than bar coding in the gift shop? The gift shop has a bottom line, measures its efficiency, and found the investment in bar coding would pay off over a reasonable time. Hospitals must look at bar coding differently. In spite of the payoff in patient safety, and the reduction of time required for the hospital personnel, how does the hospital create the business case for the investment? Any cost savings simply result in lower payments from the insurance companies. Patients usually don’t know enough about the technology to demand its use.
Check lists in surgery provide another example. Gawande and Pronovost make the case that checklists lead to dramatic reduction of post operative infection. They are also low cost to implement, though the cultural change they require (typical with the introduction of new technology) is significant. Yet where is the incentive to make checklists standard procedure across all hospitals? It would take patients or insurance companies, armed with the data, to demand these changes in the hospitals, and this does not seem to be happening.
Electronic health records represent a third example of process technology for health care. When the hospital records of every patient are available electronically, they can then be referenced by any other health care representative with the ability to access. Today this is limited to other providers within the same system, and even here the penetration of this technology is less than 50 percent in doctors’ offices, and far less in hospitals.
One way to address this issue comes, surprisingly, from the government. The new health care legislation has identified some areas of technology that have been proven through clinical trials, and have incredible payoff in efficiency or value to the patient. The legislation has dictated that hospitals use electronic health care records by 2015, so finally there is an economic incentive to do so. Not for the bottom line of the hospital, but out of concern for the fines that the legislation would bring.
In our Conversation with Jonathan Perlin we learned what happened at the Veterans Hospital system:
Over that 1996-2004 period of time, the general cost of health care, unadjusted for nominal dollars, increased by 44.7 percent. The cost of care for veterans over that same period of time went up 0.9 percent. During that period in the VA we had measurably better quality, access to care (where we increased with hundreds of clinics during that time), benchmarked improvements in customer satisfaction, and lowered death rates for preventable illness. Also during that time we had full deployment of electronic health records providing complete access to records across the system.
Perhaps in an industry like health care or electric power utilities, some outside force for change is necessary to make progress in technology implementation that reduces costs and improves outcomes.
I have watched with interest the political bickering that is going on in this country over health care. Many conservatives seem to like health care as it is, perhaps because they have insurance and thus can get what they want. Yet this is not a market driven system. It lacks the economic connection between cost and outcome. How can we better deal with the efficient, reliable, safe, and cost effective provision of health care and the technology that supports that objective? The government role in health care seems to offer some hope for progress in this area. What are the alternatives?
Comments to firstname.lastname@example.org
Al Erisman is executive editor of Ethix, which he co-founded in 1998.
He spent 32 years at The Boeing Company, the last 11 as director of technology.
He was selected as a senior technical fellow of The Boeing Company in 1990,
and received his Ph.D. in applied mathematics from Iowa State University.