Managed Care
Disease
Management

A Conversation with Jeff Goldsmith: He Connects Health Care's Dots Along Information Superhighway

MANAGED CARE May 2001. © MediMedia USA
Q&A

A Conversation with Jeff Goldsmith: He Connects Health Care's Dots Along Information Superhighway

Deny the increasing importance of the Internet at your peril, says the noted health care futurist. Faster connections, smarter applications are needed.
MANAGED CARE May 2001. ©MediMedia USA

Deny the increasing importance of the Internet at your peril, says the noted health care futurist. Faster connections, smarter applications are needed.

For nearly two decades, health futurist, author and lecturer Jeff Goldsmith has specialized in forecasting trends in medical technology, payment, and health policy. Goldsmith is president of Health Futures in Charlottesville, Va., and is an associate professor of medical education in the School of Medicine at the University of Virginia.

From 1979 to 1990, Goldsmith taught health services management and policy at the University of Chicago's Graduate School of Business. From 1982 to 1994, he served as national adviser for health care for Ernst & Young, and provided strategic advice to health care systems, health plans, and supply and technology companies.

Goldsmith graduated from Reed College in 1970, majoring in psychology and classics, and earned a doctorate in sociology from the University of Chicago in 1973, studying complex organizations, sociology of the professions, and politics of developing nations. He is a member of the editorial board of Health Affairs and the boards of directors of Cerner, a health information company, and Essent, a hospital-management concern. Goldsmith, a native of Portland, Ore., spoke recently with Senior Contributing Editor Patrick Mullen.

MANAGED CARE: At this early point in its development, how is the Internet changing health care?

JEFF GOLDSMITH: The biggest impact of the Internet on health care has been to make the system's knowledge base far more transparent to consumers. Relatively cumbersome search engines have enabled consumers, particularly women, to gain easy access to an extraordinary depth and breadth of health care information.

MC: Why particularly women?

GOLDSMITH: Women are the general managers of their families' health. If you take out e-mail, which everybody uses, the number one reason women log on is to seek health information.

MC: How is the Internet changing the delivery system?

GOLDSMITH: It's changing delivery very slowly. There's a profound reluctance to take up network computing applications that stems in large part from the immaturity of the applications. There's also a high level of skepticism of the current vendors on the part of health care providers. There was a tremendous wave of hype about health care Internet sites and applications. Most provider organizations got about as far as putting up a web page publicizing their institutions. More significant Internet activity has been much more difficult to come by. It's also been slowed by security concerns and by physician concerns about adopting the technology. Depending on the data you look at, only between maybe 6 percent and 12 percent of physicians in the country communicate with patients via e-mail and less than 5 percent have electronic records in their offices. The vast majority of hospitals think an Internet strategy is synonymous with having a web site. They think of it as a marketing tool, and as a consequence, fall far short of the capabilities of network computing applications to change how they operate.

MC: Will change come first to back-office financial applications and then to the delivery side?

GOLDSMITH: It's difficult to do anything meaningful in the way of connectivity as long as a provider's core financial and clinical records are on paper. Until both sets of data have been digitized, it's going to be very difficult to leverage network-computing applications and change much of the delivery system.

MC: How do we get to where you feel the system could be?

GOLDSMITH: The source data has to become digital to be moved through the network effectively. That will require a tremendous amount of work. Many start with financial data because it's easier and requires less physician involvement. Clearly, there is money to be saved by transmitting, adjudicating, processing, and paying medical claims electronically. That was the huge target that Healtheon shot for and missed. The opportunity continues to be large. The Health Insurance Portability and Accountability Act of 1996 will ease the transition by standardizing transaction formats and coding schemes used in transmitting data electronically.

MC: Given the existing skepticism among providers, where will the pressure for change come from?

GOLDSMITH: There are tangible and immediate savings in reduced accounts-receivable days and reduced clerical costs with digitizing financial transactions, so that's an incentive. An alternative to turning everything over to a national clearinghouse is to create a regional clearinghouse. We've seen a lot of regional efforts to do this. One mature example is something called NEHEN.

MC: The New England Healthcare EDI Network.

GOLDSMITH: It's a great example. The economics are compelling. It's been a slow process that's been under way for five years. They started, I think appropriately, with eligibility, but the real payoff comes not just from eligibility and submission of claims, but from a complete interactive data system that also corrects, adjudicates, and pays claims. You have to get through all those steps to realize the bulk of the savings. The complexity of the national accounts picture and the issues related to the breadth of plan coverage nationally favor building out regional solutions rather than imposing a national architecture. This is an area where providers and major insurers like local Blue Cross plans could do a lot of good.

MC: Once payment is digitized, will the benefits of digitizing clinical information become evident?

GOLDSMITH: That's a tough question to answer. A couple of things are holding back the digitization of clinical information. One is the barrier between records maintained in physicians' offices and hospital or health system records maintained in some central location. Crossing that barrier is a huge political problem, given the bad state of hospital/physician relations. HIPAA is clearly fogging a lot of people's minds. I don't think people realize that they need to provide some degree of standardized certified privacy protection of patient records before patients will be willing to use the Internet to transmit and receive sensitive information. HIPAA is not a burden being imposed on them. It's a precursor to gaining public confidence in the use of the Internet.

MC: People are becoming more willing to use credit cards for online purchases, so attitudes seem to be changing.

GOLDSMITH: Public acceptance is slow. It took time and an industry consortium to gain people's acceptance of online financial transactions. American Express and other financial firms got together about four years ago and created a consortium on data security. The axiom about e-mail is that you should never put anything in an e-mail that you're not willing to put on a postcard. Nobody wants the results of sensitive medical tests essentially broadcast in this way.

MC: My understanding is that health systems that use e-mail don't send test results. They send the patient a message saying, "Your test results are ready. Come to your secure site to see them."

GOLDSMITH: That's the right way to do it, with some type of security gate to go through.

MC: How will the Internet change in a few years?

GOLDSMITH: Two big things that will improve the capacity and intelligence of the network will be important for health care. One is the emerging explosion in broadband. Broadband is important because health care data sets are thick and chunky, particularly digital radiographic images and patient records. You need bandwidth and storage capacity. The second thing, which will make a bigger difference than broadband, is the emerging computer markup language XML. XML tags information inside a web site based on what it means rather than how it looks on a web page. As XML becomes broadly accepted over the next three to five years, health care web sites are going to be converted from an enormous magazine collection into an enormous searchable database. It will enable a quantum increase in the intelligence and focus of the search process. That in turn will help weed out information that shouldn't turn up in searches.

MC: The era of the passive patient who does what the doctor tells him is coming to an end. How will consumer expectations change the way health plans, physicians, and hospitals respond to their customers?

GOLDSMITH: I want to challenge that assumption. My consumer-research friends are discovering that patients have distinctly different styles of using the health system. A subset of patients does not want to be bothered with the search process; they do rely on their doctors for medical knowledge and to tell them what to do. I don't think that's going away. The Internet has made things enormously easier for those people who wish to manage their own health. The central thrust of successful Internet strategies for health plans and providers will be facilitating that self-management. That will be particularly true for patients with chronic illnesses, the principal disease burden of our society. The primary caregiver in most chronic-illness situations is a family member, typically a woman — a wife, mother, or daughter. The Internet gives these front-line caregivers an extraordinarily powerful tool to acquire information about the condition they're managing and to help them make better decisions about the health plans and providers that they select. Once they select them, they enter the dialogue with those folks at a much higher level. A tremendous caregiving burden that does not enter into our national health accounts is the burden imposed on family members by intractable disease. Think about those who are taking care of Alzheimer's patients at home.

MC: And of course, those numbers are going to climb.

GOLDSMITH: They're going to skyrocket. We are entering an era of a demographically driven epidemic of degenerative diseases of the nervous system. In terms of the selfish interest of us baby boomers, the decade of the brain should have been the '80s. We're not going to remember why we were doing all that jogging and passing up the steak and eggs. There will be a business returning lost joggers to their retirement communities.

MC: So you're not optimistic that pharmaceutical companies will solve that problem by coming up with a memory pill?

GOLDSMITH: I recently sat through a compelling biotech presentation on that precise subject. There actually are some promising developments in the area of memory-enhancing therapeutics.

MC: Maybe we'll end up with the opposite problem, that we remember everything.

GOLDSMITH: Speaking of trying to remember everything, one thing that is clear from the second Institute of Medicine report is that it is absurd to expect physicians to carry around in their heads the entire active knowledge base of what needs to be done in a particular patient situation. Physicians presently lack the ability to archive and store information in a way that they are confident they can retrieve when they need it. That's another benefit of knowledge-management applications. XML, emerging expert systems, and artificial intelligence are so important in medicine because they will help people find new information and then stash it somewhere where they can get it later. Instead of saying, "I know the answer to this question," they can simply say, "I know where to find it. It's on my server." A firm here in Charlottesville, Unbound Medicine, built a prototype Palm-based knowledge-archiving tool that is one of the slickest things I've seen in my life. I'm actually going to buy a Palm because he promised to give me access to the software.

MC: That's the definition of a killer software application, one that's hot enough to get you to buy the hardware.

GOLDSMITH: It is. There are four things I'd want as a nonphysician. One is a differential-diagnosis tool that lets you enter the symptoms and it goes through a decision tree. The second piece, which isn't that unusual, is a database to look at a drug's effects, possible interactions, and allergy risks. But the two things that really turned me on are a literature-scanning application and a clinical-query tool. You can list the journals that you read and the topics that you want to track and the server will push new articles or books onto your Palm in abstract form. You go to the server and check off what you want to retain and in what depth. You can retain full text for many of the journals. The query tool really blew me away. This will be really useful for physicians who hospitalize a lot of patients. If you're a doctor at a teaching hospital and a question occurs to you, you can turn to the fellow or resident and say, "Go find out X." If you're not lucky enough to have a fellow or a resident there, what do you do?

MC: Some argue that doctors are technophobes. Others say that's ridiculous — that physicians just don't want to be slowed down.

GOLDSMITH: I am on the latter side of that. Doctors in the main are technology-friendly. There's a very high rate of adoption of Internet applications by doctors for personal e-mail and things like managing their personal finances. On the other hand, there is a high functionality hurdle that clinical applications have to surmount. It is not unreasonable for a physician to hold out for a prescribing function that is as efficient in the use of time as their prescription pad. The real opportunity here will be with voice-activated systems and natural language processing. Speech recognition is here now. The gap comes with the natural language processing that converts speech into digitally actionable data. We're still a couple of years away on that front.

MC: How well are health plans figuring out the changes they need to make to survive?

GOLDSMITH: You'd have to be stone deaf not to have picked up a lot of consumer dissatisfaction with health plan operations and with the product itself. In varying degrees, health plans have responded aggressively in trying to use Internet computing applications to change the relationship with their subscribers. The firms that have made the biggest push early are the two West Coast group model plans, Group Health Cooperative and Kaiser Permanente. Kaiser is reportedly investing a couple billion dollars in digital enterprise and network computing applications. Its goal is making the health plan essentially transparent to the subscriber, not merely enabling him to track financial transactions. Of course in a group-model HMO, the issue isn't billing, it's the status of orders, the ability to make appointments online, to be able to get medical advice without seeing a doctor. That could be helpful in screening out a lot of unnecessary clinic and ER visits. More broadly, plans are engaging subsets of the population in disease-management activities that can make a difference in their health status. For a lot of years the user interface, to use IT jargon, was the bottleneck in disease management. You either had to send a nurse to someone's home, which costs a hundred bucks minimum, or you had to take the patient to the clinic or hospital, and Lord knows what that costs. It's pretty clear that the new user interface for DM is a web page. Patients can receive health information and structured health advice and can feed back their health status. It's a marvelous tool, whether the delivery system is attached to the health plan as with staff-model HMOs or for insurance carriers that want to minimize the number of unnecessary physician visits. In either case, you win by making it easier for patients to acquire information. Someone at Humana told me that they get 20 million calls a year from their subscribers. It costs twenty dollars to answer each call.

MC: So clearly there are savings to be had, plus happier members.

GOLDSMITH: Exactly. Of course, the social norm with e-mail is that you're going to respond quickly. So the need to be responsive doesn't go away; it becomes somewhat more urgent.

MC: Which companies are doing the best job making the transition?

GOLDSMITH: United Healthcare and Aetna U.S. Healthcare have been working hard on discrete Internet applications. Certainly Aetna U.S. Healthcare got a lot of publicity for encouraging electronic submission and prompt payment of physician claims. The company that seems to have moved most aggressively toward digitizing its whole operation is Humana, at the personal intervention of cofounder and chairman David Jones, who basically had his executives start over.

MC: That's been his history. He's taken Humana from being a nursing-home company to a hospital company to a hospital/managed care company to a managed care company.

GOLDSMITH: He may have the appropriate level of urgency about creating a new business model; the near-term economic prospects of managed care plans are fairly grim. The nice accepting pricing environment is fading quickly as the economy cools. It's unreasonable to think employers will tolerate a couple more years of double-digit premium increases when inflation is stable at 3.5 percent. And managed care companies shouldn't expect any help from providers, given that the fear of managed care plans in the provider world has been replaced by rage.

MC: I wouldn't want to be the HMO executive who has to rely on physicians to save the company.

GOLDSMITH: The response to that would be a brick in the face. The real imponderable is whether, as in the past, there are a few health plans hungry enough and with enough cash flow to use this pricing environment to grab market share. It's clear that an opportunity exists to liquidate a fairly large chunk of the spread between premiums and medical expenses. There's an equally large opportunity to make more effective use of the medical expense portion. So there are two hits for plans that actually make it over the fence and implement digital systems correctly. They're going to have a different business model when they're done. No one is ever completely in control of how the re-engineering of his business turns out. The people who figure out how to make this digital transformation first will gain tremendous leverage in profitability and in customer service, which will be rapidly reflected in improved subscriber satisfaction levels.

MC: And the share of the health care dollar that actually goes to patient care will go up.

GOLDSMITH: More importantly, the money spent on patient care will buy better results, and that is ultimately a bigger hit. Back-office savings are one-time savings. The systems investments we're talking about for a decent size national health plan are in the nine-figure range. This is a huge cost, and it's a huge consulting opportunity, since only a handful of companies have a large enough in-house IT staff to do it themselves.

MC: Are clinical innovations about to dramatically change how the system operates?

GOLDSMITH: Most of the increase in the intensity of health care is going to come from new pharmaceutical agents. How to manage pharmacy expense is a huge issue. Three things came together to spark the explosion in prescription drug expense: the arrival of exciting but expensive new drugs like Enbrel and Vioxx, increased demand for drugs due to direct-to-consumer advertising, and expanded drug coverage. Not surprisingly, if you combine straw, kerosene, and a match, you're going to get fire. The logic of broadly sharing risk is applicable to the whole cost of health, not just pharmacy costs. People are going to have to re-examine how, where, and how much consumers pay out-of-pocket for health care use.

MC: Do you mean consumers should pay more?

GOLDSMITH: When they're making expensive choices, you bet. Here's an example. We know that there is anywhere from a three- to five-fold variation in mortality risk depending on which hospital a person goes to for some complex surgical procedures, like open heart surgery. Why shouldn't health plans publicize the mortality rates and relative costs of services provided by the broad panels they're required to carry, and give rebates to consumers who voluntarily choose the highest quality programs? With procedures like open heart or transplantation or maybe even cancer therapy, if we ever get to the point where Sloan-Kettering and some of the other national centers publish their results, we may discover that they're more cost-effective than local cancer centers. If that's the case and they have better results, the health plan wins by varying the cost-sharing depending on where someone goes.

MC: Thank you.