In the US healthcare industry, payors and providers¹ are spending on IT like never before, convinced that it will be a key strategic factor. Hospitals alone spent more than $7 billion on IT in 1994—nearly 2.5 percent of their net revenues. Payors and providers as a whole spent nearly $17 billion. These sums are expected to increase to at least $44 billion and perhaps even as much as $75 billion by 2004.

Is this heavy spending bringing anything in return? So far, payors and providers have mostly succeeded only in adding to costs. Productivity is showing few signs of improvement. The chief beneficiaries of the spending boom have instead been IT suppliers. They have enjoyed average compound annual growth of 17 percent in revenues and 88 percent in profits since 1991 (Exhibit 1).

Yet payors and providers are right to see an important source of competitive distinction in the information they possess. As healthcare markets evolve, information will visibly reduce costs and raise clinical quality for purchasers. IT will be essential in managing the extra risks associated with capitated revenues,² as well as in tackling the new managerial challenges presented by the industry’s ongoing vertical and horizontal integration. Advantage will accrue to payors and providers that apply IT to address these market forces. In fact, we believe that a well-thought-out and well-executed IT strategy will differentiate winners from losers.

To succeed, participants will have to sidestep the pitfalls that have so far prevented their IT investments from paying off. The first of these pitfalls is failing to focus on productivity in purchasing and implementing technology. The second is allowing other players within and outside the industry to capture the value created. The third is that ancient managerial plague, poor execution. And the fourth is the temptation to buy gadgets (such as fancy point-and-click physician systems) that neither accomplish cost objectives nor improve quality—mere toys, instead of tools.

Many industries go through similar periods where IT investments run up costs while delivering neither improved productivity nor the competitive advantage that higher productivity is likely to yield. It took years for the commercial banking industry, for example, to wring acceptable profits from its heavy IT spending in the 1980s, as it struggled to match investments to strategy, improve productivity, and use technology to change consumer behavior. Payors and providers can learn from such experiences by following a few simple guidelines in order to capture value from IT:

• Focus, above all else, on improving the productivity of front-line care givers
• Make your IT investments in the right sequence
• Match your IT investments to your market position
• Develop superior IT executional skills.

Focus on productivity

Some might argue that productivity discussions have no place in an industry where the quality of outcomes is so important. Certainly, it would be hard to find anyone who would advocate reducing the quality of outcomes in the pursuit of competitive advantage. Almost everyone, on the other hand, would be in favor of achieving the same outcomes at a lower cost, or better outcomes at the same cost.

For payors and providers, the issue is clear: if you do not increase your levels of productivity—defined as the ratio of outputs to inputs—you may well find yourself on the short end of the competitive stick. Analysis of health maintenance organization best practices reveals a cost reduction opportunity of up to $270 billion in the US healthcare industry (Exhibit 2).

Given current market pressures, the biggest opportunity—at least for the moment—will lie in reducing the input side of the equation. The US healthcare industry has yet to curb the tremendous inefficiencies arising from unmanaged care by individual physicians. Although treating any one patient will involve a unique combination of complex decisions, aggregating patient populations and examining variations in physician decision making will yield valuable insights for practitioners. At present, most physicians practise in relative isolation from their peers; considerable practice variability exists between providers (which frequently follows a consistent pattern); and only very limited information about care decisions is fed back to physicians.

Inefficiencies emerge clearly when we examine provider practice variations. Even after adjustments are made for patient populations, many provider organizations have sizable differences among their physicians in terms of the quantity of inputs needed to achieve a given outcome. Some physicians spend a lot more money than others in order to remedy the same illness.

How often, for example, do physicians see a diabetic patient? Once every two months? Every four months? How often do they screen for glaucoma? What lab tests do they order? Which medications do they prescribe? If physician A consistently achieves the same outcome as physician B for two-thirds the cost, what can physician B learn from physician A?

There is a real opportunity for physicians to use information to support the adoption of standards, or protocols, along the lines of what in an industrial setting would be called best practice. It is not that the healthcare industry actually lacks standards or protocols; rather, it is getting physicians to seek them out and actively use them that is the challenge. Information is a crucial enabler in this process, allowing physicians to engage in constructive discussions about practice pattern variations. If you have similar patient populations, adjusted for demographics and illness severity, you can look at the total costs of treating individuals and consider what drives the differences between them. In one case it may be pharmacy costs; in another, inpatient utilization.

Information holds the key to improving productivity. Providers will have to change their approach to managing information—and information technology. Improving productivity will require providers to excel at three other things.

Get the sequence right

Sequencing IT investments correctly can go a long way toward helping them pay off. (By IT investments, we mean investments in IT applications—whether developed in-house or purchased off the shelf—and in the hardware and executional skills that support them.) Getting the order right involves three steps that apply to all payors and providers, no matter where they compete in the healthcare delivery chain (see Exhibit 3):

1. Integrate basic information (such as care data) across the business system (inpatient, outpatient, laboratory, pharmacy, radiology) to create a complete picture of what was done to the patient, who did it, and what it cost.

2. Begin capturing clinical data regarding outcomes such as functionality or recovery time. Most payors and providers gather only financial claims and accounting data, which, though useful, hardly help in the understanding of practice variations.

3. Start gathering data longitudinally. Accumulating data over several years will prompt the kinds of insights into disease management that lead to more fundamental care delivery redesign.

1. Integrate information

There are several reasons why integrating information across the business system should be the first step in this sequence. For one, it is the easiest to implement. Much of it can be done by automating current procedures; no process redesign is required. Claims data, for instance, is already captured on relatively standard forms, making collection and comparison straightforward. And the analytical tools that will allow payors and providers to get good value out of basic data already exist.

The second reason for tackling this step first is that a good deal can be accomplished by integrating basic care data. The new insights that can be gained are significant and reasonably easy to act upon. An organization might use the information to decide, for example, which diseases and patients to focus on. Our experience suggests that the 80/20 rule as applied to costs holds here—that is, 80 percent of costs are driven by 20 percent of patients.

Third, time is of the essence. Industrywide efforts to reduce costs in such areas as inpatient days and outpatient procedures and visits, not to mention current trends toward increased capitation and vertical integration or alliances, are fueling new data needs.

Finally, this step also goes first because it is the necessary precursor to capturing value from the steps that follow.

Players obviously vary in the ease with which they can obtain and process information. Integrated health plans should be in the best position to capture the value from this step quickly, while individual hospitals and physician groups need to be more creative to overcome the organizational barriers that often exist to collecting data.

2. Capture clinical data

Most existing IT systems are designed to answer three questions: what procedure did we perform, what did it cost, and did we get paid for it?

Here the challenges begin getting stiffer. Most existing IT systems have been built around the patient’s bill, and are designed to answer three questions: what procedure did we perform, what did it cost, and did we get paid for it? Shifting from this sort of billing data to information that captures diagnosis, severity of condition, functional status after treatment, and treatment paths will provide great value, and could become a real differentiating factor for payors and providers.

But accomplishing this shift will not be easy. First, basic care data must be integrated across the business system. Next, selective process redesigns may be needed to ensure that the captured data is "clean." And since most clinical data currently exists only on paper, significant new software investments may also be required.

Why do it at all? Take a patient who needs a hip replacement. Clinical data from inpatient, outpatient, and long-term care is necessary when making decisions about appropriate (and cost-effective) treatment. Equipped with information from only one side of the care continuum (say, inpatient), the provider will have difficulty answering questions about the delivery process and its outcomes, such as: Is the patient better? How long did the treatment take? How important were individual elements of delivery in achieving a positive outcome? How can care delivery be reconfigured to employ lower-cost resources?

Only clinical data can demonstrate the efficacy of the care provided—or, indeed, show whether it was necessary at all

Another reason for assembling clinical data is that only limited insights can be drawn from claims data in the absence of corresponding clinical or outcomes data. Only clinical data can demonstrate the efficacy of the care provided—or, indeed, show whether it was necessary at all. Moreover, physicians are skeptical (and rightly so) about information that fails to incorporate outcomes.

One problem in collecting clinical data is that there are as yet no widely accepted standards within the medical community for measuring outcomes. New analytical tools are needed. How, for instance, do you define functional status? The challenge will be to ensure that IT investments comply with evolving industry standards.

Note that capturing clinical data is likely to be a "ticket to play" for providers as healthcare markets mature, but it will not be enough to secure competitive advantage. Since nearly all providers will be seeking to use clinical data as a way to create efficiencies, doing so will merely ensure that you keep pace.

Too many providers have yet to reconcile themselves to the fact that they will have to compete as low-cost suppliers

Keeping pace will also mean using the data you capture to reduce costs. Too many providers, hospitals in particular, have yet to reconcile themselves to the fact that all but a few of them will have to compete as low-cost suppliers. The exceptions—those that, like Memorial Sloan Kettering and the Mayo Clinic, can compete on the basis of a national quality brand—will be rare.

3. Pursue longitudinal data

This step is last in the sequence for three reasons: capturing the data will take several years; this step builds on the efforts that have gone before it; and at present only a few competitors have patient bases large enough to support longitudinal analysis. Collecting longitudinal data should begin only when the previous two data collection efforts are well under way. Without them, this third step would have limited value. How, for instance, do you use longitudinal data to determine whether a particular patient actually got better after a hip replacement operation ten years ago if you do not have inpatient and rehabilitation information?

Keeping track of data over a period of years makes it possible to monitor the real impact of clinical decisions, whether preventive or not. This, in turn, will promote the acceptance of practice guidelines. Should breast cancer screens be performed every two or every five years for women over 45? Such questions can be answered much more readily if solid longitudinal data is available to underpin the debate.

The challenge of collecting the data should not be underestimated. The storage requirements alone are immense

Gathering longitudinal data is in many ways the most important of these three steps, and may lead to fundamental changes in care delivery: say, fewer physicians or more nurse practitioners for a specific disease. But the challenge of collecting the data should not be underestimated. The storage requirements alone are immense. The necessary analytic tools are only just emerging. And the minimum efficient scale calls for a large patient base—probably bigger than a single hospital, for instance, will be able to muster.

Match strategy to market position

Payors and providers should also match their IT strategies to the stage of market evolution at which they find themselves. Healthcare markets in the United States follow a consistent evolutionary path, from a fee-for-service stage through timid, turbulent, and potentially restructured stages.³ Since the basis of competition shifts as the market evolves, IT strategies should change accordingly.

Payors and providers finding themselves in a fee-for-service or timid stage of market evolution should confine themselves to investments aimed at integrating data across the business system. For them, claims administration, risk management, and the strength of physician relationships act as the basis of competition. IT investments are best geared to back-office systems, typically those designed to capture claims data.

More turbulent markets call for investments in clinical data systems. These become necessary as aggressive competition requires rapid reductions in costs and prices in order to maintain share position. IT investments during this stage of market evolution should be focused on utilization review, quality assurance, and other systems for reconfiguring care processes and reducing practice pattern variations among physicians.

Getting practice guidelines to the point of care will boost productivity, increase quality, and reduce costs

In potentially restructured markets, the basis of competition may shift to a focus on clinical outcomes. Here, IT investments should be devoted to helping to implement clinical protocols. IT systems must strive to get real-time information into the hands of front-line providers. Getting practice guidelines to the point of care will boost productivity, increase quality, and reduce costs. This reflects the increasingly "partnered" role that providers must adopt with care givers as markets mature and the old "stick" approach to managing relationships becomes obsolete.

Base your strategy on your starting point

Obvious though it may seem, payors and providers should examine where they are in the healthcare delivery system before jumping in and investing in IT. Your starting point determines which strategies will work best for you.

There is value to be found in IT investments at every point in the delivery system. In general, though, hospitals are in a weak position in relation to other payors and providers because they have little influence over their delivery costs, most of which are controlled by physicians. They must find ways to strengthen their position by collecting more clinical data across the business system and by helping physicians to change delivery decisions. Such an approach can be difficult for hospitals because it means reducing the utilization of their biggest asset—beds.

Hospitals should focus on reducing costs, creating focused clinical value (probably a viable strategy only for the national quality brand leaders, such as Memorial Sloan Kettering), and integrating information with physician practices so as to get up to scale. They will have to structure their contracts with these practices carefully, however, to avoid letting physicians capture all of the value created.

Physician groups are probably in the best position to benefit from information systems in the future. Ultimately, physicians are the care providers and can best put new insights into practice. But physician groups are mostly too fragmented to gain access to capital and make the necessary investments. All the same, some large groups—Mullikin and Pacific Physicians among them—have been spending heavily on IT, and the early signs are that they are reaping substantial efficiencies from their investments. Such groups are growing rapidly, but still represent only a small percentage of physicians.

If health plans fail to secure a role in actual care management, they run the risk of being disintermediated altogether

For their part, health plans enjoy the best current position for gaining strategic advantage from IT investments, but they must find ways of using information to redefine their value proposition as local markets mature. If they fail to secure a role in actual care management, they run the risk of becoming marketing and administrative appendages to provider organizations—or, worse still, of being disintermediated altogether. In addition, health plans will have to change their approach to provider management and form more effective partnerships with physicians if they are to maintain the same kind of growth in revenues and profits that they have seen in the past eight to ten years.⁴

Improve execution

This last guideline is probably the most important—and most difficult—of all. Healthcare IT skills are among the lowest in US industries, mostly because IT has not been strategically critical to success until now, and payors and providers have not had to focus on the kind of executional excellence that is needed to make IT investments pay off.

Given the shortage of in-house talent, payors and providers would do well to buy software off the shelf and outsource much of their IT capability

Given the shortage of in-house talent, payors and providers would do well to look for opportunities to buy software off the shelf and outsource much of their IT capability. The experience of other industries indicates that major in-house development projects—in which teams of software writers develop code for a customized application—rarely deliver value, take a very long time, and unwittingly impose constraints on operational flexibility. Such projects often overreach as managers try to find the IT Holy Grail, a solution to all of their problems. Winning organizations will take more of a "Do it, try it, fix it" approach to implementation, making sure they exhaust all opportunities to build user-friendly front-ends for current systems and seek out package solutions before they begin developing their own software in-house and buying new hardware.

Executional excellence in IT also means integrating systems with performance measures. Organizations must have a good idea of where value is to be created before they begin spending. The best approach is to focus IT investments where there is greatest leverage: for instance, on the 20 percent of members who consume 80 percent of medical costs, or on the seven diseases that account for 60 percent of expenditure. IT systems must also be linked to both monetary and nonmonetary measures of physician performance.

Payors and providers need to understand how physicians interact with technology. Several years ago, a group of hospitals spent large sums on linking their physicians to their network by personal computer so as to share laboratory and other clinical information. The results have been disappointing: the PCs are mainly used by hospital office managers for billing and inventory purposes. However, other hospitals that established simple automated fax procedures for forwarding clinical data to physicians have been much more successful in improving physician productivity—and have spent a lot less.

The best way to prevent physicians rejecting a technology "solution" is to have them actively engaged in the design of the applications

Experience suggests that the best way to prevent key user groups (like physicians) from rejecting a technology "solution" is to have them actively engaged in the design of the applications.

Payors and providers that follow the guidelines above may have an easier time than their competitors in avoiding IT applications that add to medical costs without changing the way in which physicians practise medicine. Investing in these IT "toys" will line the pockets of IT suppliers, but it will not reduce costs. Nor will it produce the improvements in clinical quality that create competitive advantage.