A comprehensive central repository of medical data could be an answer to the cost and quality problems plaguing US health care. Besides the medical histories of individual patients, the database would aggregate information spotlighting medical best practices and their impact on patient care as well as help researchers to identify health-related trends. It would be accessible to doctors, patients, insurers, and researchers via secure and confidential networks.

Without question, such a scheme would greatly improve the quality and consistency of health care practices. Improved data from health care providers would help patients to make better choices—a vital step toward self-directed care. In time, too, patients would benefit from access to standard, portable medical records containing all aspects of their medical history, from physicians' notes and prescriptions to X-rays and laboratory results.

For all of these undoubted advantages, the creation of a nationwide system of electronic health records (EHR) is far from a foregone conclusion. A sizable investment—as much as $200 billion initially—would be required to transform the industry's IT and telecommunications infrastructure to enable data capture and reporting by all participants. Furthermore, such a system would be extremely complicated, requiring the cooperation of a multitude of physicians, specialists, nurses, and technicians in hospitals, clinics, and laboratories (Exhibit 1).

Skeptics argue that a national network is a generation away. They point out, correctly, that a consensus on basic technology standards for data capture and interchange—let alone the thornier issues of personal privacy and data security—is a long way off. In the near term, they contend, the best hope is to experiment with EHR concepts and tools in narrower, controlled environments such as coordinated regional networks or large provider groups.

Champions of EHR see things differently. They point to the many catalysts that could move EHR forward: the continuing rise in health care costs; the increasing interest in EHR by some large private-sector bodies, including big employers (which pay for the health coverage of employees), large hospitals, and the makers of drugs and medical devices; and advances in information technology. In addition, they note that the federal government is pushing in this direction; in 2004, President George W. Bush formed the Office of the National Coordinator for Health Information Technology and appointed David Brailer to lead an effort to bring the US health care system into the digital age.

However, skeptics and champions alike are missing a more important issue that must be resolved if EHR is to succeed. While the costs involved are daunting (Exhibit 2), and the technical standards and policy challenges far from trivial, the key obstacle to EHR is the misalignment of costs and benefits among different groups in the health care system.

Misaligned costs and benefits

Certain industry participants would gain more from EHR than others would. Funders—the consumers and institutions that ultimately pay for health care, including state governments and the federal government (through Medicare and Medicaid) as well as businesses—stand to reap much of the savings, estimated to be worth from $75 billion to $100 billion a year.

Broadly, these benefits would come from higher quality and the reduced cost of care (Exhibit 3). EHR would improve efficiency by eliminating redundant treatments and directing patients to better providers. Employers large and small would be able to cut the cost of employee benefits, while the federal government and the states could reduce Medicare and Medicaid fraud and abuse through the better tracking of patients, providers, and procedures. The cost-benefit equation is especially attractive for the funders of health care because they would bear only a small proportion of the up-front costs (Exhibit 4).

Pharmaceutical companies and medical-device manufacturers too are well placed to benefit. Despite the need for anonymity, large pools of data about patients (and their conditions and diagnoses) and about procedures and physicians would provide a rich trove of research and clinical information to bolster R&D programs, help companies recruit patients and physicians for clinical trials, and inspire new ways of marketing to doctors and consumers. The same technology that physicians would use to gain access to and update electronic health records at the point of care could also be used by investigators conducting clinical trials, who would be able to input their findings more quickly. Timely, higher-quality data on trials would greatly reduce their cost and speed the progress of new drugs and devices to market. As with the funders, up-front costs for this group are likely to be relatively low.

The balance of benefits to costs is less clear, though, for health insurers such as Aetna and UnitedHealthcare, which underwrite the health care costs of employer groups and manage and resolve claims for providers. Competition would force these companies to pass on the bulk of the savings achieved through a national EHR system to customers in the form of lower premiums. Of course, with improved data access and exchange, insurers could process claims more efficiently—but this improvement would provide only a modest financial benefit, which also might have to be passed on to the insured. On the cost side, health insurers would both provide and receive EHR information, so they would have to make substantial investments in IT infrastructure and processes.

But the largest portion of the initial costs of a national EHR system will fall on health care providers. Larger ones, including organizations that offer integrated health plans (such as Kaiser Permanente and major hospital chains such as Sutter Health, Tenet Healthcare, and HCA), face large, multiyear investments to prepare for EHR. Some are already planning their infrastructure and adopting new processes, and for a simple reason: despite the high cost, more medical information is likely to improve the quality of care and even reduce costs. For these providers, EHR is an expensive and complex proposition that is nonetheless worthwhile.

It is another matter for small groups of physicians and for subscale, unprofitable hospitals and clinics. Such players would need to make huge investments in PC- and Internet-based technologies to capture and distribute data about patients and medical episodes: the symptoms patients reported, the diagnoses doctors made, the tests run, and the prescriptions written. Today that information is held in paper files and in small, distributed systems in the locations of individual providers. Even a basic EHR system would require those records—and any new data—to be collected in standardized digital formats. Hospitals, clinics, and physicians' offices would therefore need to retool their support systems and work processes. Physicians' notes, for example, would have to be typed and input into the health record; tests would have to be loaded into some standardized model. All of these efforts would require substantial investments in technology, including core systems, clinical systems, desktop technologies, data storage, and telecommunications.

In addition, providers face the considerable task of training their office staffs on these new systems. Initially, bringing IT to the doctor's office could slow things down—entering data into the new system would be more time consuming than jotting a few notes on a patient's chart. Moreover, patients might view the new technology as an intrusion that detracted from the quality of care and attention they received. In the long run, of course, the data gathered through EHR could be used to reveal which physicians were providing the most effective care and which were substandard. If care outcomes went online, a physician's track record would be laid bare. Armed with this information, insurers and funders would probably start "paying for performance," putting more pressure on all physicians and reducing the income of the weaker ones.

Breaking through the blockages

There are certainly potential drawbacks for physicians, but the information they provide would be the cornerstone of a national EHR system, which can't be achieved without their comprehensive cooperation. Some way must therefore be found to encourage the adoption and use of IT by these frontline practitioners. We believe that if smaller providers and physicians are to come on board and thus maintain EHR's current momentum, other parties involved with the industry, including the government, must find ways to help spread the costs and benefits more evenly.

Over time, the willingness of large providers such as Kaiser Permanente and the Veterans Health Administration to invest in EHR and to use electronic records in real care settings will help smaller groups of physicians to follow along. Leading hospital groups and academic medical centers have been investing significantly in IT infrastructure and solutions for more than a decade, laying much of the foundation necessary for a more comprehensive EHR system. Links between academic hospitals—where many specialist physicians already work—and smaller practices could help drive these developments.

The federal and state governments also have a role to play and are starting to play it. For the most part, it will be to help encourage the establishment of basic data requirements and technology standards for the system overall. This point is already a key element of a federal initiative. More directly, Congress and the state legislatures could set directions on policy and privacy issues and mandates on data reporting. They could also choose to subsidize "poorer" industry players: the Medicare system, for example, is experimenting with pay-for-performance methodologies, which if structured properly could offer physicians financial inducements to invest in the necessary technology and information. In addition, legislation recently introduced in Congress would support pay for performance by providing extra payments to hospitals, physicians, and other health care providers prepared to adhere to preestablished IT and data standards. A bill aimed at easing the load on providers would waive certain provisions of the Stark and antikickback laws, thereby allowing hospitals to provide physicians' practices with IT systems.

Such government-sponsored performance bonuses or subsidies could encourage small providers to invest in EHR but don't go far enough toward addressing the basic imbalance between costs and benefits for this group. One remedy would be for some of the private-sector bodies (such as large funders, health insurers, and pharmaceutical companies) that stand to gain most from the technological shift to catalyze the adoption of EHR by dipping into their very deep pockets to help pay for the effort.

Of course, whatever the equation's theoretical attractions, these companies still need a solid business case. Few will be willing to invest money now for the promise of a payout five to ten years down the road. Funders and health insurers are, however, already working with providers to establish standards for delivering better-quality care at a lower or more competitive cost, and these efforts are creating some regional data-gathering and -reporting networks. Pharmaceutical companies could enter into partnerships with regional health care providers to gather data (on procedures and outcomes) that would improve drug-development programs.

Funders, and perhaps health insurers, are also in a position to offer incentives of both the carrot and stick variety. On one level, they can continue to lean on providers to push the EHR agenda more vigorously—by paying the providers to tighten the link between cost and quality, for instance, or by reimbursing more generously those physicians in their networks who invest in EHR. They could also opt for a more aggressive approach: actually subsidizing or purchasing technology solutions for their independent physician partners. WellPoint, the largest publicly traded US health benefits company, recently offered PCs to members of its local physician groups to encourage the adoption of information technology (see "IT remedies for US health care: An interview with WellPoint's Leonard Schaeffer").

Such measures, combined with the growing population that has Internet access, the decreasing cost of technology, and greater technical savvy among physicians, will make it easier for providers and their practices, large and small, to move toward EHR. Those that choose to stay out risk losing in the marketplace as payers and consumers of health care begin to base their choice of doctors, clinics, and hospitals on cost and quality metrics.

Health care has always been a slow-moving industry, and the shift toward a national EHR system will proceed at an evolutionary pace. But the system will be built; the benefits to individual participants and the overall system are too great for it not to be. Right now, all industry players, as well as the government, must begin to understand the mismatch of costs and benefits that is hindering progress and collaborate to resolve it.