Requirements for Helping U.S. Physicians to Adopt EHRs

Numerous regulatory and private activities focus on increasing physicians' use of EHRs. Experience has shown that success in physician adoption requires much more than simply paying for the cost of software as implemented by the typical vendor. This research document focuses on the considerations that enable successful adoption. It is written from the point of view of an organization that sponsors EHR adoption for physician practices by providing subsidization, implementation support, or both. Sponsoring organizations may be hospitals, healthcare payers, health information exchanges, state and local governments, extension centers funded under the ARRA, and other nonprofits set up for this purpose.

The ARRA includes incentive payments of up to $44,000 through Medicare and Medicaid for physicians that adopt EHRs by 2012. In addition, it provides for grants to establish extension centers to support adoption, modeled after agricultural extension centers. Other grant programs support the rollout of IT infrastructure necessary for interoperable EHRs. Prior to the ARRA, the U.S. Department of Health and Human Services (HHS) issued regulations exempting donations from hospitals of up to 85% of the cost of EHR systems from the anti-kickback statute, the Stark federal physician self-referral law. Moreover, the U.S. Internal Revenue Service assured nonprofit organizations that they would not lose nonprofit status by making such donations.

Some other sponsoring organizations already had subsidy programs under way or planned. Although ARRA somewhat changes the economics, it did not eliminate the ability to provide other subsidies. Some of the same sponsoring organizations will continue to provide subsidies to reach out to physicians who:

Are likely to qualify for a subsidy under Medicare or Medicaid
Need incentives beyond those in the ARRA
Are unable or unwilling to pay for the EHR in a given year, hoping for incentive payments in future years

This report centers on the U.S. market. Therefore, it follows the terminology set forth by the U.S. government laws and regulations. It uses the term "electronic health record" to describe a system that contains clinical information about individual patients where the information is controlled by a single provider or healthcare organization. Although the U.S. definition and much of the applicable legislation apply to both ambulatory and hospital settings, this report applies to EHRs to be used in physician practices, because this is where the need for support is most acute. Gartner generally prefers the term "electronic medical record" (EMR) for systems that are targeted at the ambulatory setting (see "Global Definitions of EHR, PHR, E-Prescribing and Other Terms for Healthcare Providers").

The ARRA stipulates that physicians must make "meaningful use" of an EHR to receive the incentive payments. The law leaves the definition of "meaningful use" to regulations that will be issued by HHS. As of the publication of this research, these regulations have yet to be issued. The regulations will likely use criteria for meaningful use that benefit the overall healthcare system, focusing on information exchange to improve patient safety and measure quality. The regulations will likely set a low bar initially so that most physicians who install and use certified EHR products will qualify for the stimulus.

While meaningful use will be important in terms of qualifying for incentive funds, sponsoring organizations must keep in mind the overall trade-off between the benefits and deleterious impacts of an EHR from the specific point of view of the physician. In this report, we use the term "effective use" to describe the situation where a practice has selected and implemented an EHR so that, on balance, the EHR is perceived as making the physician's life easier rather than more difficult. The requirements for effective use are distinct from the eventual definition of "meaningful use."

This might not be an issue except that the history of EHR adoption in the U.S. has included many efforts that completely failed or resulted in insufficient improvements to justify the cost and disruption to practices. These difficulties have led to physician resistance to EHRs that can almost be characterized as "EHR rage." Implementations that are targeted solely at achieving "meaningful use" will likely fail. If physicians' use does not rise to effective use, their disenchantment will cause them to give up altogether. As their dissatisfaction is passed among their peers by word of mouth, the number of physicians who will opt not to install an EHR will likely increase, despite the availability of incentive payments.

Of course, this does not mean that organizations that assist physicians in getting EHRs can ignore achieving meaningful use. It just means that simply helping physicians access the incentive payments is not sufficient to achieve long-term EHR use or create a long-term affinity with physicians.

In this research document, we examine the strategic value that sponsoring organizations can achieve, as well as the broad capabilities of EHR products that are critical to achieving effective use. However, we do not provide an exhaustive rundown of features or products. Instead, we highlight specific characteristics of products that differentiate their value to the subsidizing organization.

We also look into critical organizational practices that will help those that subsidize or assist in implementation to ensure that the use of EHRs is effective for the physicians. Finally, we offer recommendations to spur physicians to install and use EHRs.

Several categorical sources of funds enhance the rollout of EHRs. Each of these categories is driven by different considerations. They include:

Governmental agencies that seek to introduce changes to healthcare to slow the growth in costs, ensure better access, and improve quality
Private healthcare payer organizations wanting to demonstrate to their customers (primarily, employers) programs that slow the growth in costs and improve quality
Large care delivery organizations (CDOs) attempting to "bond" with physicians or structure an integrated care process that coordinates the work of community physicians

Programs that are responsive to these drivers must not lose track of the needs of the individual physicians and practices. If they do not look at the features and support requirements of these individual practices, they will not be able to force acceptance of EHRs by physicians and will have little chance of achieving the broader goals. This is a primary tenet of this report.

One of the long-term motivators in subsidizing EHRs is the broad concern that the U.S. supply of physicians is growing far slower than the workload required to serve the aging population. Last year, J. M. Colwill and others estimated a deficit of 35,000 to 44,000 adult care generalists by 2025. The problem will become even more acute if healthcare reform creates access to healthcare for the 45 million or more Americans who are currently not covered. Generalist doctors have been buffeted by:

Falling incomes
Pressure to see more patients
Increasing preoccupation with administrative and coordination issues
Growing pressure to measure specific aspects of the quality of the care that they give, despite their not having the tools to gather the data

Moreover, it appears that new physicians entering the medical workforce are not as willing to accept the all-consuming work ethic that was embraced by prior generations.

Many sponsoring organizations face problems as a consequence of physician shortages. If their regions have difficulty competing for the dwindling supply of doctors, the enterprises cannot function well.

As a rule, it is not within the power of the sponsoring organization to directly increase the fees associated with providing care. Therefore, it is in their interest to make the practice of medicine more comfortable for independent physicians. This includes helping them to be more efficient and enabling them to focus more on patient care instead of administrative chores.

Helping physicians to acquire and use EHRs can help attract and retain physicians. Even when physicians are unsure that an EHR will deliver benefits, they increasingly acknowledge that getting an EHR is inevitable, and they welcome financial and process assistance in doing so.

The Obama administration has made it clear that the stimulus funding for healthcare IT is "a down payment on healthcare reform." Whatever form healthcare reform takes, it will probably include goals of achieving better care coordination, measuring quality, paying for measures that are likely to improve quality, or even paying a fair amount for "cognitive medicine" versus procedures. These goals cannot be met if data is illiquid — locked up in paper charts or even in silos of individual electronic medical record systems.

These factors offer the most danger of driving physicians away from EHRs. If meeting the goals compels physicians to spend ever more time documenting cases for years before healthcare reform propels fundamental changes in how they are paid, they will likely reject the EHRs.

Large CDOs in competitive areas recognize that the "soft" benefit of making it more desirable for physicians to practice in their areas is difficult to justify. After all, competitors benefit from the investment even if they fail to make a comparable investment. This often leads to discussions of "physician bonding" — in other words, using an EHR subsidy to attract physicians to admit patients to the CDO hospitals and use the CDO's diagnostic services.

Physicians will bond to a CDO if that entity:

Helps them practice more efficiently
Shares information with them
Makes it easy for them to work with the CDO
Helps them financially each year

The notion of physician bonding is undergoing a long-term change from simply ensuring that community physicians refer to the specialists and secondary-care facilities of large CDOs. Increasingly, leading-edge CDOs are attempting to take on the coordinating role of the medical home and enlist participating physicians in standards of care branded to the large CDO.

Physician bonding can be a strategic benefit, although it requires some finesse. Physicians are protective of their practices' autonomy. Even where they know that they conduct all or most of their business with one hospital, they are leery of an arrangement that would lock them in by making it difficult for them to receive electronic reports from competitive hospitals.

Physicians are even more concerned about their business data being accessible to hospitals or other sponsoring organizations. Finally, although they welcome help in choosing an EHR and learning to use it, they will react negatively if they sense that a single product or set of workflows or templates is being forced on them.

Physicians who are the targets of EHR subsidies are rarely the sole source of care for patients. Indeed, one of physicians' primary sources of lost time, distraction, and frustration is the coordination of care. Based on minor variations in administrative and clinical characteristics of a case, their patients will take different paths among the large array of care delivery entities that comprise the healthcare "system" in their area.

For this reason, Gartner believes that no EHR subsidy program will be effective if it focuses only on improving operations within the practice that receives the EHR. This may appear contradictory to the overall theme of this research report, which is to focus on effective use from the physician's point of view. However, this is not a real contradiction. Among the tasks that beleaguer physicians is coordinating care across the community. A well-implemented EHR with community information sharing will be appreciated by physicians as making them more effective.

The previous section notwithstanding, there is some evidence that physicians are softening in their attitudes about open sharing of clinical data with hospitals and other clinicians in a community related to the hospital. This occurs primarily where the sponsoring organization is a hospital that is taking on an active role in care coordination, payment bundling, or even providing "brand name" standards for the care of specific conditions. Such hospitals are finding community physicians more willing to become participants of a "single chart" for the patient, maintained on the hospital's enterprise EHR. To date, we have found this paradigm-shifting phenomenon primarily in hospitals that use Epic's EHR.

To be accepted, EHR products must contain features that create ongoing value directly to the practice. That is, they must make the physician more effective in areas that matter to the physician (not only in matters that are critical to the sponsoring organization). If such features are absent in a product or are implemented improperly, physicians are unlikely to accept EHRs or continue to use them.

EHR products must offer a range of options for physician documentation, including structured entry, voice entry, external transcription, and scanned paper. For the most part, physicians will end up with a mixture of these attributes.

Moreover, the EHR product must help make the physician more effective by providing answers to these questions:

How will the proposed product affect the time taken in the physician's work day?
Will it improve the physician's revenue or profits?
How will the product support physicians in meeting the growing mandates for quality measurement data?
How will it help the physician improve the patient's care experience?

Acquiring an EHR will offer many physicians their first opportunity to use population data about their patients. The product must make this easy for physicians who are not technologically sophisticated. Moreover, it must support highly configurable alerts to capture quality data during documentation.

One of the most tedious aspects of capturing data for quality mandates is documenting contraindications. Therefore, the EHR must appropriately identify situations in which a recommended therapy or diagnostic step is not indicated for a specific patient and include an easy option to document the reason.

Patient expectations are growing about online appointment scheduling, e-visits, physician-endorsed content, and bill reconciliation. EHR products vary substantially in their ability to support these capabilities as an integral part of using the EHR and associated practice management capabilities.

Gartner does not recommend establishing a partnership with an EHR vendor, unless it offers a truly integrated product that includes both practice management and EHR. Moreover, we recommend replacing existing practice management systems with the integrated product when installing the EHR (see Figure 1).

The most immediate benefit of integrating practice management and the EHR is that the workflow that follows a patient through encounters merges between the two systems. The traditional notion that these transitions could be handled by interfaces does not support clinical and financial optimization. The next section explores the reasons for this.

Administrative people often need access to clinical data — for example, when they are establishing referrals. Furthermore, some clinical decisions are informed by administrative data, such as the health plan associated with the appointment, the patient, or the patient's spouse. In some cases, the ability of physicians to offer e-visits is health-plan-specific. On occasion, the justification for failing to take a routine precaution (such as recommending aspirin) may only need to be collected by the quality programs of some payers.

A major complication in the practices of generalist physicians is the requirement to create and justify evaluation and management codes for most visits. If physicians undercode a visit, they get paid less, but if they cannot justify a higher level of coding, they risk jail for fraud. In the long term, healthcare reform will hopefully create a payment approach that does not put physicians through the requirement for such tedious accounting. However, in the near term, physicians will continue to have to chart information to support the bill, requiring tight coupling between the EHR and the billing system.

The linkages between administrative and clinical aspects of the case are very likely to grow. It is a strategic necessity to ensure that detailed data in a common data model is available to administrative and clinical functions. Finally, it is possible to subsidize an integrated system under Stark relief.

Another reason for a practice to use a combined EHR and practice management system is the potential of such a system to enhance the patient experience through a Web portal. Patients do not distinguish between administrative and clinical interactions and will likely have a better experience if they access online information from a single application instead of two separate ones.

Of course, some large practices may be satisfied with their practice management product and be unwilling to make the transition. Subsidizers should be prepared to support a few exceptions.

In a separate initiative under HIPAA, the U.S. government has mandated that physicians move from the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) to ICD-10-CM for submitting diagnoses on bills and other administrative transactions. This must be accomplished before 1 October 2013.

ICD-10 identifies twice as many diagnoses as ICD-9 and 20 times as many injuries. If nothing else, it represents the state of clinical knowledge of the 1990s rather than of the 1970s.

The new code set promises more accuracy in interchanges with payers and more-precise statistical insight into business and clinical activities within the institution. It may also enhance clinical decision support. In academic medical centers, it can simplify cohort formation and other aspects of research. ICD-10 is also logically organized and has room for expansion as medical knowledge grows.

These benefits notwithstanding, ICD-10 diagnoses represent a challenge to existing workflows in physicians' practices. Previously, physicians relied on a combination of check-the-box "super bills" and billing coders who had years of experience to create the diagnosis code that would not run afoul of payer rules for reimbursement. Forms and the professional knowledge locked up in coders' heads will have to change. For some specialties, the larger number of codes will make the use of the super bills and manual coding impractical.

Physicians, particularly in small practices, are unlikely to be aware of the pending conversion to ICD-10. More significantly, they are very unlikely to appreciate the impact of this change on their practices.

When selecting a combined EHR and practice management system, physicians should understand the upcoming challenge of ICD-10 and the particular benefit that may be obtained when the process of constructing the bill is assisted by software that has access to a coded clinical record of the encounter. Physicians should evaluate candidate vendors based on their demonstrated ability to meet this need.

Sponsoring organizations should use their technical and functional knowledge to distinguish between hype by vendors and demonstrated functionality. They should help physicians understand the difference and recognize the importance that an effective ICD-10 capability will have on their practices.

A worst-case scenario is to select a product that is weak in this area and be in the middle of implementation in 2011 as payers start to reach out to physicians to alert them to the conversion. While it is too early to evaluate vendors' implementations of ICD-10, sponsoring organizations should question candidate vendors on their proposed approach. If they cannot demonstrate that they have already thought this through, they should strongly consider other vendors.

E-prescribing allows a provider to electronically send an accurate, error-free, and readable prescription directly to a pharmacy from the point of care. An EHR must enable a physician using e-prescribing to:

Enter prescriptions directly into the EHR, where information is saved as structured data.
View the patient's full medication history available from a source that consolidates prescriptions written by all physicians, while evaluating the patient and creating new prescriptions.
Alert the physician to allergies and dangerous medication interactions when a prescription has been entered.
Alert the physician to administrative issues, such as a medication not being on the formulary associated with the patient's health coverage.
Print a legible physician signature that meets U.S. Centers for Medicare & Medicaid Services (CMS) anti-tampering requirements.
Fax prescriptions to the pharmacy or mail order fulfillment firm of the patient's choice, when requested by the patient or if the pharmacy does not yet accept online prescriptions.
Transmit prescriptions to drugstores or mail order fulfillment firms as structured data.

Most of these components of e-prescribing contribute to physician effectiveness. However, from the physician's viewpoint, that last point is dubious. Today, structured prescriptions cannot be used, by law, for approximately 15% of the prescriptions that are for controlled substances. Moreover, many drugstores do not yet accept structured prescriptions, and there are other situations when the patient does not want the prescription to be sent electronically.

Physicians cannot tolerate different workflows according to whether electronic transmission works or doesn't work. However, they will be satisfied if the EHR chooses when to transmit electronically so that the physician's workflow is essentially unchanged. Despite the dubious value to the physician, actually transmitting structured prescriptions is likely to be a key criterion for meaningful use. Therefore, getting this capability actually implemented in a manner that is not disruptive to the physician is critically important.

An EHR reaches its full benefit only when information flows directly in and out of it. This is true when considering value in terms of enhanced physician effectiveness and the value to the healthcare system of providing incentives for physicians to use EHRs.

However, it is important to recognize that enhancing physicians' effectiveness requires that the information exchanges with external entities be integrated into the workflow of the physician and the office staff. Initially, the single biggest requirement has been to receive structured lab data that permits trending analysis, detects critical values, and supports physician sign-off.

Increasingly, physicians seek the same convenience for specialists' reports, discharge notes, and other reports about the patient's health status. They also desire a complete problem list, medication list, and allergies.

Physicians frequently have Web-based referral products for one or more health insurance plans or other referral managers. However, they recognize that "swivel chair automation" between the Web site and the EHR is a substantial source of error and patient and staff dissatisfaction, because the practice is unable to manage the workflow of obtaining referrals.

Ideally, all EHRs would interoperate with all other data sources and receivers automatically. As well, the networking necessary for that to occur would be provided by a single agency for the whole community. However, that ideal is currently unattainable.

The leading EHR vendors, particularly those that are remotely hosted, can add substantial value by arranging for interoperability with EDI networks, major reference labs, and e-prescribing networks. This is an important feature for immediate benefit and strategic value, and the requirement to participate electronically in a broader community will only grow over the years. Any program to assist physicians in acquiring EHRs should select products from vendors that offer this capability and ensure that interoperability features are actually implemented. Some of these interoperability features will likely be included in the regulations that define meaningful use.

A long-recognized dilemma in architecting EHR systems has been the trade-offs of supporting subsecond interactions on functionally rich screens, versus remotely hosting products to avoid depending on the practice staff for reliable and secure computer operation.

In the past, physicians favored in-practice, locally administered systems. More-recent products, however, have moved toward remotely hosted, Web-based applications. Increasingly, Ajax and other rich Internet architectures provide the required level of interactivity for remotely hosted applications. Moreover, in urban and suburban locations, relatively low-cost, carrier-managed, commercial-grade Internet connections are widely available.

In addition, voice entry of text has crossed one important threshold — physicians prefer using it to tapping on a keyboard. The September 2008 release of Dragon NaturallySpeaking Medical v.10 shows promise in moving physicians to use voice as means of entering structured data.

"EHR lite" products give physicians some benefits of EHRs without necessarily replacing the paper chart or automating the full workflow of a practice. These products are generally remotely hosted, strongly emphasize interoperability, and are available to physicians at a modest cost. Examples include products that automate the receipt of reports from labs, diagnostic practices and hospitals, and remotely hosted community registries funded by independent physician associations or other entities that manage population risk.

Because the key benefits of EHRs arise from interoperability, it is not surprising that some health information exchange products compete with EHR adoption at the community level. For example, when the Axolotl health information exchange product was introduced in Santa Cruz County, California, through an independent physician association, it caused some physicians to stop using EHRs. Ten years later, however, more than 60% of the physicians in Santa Cruz County use EHRs.

Until the ARRA was passed, many subsidizers were considering a dual-track approach — subsidizing EHRs and enabling EHR lite applications without providing the subsidy. They currently have this consideration on hold, while they wait to see the regulations that define certification and "meaningful use."

Although an emphasis on EHR lite could be the best approach for achieving interoperability across most practices, Gartner believes that the deadlines associated with the ARRA are prohibitively tight for such a sharp change in direction. Therefore, we advise sponsoring organizations to put plans for EHR lite on hold. Alternatively, an EHR lite approach could be implemented if the economics are shown to be better when using EHR lite and forgoing physician incentive payments for use of certified EHRs.

At the same time, it is not out of the question that HHS could suspend some certification criteria for product functionality and establish meaningful-use criteria based entirely on interoperability. Sponsoring organizations should have a contingency approach in mind for this possibility. This could entail selecting and supporting the implementation of health information exchange products for communities of physicians or establishing "lite" implementation approaches for full EHR products.

The initial implementation of the EHR must focus on simple accomplishments that will deliver short-term benefits to physicians. They may not, for example, emphasize capturing encounters in a detailed structured form, even though this may limit the ability of the EHR to provide clinical decision support.

Many physicians will overdo simplicity; they will attempt to shortcut the recommended implementation processes. If this goes unchecked, the physicians could find that the EHR is more trouble than it's worth. For example, attempting to automate a bad workflow does not make a good workflow.

If a practice attempts to implement an EHR without reworking the workflow, the staff will "work around" the system by fooling it or simply ignoring it. In such a situation, there is little hope that the EHR will actually improve the practice's efficiency or job satisfaction.

The same is true for setting up templates. If the templates used to collect encounter data are so simple that they fail to capture the fundamental information necessary to ensure that the physician avoids undercoding, the financial benefits of the system will not be realized.

The keys to effective implementations are:

Establish specific and immediate goals.
Where achieving the goals implies changes in workflow, manage the implementation to this change; do not let physicians evade necessary workflow changes.
Use an approach that focuses on the immediate goals and achieves them as soon as is realistically possible.
Identify specific incremental gains that should be deferred until the physicians and their staff have initial experience using the system. Follow up a few months after the system goes live to initiate changes in workflow and configuration to obtain the extra benefits.

CIOs, CMIOs, and CMOs of large healthcare providers, as well as CEOs and program directors of state and local governments and nonprofits:

Establish partnerships with EHR vendors that place the overall responsibility for implementation on the sponsoring organization rather than on the vendor.
Don't rely on vendors to determine the readiness of practices for EHR implementation; the organization providing support must work with the practice to make the determination.
Take and maintain a critical role in advising physicians, monitoring progress, and sustaining the ongoing relationship.
Choose products that integrate practice management and EHR capabilities.
Strongly encourage practices to replace practice management systems with an integrated product.
Evaluate and advocate products that embrace technologies that facilitate reliability, networking options, and remotely hosted EHRs that do not require a thick client.