Physician Capability to Electronically Exchange Clinical Information, 2011
Vaishali Patel, PhD, MPH; Matthew J. Swain, MPH; Jennifer King, PhD; and Michael F. Furukawa, PhD
The capability to electronically share and view clinical data has the potential to enable clinical information to follow patients wherever they go to seek care and thereby improve the safety, quality, and efficiency of healthcare.1 Despite promising benefits, historically physicians have not exchanged clinical information electronically due to the high costs associated with implementation and limited incentives for data sharing.2 Exchange activity has largely been confined to regions of the country where there are operational health information organizations that support clinical data exchange within their community.3 Furthermore, physicians have typically had to use stand-alone e-prescribing systems or proprietary portals that support the exchange of specific types of clinical data (eg, viewing lab data), which can be costly, difficult to incorporate into their clinical work flow, and possess limited capability to support integrated data as with an electronic health record (EHR).4-7
A number of federal programs and other initiatives are under way to help address some of these barriers. The Health Information echnology for Economic and Clinical Health (HITECH) Act of 2009 includes up to $22.5 billion in financial incentives for eligible professionals who demonstrate “meaningful use” of interoperable EHRs capable of electronic exchange. HITECH also awarded more than $540 million to the Office of the National Coordinator for Health Information Technology (ONC) State Health Information Exchange (HIE) Program, which provides support for state-designated entities to ensure mechanisms are in place to enable providers to exchange clinical information.8 Furthermore, ONC’s Health Information Technology Certification Program seeks to ensure that EHR products include functionality that enables electronic exchange.9 In addition to the HITECH incentives and programs, a public-private initiative provides relatively simple technical solutions to enable directed exchange between 2 known providers.10 A community of participants from the public and private sector focus on providing tools, services, and guidance to promote functional interoperability.11
In the first stage of meaningful use, it was sufficient for providers to perform a test to demonstrate their EHR’s capacity to electronically exchange information.12,13 Stage 2 meaningful use requirements related to HIE have evolved to become more advanced. Physicians must go beyond demonstrating capability to exchange; they must actually electronically exchange key clinical data among providers and patient-authorized entities. Additionally, physicians must demonstrate the capability to send summary-of-care documents electronically to recipients with a different EHR vendor.14
Yet little is known about current physician capability to electronically exchange clinical information at a national or state level, both of which are relevant in implementing ONC’s strategy and in assessing its potential for success. We used a nationally representative survey of office-based physicians conducted in 2011 to provide a snapshot of physicians’ capability to electronically exchange clinical information associated with key national priorities: pharmacy exchange(e-prescribing), laboratory exchange (including receipt of results and lab orders), and clinical summary exchange with patients and providers.15 This assessment provides both a portrait of exchange capability as of stage 1 meaningful use and a baseline for monitoring progress going forward as new policies and initiatives to accelerate HIE are implemented—in particular, stage 2 meaningful use. Future trends in physicians’ HIE capability could help assess the effectiveness of these policies. We describe physician exchange capability geographically across states and by EHR vendor. Finally, we examined the association between physician and practice characteristics, including adoption of EHRs, with physician capability to exchange different types of clinical information.
Data Source and Collection
We analyzed the 2011 National Ambulatory Medical Care Survey Electronic Medical Record Supplement (NAMCS EMR supplement), a cross-sectional nationally representative survey of nonfederal office–based physicians who provide direct patient care.16 The overall sample consisted of 4326 hysician respondents, with a 61% weighted response rate. The sample size is sufficient to generate state-level estimates.
Measures of Electronic Exchange of Clinical Information
The NAMCS EMR supplement asks physicians to report on their practices’ capability to electronically exchange key types of clinical information, including pharmacy data, laboratory data, and patient clinical summaries (eAppendix A, available at www.ajmc.com).
To assess the capability of physicians to send pharmacy data and to send and receive laboratory data, we examined the percentage of physicians with the ability to send prescriptions electronically, send lab orders electronically, view lab results electronically, and incorporate lab results into the EHR. The latter item provides an indication of whether physicians receive results in a structured format that enables them to view and track individual test results.
To assess the capability of physicians to exchange patient clinical summaries, we examined the percentage of physicians who indicated that they “exchange patient clinical summaries electronically with any other providers” by either receiving and/or sending patient clinical summaries. This measure goes beyond capability to exchange, capturing actual exchange activity.
We calculated estimates of the electronic exchange measures at the national and state levels, and examined whether physicians’ capability to exchange these types of clinical information at the state level differed from the national average. We used t tests to test for significant differences at the P <.05 level.
We used multivariate analyses to examine the association between the capability to exchange different types of clinical information and physician and practice characteristics, including EHR adoption and state fixed effects to control for confounding within each state. Given the dichotomous nature of the dependent variables, we estimated probit regression models and used the regression results to calculate the incremental effects associated with each independent variable (all were categorical variables). These incremental effects represent the percentage point change in the outcome that is associated with a given characteristic (compared with a reference category). Analyses were conducted using Stata 12.0 (StataCorp, College Station, Texas). We used weights to account for nonresponse and standard errors, which were adjusted to account for the complex sample design.
We assessed the extent to which exchange capability varied by vendor. We only included vendors that had at least 1% of the market share (representing a total of 55% of physicians with EHRs). Vendors with at least 1% of the market share included the following: Allscripts, Cerner, eClinicalWorks, Epic, eMDs, GE/Centricity, Greenway Medical, NextGen, and Sage.
Physicians’ capability to electronically exchange clinical data varied by type of information (Figure 1). More than half of all physicians (55%) reported that their practices have computerized capability to e-prescribe. A majority of physicians (67%) reported that they are able to view lab results electronically, but fewer physicians (42%) were able to incorporate lab results into their EHR. More than one-third (35%) reported they are able to send lab orders electronically. The computerized capability to provide clinical summaries to patients was reported by 38% of physicians.
Among those physicians who reported exchanging clinical summaries with other providers (31%), approximately three-fourths (76%) reported both sending and receiving clinical summaries (eAppendix B, available at www.ajmc.com). About one-fifth (19%) of physicians reported that they send clinical summaries only to other providers. Of all physicians who exchange clinical summaries with other providers (eAppendix C, available at www.ajmc.com), the most common method of access was through an EHR vendor (64%), followed by a hospital-based system (28%).
Physicians’ capability to exchange varies by EHR adoption status (Figure 1). A large majority of physicians with an EHR have the capability to send prescriptions electronically (78%) and view lab results electronically (87%). Substantial minorities of physicians with no EHR also have these capabilities (23% and 42%, respectively), highlighting the role that stand-alone e-prescribing products and proprietary portals continue to play in facilitating physician access to pharmacy and lab result data. However, very few physicians without an EHR have the capability to electronically exchange clinical summaries and lab orders.
Physician Exchange Capability by State
Physicians’ capability to exchange clinical information widely varies by state (eAppendix D, available at www.ajmc.com). The capability to receive lab results electronically ranged from 88% (Wisconsin) to 44% (Louisiana). The capability to then incorporate those results electronically into an EHR among those providers able to receive results electronically varied between 73% (Minnesota) and 21% (Louisiana). Electronic lab ordering ranged from 58% (Washington) to 19% (Delaware). The capacity to electronically exchange clinical summaries with patients varied from 55% (Minnesota) to 18% (Louisiana). The proportion of physicians who exchange clinical summaries with other providers varied from 61% (Wisconsin) to 15% (Alabama).
Although the rates vary widely across states, physicians in certain states exchange these data at consistently higher or lower rates compared with the national average (Figure 2). Physicians in 7 states reported the capability to exchange clinical information at rates significantly higher than the national average across at least 4 out of the 6 measures of exchange: Massachusetts, Minnesota, North Dakota, Oregon, Vermont, Wisconsin, and Washington. Conversely, Louisiana, Nevada, and New Jersey have significantly lower rates of physician reported exchange capability across at least 4 out of the 6 measures.
Physician and Practice Characteristics Associated With Exchange Capability
Practice Characteristics. Although a number of practice characteristics were significantly associated with physicians’ capability to exchange different types of clinical information (Table), the single strongest significant predictor was adoption of an EHR (P <.001). Adjusting for other physician and practice characteristics, physicians with an EHR were 49 percentage points more likely to have computerized capability to provide clinical summaries to patients and 37 percentage points more likely to exchange clinical summaries with other providers compared with physicians with no EHR. The association between EHR adoption and the capability to exchange pharmacy data and lab data were similarly strong.
Practice ownership was also an important predictor of exchange capability. Practices owned by health maintenance organizations or healthcare corporations were significantly more likely to exchange all 6 types of clinical data compared with physician-owned or physician group–owned practices. Community health centers were significantly less likely to have the capability to e-prescribe (P <.001) or receive lab results electronically (P <.05) compared with physician-owned or physician group–owned practices. Practices owned by hospitals or academic medical centers were significantly less likely to have the capability to e-prescribe (P <.001) but more likely to have the capability to receive lab results electronically (P <.001) and exchange clinical summaries with other providers (P <.001).
Larger practice size was significantly associated with the capability to conduct all facets of electronic lab exchange and was a particularly strong predictor of the capability to incorporate lab results into an EHR. Practice size was also significantly associated with the capability to e-prescribe (P <.001). However, practice size was not a significant predictor of capability to exchange clinical summaries.
Physicians practicing in multispecialty practices were more likely to have the capability to e-prescribe (P <.05), send lab orders electronically (P <.001), and exchange clinical summaries with other providers (P <.001) compared with singlespecialty practices. However, overall, practice type was not a strong predictor of exchange capability.
Physician Characteristics. Primary care physicians were more likely to have the capability to e-prescribe (P <.001) and engage in various forms of lab exchange (P <.001), as well as provide clinical summaries to patients (P <.05) compared with specialists. Age differences were not associated with most forms of exchange capability; physicians aged less than 50 years were only slightly more likely to possess the capability to provide clinical summaries to patients (P <.05) compared with older physicians.
Variation in Physician Exchange Capability by EHR Vendor
Among physicians using an EHR, capability to exchange different types of clinical information varied according to EHR vendor (Figure 3). Between 73% and 94% of physicians using an EHR reported that they possess the capability to eprescribe, depending upon their vendor. Most EHR vendors (between 80% and 99%) provide the capability to receive lab results electronically. Depending upon the EHR vendor, between 62% and 95% of physicians reported the capability to incorporate lab results into their EHR. Physicians’ capability to send lab orders electronically ranged from more than onethird (38%) to 87% depending upon the EHR vendor.
Large differences existed across EHR vendors with regard to capability to exchange clinical summaries. EHR vendor capability to provide clinical summaries to patients ranged from 22% to 90% of physicians. Between one-fourth (24%) and three-fourths (77%) of physicians reported the capability to exchange clinical summaries with other providers, depending upon their EHR vendor.
In 2011, a majority of office-based physicians possessed the capability to electronically exchange lab and medication data, and approximately one-third could exchange clinical summaries with patients or other providers. These estimates represent a baseline in measuring progress of national programs under way to promote greater exchange activity.
Variation in exchange capability may be driven by market factors and policy levers, as well as idiosyncratic work flow issues associated with each type of clinical data.17 Greater physician capability to view laboratory results and e-prescribe may be due in part to the role of large vendors (eg, Surescripts) and to hospital-physician portals.7 In addition to the implementation of meaningful use stage 1 requiring physicians to eprescribe using a certified EHR, other national policies, most notably the Medicare Improvements for Patients and Providers Act, have also likely spurred e-prescribing uptake.18,19 Lower capability to incorporate lab results into an EHR may be due to technical challenges in transforming unstructured laboratory results into a structured format that involves agreement on vocabularies and standards.20
Lower rates of physicians’ exchange of clinical summaries may reflect technical and operational barriers to care coordination.7 For example, some physicians reported that summaries are not integrated within their work flow, prompting them to fax information between providers with EHRs rather than transmit summaries electronically; some physicians also reported that standard clinical care summaries can contain extraneous information, making them difficult to use.4 Rates of exchange of clinical care summaries may also be lower because the survey assessed exchange activity as opposed to the capability to exchange. However, rates of clinical care summary exchange may increase with the implementation of stage 2 meaningful use, which requires physicians to exchange clinical summaries with other providers during transitions of care.
Although in theory EHRs are expected to facilitate exchange, few empirical studies have examined EHRs as a vehicle for exchanging data among providers. To date, stand-alone systems and proprietary web portals have facilitated accessto clinical data. Studies demonstrating the impact of HIE in an ambulatory care setting have largely been limited to portal users.21,22 Furthermore, EHRs have been criticized for their lack of interoperability, including their inability to facilitate exchange of data easily between providers.23 We found that a majority of physicians who have EHRs reported they have the capability to exchange different types of clinical information, and multivariate analyses demonstrated that having an EHR was the single strongest predictor of exchange capability for e-prescribing, laboratory test viewing and ordering, and exchanging clinical summaries. Although stand-alone systemsmay continue to play a significant role, our findings indicate that EHRs have the capability to exchange a wide variety of clinical information with patients, providers, and other entities, in addition to serving as a more comprehensive tool to improve patient care.
However, our findings indicate that EHR systems do not all offer equivalent exchange capability. Although most physicians with EHRs reported capability to e-prescribe and view laboratory results, we found wide variation across EHR vendors in capabilities for lab ordering and exchange of clinical summaries. This variation may be due to how recently the physician implemented the EHR and whether the product met certification criteria requiring it to support clinical summary exchange. Some EHR vendors have designed products using outdated technology that does not enable interoperability with other systems.24 Variation in HIE capability across vendors may diminish over time as physicians upgrade their systems or purchase newer systems that meet more rigorousstage 2 meaningful use certification criteria related to HIE. The requirement to demonstrate the capability to exchange clinical summaries across different vendor platforms may increase the number of vendors offering products featuring this functionality, paving the way for an increase in the proportion of physicians using EHRs to exchange clinical care summaries with other providers. Evidence suggests that after the implementation of stage 1 meaningful use requiring physicians to e-prescribe, the proportion of providers e-prescribing on the Surescripts network using an EHR increased almost 7-fold from 7% in 2008 to 48% in 2012.25
Another potential driver of HIE is healthcare payment and delivery reform. As payment and organization continues to evolve, rewarding providers who better manage patient populations and penalizing those that do not will likely spur demand for HIE and the ability to facilitate transitions in care electronically.26 Specifically, accountable care organizations, which allow entities to share cost savings, may create a business case for HIE by giving providers greater financial incentives to exchange information regarding their patients with each other. Additionally, financial penalties for high hospital readmission rates, for example, may spur greater care coordination between hospitals and ambulatory care providers to better manage transitions of care through the use of care summaries.27,28
Physician exchange capability also varied widely across states. Prior state health information technology investment, state legislation and regulations, and existing local/regional health information organizations may help explain some of this variation.29 Furthermore, variation in EHR adoption rates may also explain these differences. Except for Vermont, all the states with above-average (relative to national) exchange capability (across multiple measures) also have significantly higher levels of EHR adoption.30 Understanding how states such as Minnesota have facilitated higher rates of physician exchange capability across a number of measures may provide other states with best practices in promoting greater exchange activity. These findings suggest that ONC's State HIE program has an important role in ensuring that physicians, regardless of their geographic location, have the capability to exchange clinical information with patients and other providers. The measures examined as part of this baseline study of exchange capability are being used by the State HIE program to monitor the performance of states and set national goals.15
In addition to EHR use, other physician and practice characteristics were associated with greater exchange capability. However, these were not as strong predictors as EHR adoption, nor were the patterns necessarily consistent across the different types of clinical data. Smaller practices have previously reported experiencing barriers to HIE, and our findings suggest that smaller practices do have less capability to exchange with pharmacies and laboratories compared with larger practices; however, we did not find differences with regard to clinical summary exchange.31 Primary care physicians’ higher rates of exchange capability compared with specialists—which persisted after controlling for EHR use—may be driven in part by primary care physicians’ greater need to exchange information with other providers to coordinate and manage their patients care.28
This study has some important limitations. These findings are based on self-reported survey data, which could not be independently verified. The survey did not capture detail on all mechanisms or tools by which physicians may exchange clinical information, including stand-alone systems or proprietary portals. Our measure of pharmacy exchange was limited; the survey did not capture potential exchange from the pharmacy or pharmacy benefit manager such as prescription renewals, eligibility/formulary verification, and medication histories.32 Nor did the survey report on whether physicians exchanged information with providers outside their practice or organization. We were also not able to measure physicians’ capability to exchange data across different EHR vendor platforms, which is required for stage 2 meaningful use. In addition to examining trends in exchange capability, future research should examine physicians’ actual use of exchange functionality.33
This baseline study of physician capability to exchange key types of clinical information found room for improvement but also a foundation on which to build and expand physician exchange capability. Our findings suggest that while most physicians possess the capability to meet certain HIE-related meaningful use requirements such as e-prescribing, physicians are less prepared to meet other upcoming stage 2 core HIE requirements such as clinical care summary exchange. It will be important to monitor these key measures of exchange capability as ambulatory providers implement EHRs meeting meaningful use stage 2 requirements and as shifts in healthcare delivery require robust exchange of health information.