Despite Regulatory Changes, Hospitals Cautious in Helping Physicians Purchase Electronic Medical Records

Federal Rule Changes Opened the Door, but Hospitals Moved Slowly on Physician EMR Subsidies

Community physicians in the mid-2000s were notoriously slow to adopt electronic medical records (EMRs). The costs of purchasing, installing, and maintaining such systems fell squarely on individual practices, and many doctors concluded the expense was not worth it. Federal policy makers, hoping hospitals could serve as a catalyst for broader adoption, relaxed long-standing fraud and abuse regulations in August 2006 to let hospitals subsidize physician EMR purchases. Yet a Center for Studying Health System Change (HSC) study -- Issue Brief No. 123, September 2008, by Joy M. Grossman and Genna R. Cohen -- found that most hospitals were still on the sidelines more than a year after the rules took effect.

The Regulatory Backdrop

Two federal statutes -- the Stark physician self-referral law and the anti-kickback statute -- had long prohibited hospitals from giving financial incentives to referring physicians. The Department of Health and Human Services carved out a new IT exception to the Stark law and a corresponding safe harbor to the anti-kickback statute, both finalized in August 2006. In May and June 2007, the IRS added clarifications confirming that hospital compliance with these provisions would not violate federal tax law.

Under the revised framework, hospitals could cover up to 85 percent of the upfront and ongoing costs of EMR software and related IT support services for community physicians. Physicians had to pay the full cost of any hardware and at least 15 percent of other expenses. The EMR had to be interoperable -- capable of communicating with multiple IT systems, not just the subsidizing hospital's own network. Regulators built in that interoperability requirement to discourage hospitals from locking in referrals through proprietary technology. The provisions carried a sunset date of December 31, 2013, at which point physicians would need to absorb all ongoing EMR costs.

What Hospitals Were Actually Doing

HSC's findings came from follow-up interviews conducted in late 2007 and early 2008 with executives at 24 hospitals across 12 nationally representative metropolitan communities. Only seven of the 24 reported actively pursuing a strategy to provide financial or other support to community physicians for EMR purchases. Just four had begun or scheduled implementation. The remaining 17 were at various stages of planning and evaluation, with none expected to launch a program before 2009.

The strategies under consideration fell into three broad categories. Eleven hospitals were thinking about direct financial subsidies -- covering a portion of EMR software costs, usually coupled with volume discounts the hospital had negotiated with EMR vendors for its own employed practices. Some planned to subsidize less than the 85 percent Stark maximum, or to phase out the subsidy over time so physicians would have, as one executive put it, "more skin in the game." Five hospitals that had ruled out direct subsidies were considering extending their vendor discounts to community physicians along with IT support services like training, technical help, and data storage. The remaining eight were either exploring support services only or had not settled on an approach.

Two Motivations: Better Care and Closer Ties

Hospital leaders across the board pointed to two driving forces behind their interest. The first was quality improvement. By coupling greater physician EMR adoption with electronic data exchange between the hospital and office-based practices, hospitals saw the potential for more complete patient records, fewer redundant tests, and better-coordinated care. One executive described clinical integration through a robust EMR as "common sense and good care."

The second motivation was physician alignment. Hospital executives expected that once doctors depended on electronic interoperability with a particular hospital -- lab results flowing into practice records, imaging reports arriving automatically, discharge summaries populating charts -- they would be less inclined to shift referrals to a competitor. As one respondent said bluntly, if a physician's labs and X-rays flow easily into their records, "it will make it less likely they'll take their business across the street." This alignment strategy echoed earlier HSC findings about hospitals using web-based portals to bind physicians more closely.

Why Progress Stalled

Several obstacles slowed implementation. Hospitals that were willing to subsidize EMRs generally planned to start small -- targeting around 100 physicians across roughly 10 practices in an initial phase, although the range stretched from a two-doctor pilot to a 400-plus physician program. Budget limitations were a persistent constraint, and IT staffs were already stretched thin by hospital-based system implementations and EMR rollouts for hospital-owned practices. As one respondent acknowledged, "We certainly have enough on our plates with what we already have to do."

Physician interest was lukewarm at best. Some of the initial enthusiasm evaporated when doctors learned they would still have to cover all hardware costs and a minimum of 15 percent of software expenses. One hospital executive noted that many people assumed the Stark changes would make it simple, but the reality was more complex: "You can't hand physicians money and let them do what they want -- there are parameters." Uncertainty about tax treatment further dampened enthusiasm; IRS guidance issued in 2007 had not resolved whether hospital subsidies constituted taxable income for the receiving physician. The 2013 sunset clause also gave doctors pause, since they would eventually need to shoulder the full ongoing cost of any system they adopted.

The Interoperability Dilemma

Building the electronic bridge between hospital systems and physician EMRs was widely viewed as a critical piece of the strategy, but it was also expensive and technically demanding. Hospitals took different approaches. A few planned for full interoperability through a single clinical data repository containing both hospital and ambulatory records in one unified patient file. Most others favored a two-step process: first getting physicians onto the ambulatory EMR product used by hospital-employed practices, then building the interfaces or portals needed to connect those records with the hospital's inpatient systems.

This raised a competitive tension. The tighter the electronic link between a physician and a particular hospital, the harder it became for that physician to split work across multiple unaffiliated hospitals. Doctors who practiced at more than one institution faced the prospect of fragmented records and duplicated effort. A physician who accepted a hospital-sponsored EMR might find it difficult to store records for patients treated at rival facilities. The single-repository model offered the greatest clinical benefit but also created the highest switching costs, meaning a doctor who later wanted to change hospital affiliations could face significant barriers to transferring patient data.

Broader Implications for the Health System

The HSC researchers concluded that the regulatory changes had produced only a modest effect on hospital behavior. Strategies like extending vendor discounts and offering IT support services could have been pursued without the Stark exception. The communities most likely to see action were larger metropolitan areas where bigger hospital systems with both the resources and the strategic ambition would undertake phased, small-scale rollouts. In any given community, only a small fraction of physicians was likely to be affected in the near term.

Competitive pressure could accelerate adoption -- hospitals reported they would move faster if a rival launched an EMR program first. The approaching 2013 sunset was another potential accelerant, as was hospitals' own progress on internal IT projects that would free up bandwidth. But the ultimate outcome depended on two things: what hospitals were willing to offer and whether physicians would actually accept the help. Given the out-of-pocket costs, the compliance requirements, and the unresolved tax questions, physician uptake was far from guaranteed.

There was also a broader system-level tension. Hospital-sponsored EMR networks could improve care quality and even serve as building blocks for community-wide health information exchanges (HIEs). Several respondents suggested their proprietary networks might eventually form the backbone of an open exchange accessible to all local providers. Yet the competitive dynamics that were already impeding HIE development in many communities suggested that rival hospital-sponsored networks might actually reduce the incentive for a truly open exchange, since physicians who could already share data easily with their primary hospital might see less need for a broader system.

Regardless, the researchers emphasized that technical IT implementation alone would not deliver the integrated care delivery that policy makers envisioned. Hospitals and physicians would also need to redesign clinical workflows, restructure financial incentives, and commit to the operational changes required to make electronic records a tool for genuinely better medicine rather than just a digital filing cabinet.

Data Source

This research was part of HSC's Community Tracking Study, which conducted site visits every two years in 12 nationally representative metropolitan communities: Boston; Cleveland; Greenville, S.C.; Indianapolis; Lansing, Mich.; Little Rock, Ark.; Miami; northern New Jersey; Orange County, Calif.; Phoenix; Seattle; and Syracuse, N.Y. In 2007, a total of 453 interviews were conducted across these communities with representatives of health plans, hospitals, physician organizations, employers, benefit consultants, brokers, community health centers, consumer advocates, and state and local policy makers. Follow-up telephone interviews were then conducted between October 2007 and January 2008 with 24 executives representing two of the largest hospitals or health systems in each market, focusing specifically on the extent to which hospitals were helping community physicians purchase EMRs. Safety-net hospitals were excluded from the study.

Sources and Further Reading

Health Affairs -- Hospital-Physician Portals: The Role of Competition in Driving Clinical Data Exchange -- Earlier HSC research on hospital strategies for electronic physician alignment.

CMS -- National Health Expenditure Data -- Official data on U.S. health spending trends.

Kaiser Family Foundation -- Health Costs -- Analysis of health care costs and spending.

Health Affairs -- Peer-reviewed health policy research.

Robert Wood Johnson Foundation -- Health policy research and programs.