Evidence-based medicine is often admired but seldom practiced. It is rarely practiced because few physicians have the time to critically appraise the medical literature; an unfortunate reality considering the impact on quality of care, and the fact it could be changed.
When integrated into an electronic health record (EHR) system, evidence driven decision support is presented to the physician at the point of thought, providing crucial evidence-based literature that promotes timely and informed medical decision making. Further integration with a single platform solution that includes an electronic prescribing module provides the physician with objective, medication therapy decision support at the point of prescribing.
Integrated seamlessly into a practice’s EHR, e-prescribing provides additional information, including the cost, efficacy and adverse effects of various medication-based therapeutic alternatives to help the physician make the best prescribing decisions.
Three published estimates suggest that physicians are directing 80% of the spending in our $2 trillion health care market. Yet if you consider the information that we physicians bring to these spending decisions, frankly, it is primitive and pathetic. Imagine physicians as purchasing agents with $2 million annual budgetary authority. Studies show that we physicians don’t know how much the drugs and diagnostic tests that we order cost, and we lack comparative information about their effectiveness and adverse effects. Furthermore, our compensation is largely disconnected from the quality and cost-effectiveness of our performance. Is it any wonder that the U.S. has the most expensive health care in the world, while perennially ranking near the bottom of industrialized countries in metrics like healthy life expectancy?
Health information technologies, especially EHR systems, are often promoted as the solution to much of what ails our health care system. The implementation of technology has become a powerful political issue, but in such a fractured healthcare system, adoption remains a constant struggle for small and medium-sized practices.
Evidence-based medicine promises to displace Authority Based Medicine, wherein practicing clinicians simply followed the recommendations of expert thought leaders in the healthcare community. These thought leaders were usually identified by their affiliation with distinguished academic medical centers with successful college football programs.
The actual practice of EBM requires clinicians to formulate carefully structured questions about clinical problems in specific patients, and to then perform medical literature searches to find valid randomized controlled clinical trials containing individuals who are representative of the patient being treated. This is a time-consuming process that demands up to one hour per question.
If the true practice of EBM takes too long and is not compatible with having a financially viable medical practice, and if most physicians lack expertise in critical analysis of sophisticated medical studies, then what can dedicated physicians do to improve the quality and cost-effectiveness of their care?
Answering the Adoption Question
The answer is advanced tools and economic incentives that together optimize health care outcomes for patients. The state-of-the-art in evidence-based medical practice involves the integration of context-specific rules-based clinical decision support messages into electronic health records.[i]
On the technology side, advanced EBM tools need to follow and intuitively present medical evidence in the traditional four-step organization. This is aggregation of relevant medical studies, synthesis of such data including reviews, synopses of key studies and systematic reviews, and summaries of the relevant synopses.
Clinical decision support systems use rules behind the scenes to link relevant messages about patient care to patients who met the characteristics of the rule. Simply presenting guidelines with no validation, or offering links to online textbooks is unlikely to improve healthcare outcomes.
For example, in 2004, Michael Fischer and Jerry Avorn published a study which showed that if all Americans aged 65 and older who had high blood pressure were treated with the drugs that the best evidence proved to be most appropriate, we would save $1.2 billion in annual drug costs. There would also be enormous additional savings due to fewer heart attacks, strokes and heart failure.
But 2005 HEDIS pay-for-performance (P4P) measures reward the lowering of blood pressure without distinguishing whether this happens with a calcium channel blocker that may worsen the patient’s five-year mortality, an expensive new drug with no five-year outcomes data, or with a thiazide diuretic that clearly improves five-year mortality.
Not only is simply supporting P4P guidelines inadequate, but many EHR vendors don’t even support guidelines in the work flow in the first place. It is rare for electronic health record companies to truly integrate evidence-based context-specific decision support information into the work flow of busy clinicians. EHR vendors will be reluctant to invest in building the decision support needed to improve the quality and cost of care unless the market demands it because higher reimbursement rewards it.
Thus the incentive carrot is as important as the technology. In England, the contract between the General Practitioners and the National Health Service has 18% of the physicians’ annual income at risk-dependent on their performance against 146 quality measures. In contrast, here in the U.S., physicians are lucky if even 2 or 3% of their income depends on their performance against quality measures. In America, pay-for-performance is a plastic carrot: it looks appetizing, but when you bite into it, there is little satisfaction. Like the Soviet Union in the 1970s, doctors will continue to pretend to perform in the clinical outcomes arena as long as the health care system pretends to pay them for outcomes.
Why should we have to pay physicians more to perform well? Isn’t this their job? In fact, one might object to the thesis that physicians need incentives-that they are mercenaries and not missionaries. The reality is that physicians need incentives to be able to afford the software, hardware and implementation costs of electronic health records. Primary care physicians practice under surprisingly severe economic pressures-many are struggling for financial survival.
Between 1995 and 2003, family physicians increased their billable productivity by 35% but received an 18% inflation-adjusted reduction in average income. Meanwhile, the number of American-trained physicians choosing careers in family medicine declined by 50%. And Medicare projects physician payment cuts of about 35% between now and 2015, while physician costs are expected to escalate by another 20%.
Payers are reluctant to offer generous payment for performance bonuses. Payers generally struggle with the “free rider” problem. Physicians tend to treat all patients in their practices the same. If one payer with a 20% market share finances an effective incentive program for evidence-based care, then most of the benefits will actually accrue to the sponsoring payer’s competitors. The IHA program in California is a notable exception insofar as multiple payers cooperated to overcome the free rider problem. Nonetheless, this multipayer coalition has grossly underfunded the bonuses. They typically amount to only a 1 to 2% of a physician’s income. Experts estimate that such bonuses should be in the 5 to 10% of income range to be effective.[ii] Some of the Hawaii Medical Service Association’s performance-based bonuses reach into this range, but HMSA has an 80% market share, and Hawaii enjoys an aberrantly strong sense of community that is lacking in the mainland. HMSA is the exception that proves the rule.
Many primary care doctors are likely unaware of the most powerful financial incentive for EHR adoption, which also indirectly rewards evidence-based fiscally responsible care: CMS’ Medicare Advantage HMO plans. About 20% of Medicare recipients receive their care in these plans. They are funded through a prospectively risk-adjusted compensation formula, wherein the revenues that physicians receive for each patient in one year are determined by the severity of illness that they documented through their diagnosis coding for each patient during the previous year. In brief, doctors using an EHR with sophisticated Medicare HCC coding support may code more thoroughly and accurately, generating up to 30% more revenue than doctors who lack EHRs and do not pay attention to their diagnosis coding practices. The insurance company administering the Medicare Advantage plan often has a risk-sharing contractual agreement with physicians for a substantial percentage of this incremental revenue. Ironically, this powerful P4P program costs CMS nothing, since is a zero sum game; the risk adjustment model is rebalanced every 12 months. Doctors who code poorly and are paid poorly fund the incremental revenue for doctors who code well.
Employers are becoming increasingly interested in P4P programs; however, most businesses need to see a strong case before engaging in these programs on a large scale. One promising incentive for evidence-based care is the Bridges to Excellence (B2E) program. This program is offered in collaboration with the National Business Coalition on Health, NCQA, and Leapfrog.
The B2E program encourages physicians and physician practices to deliver safer, more effective and more efficient care by giving them financial and other incentives to do so. Thousands of physicians currently participate in the Physician Office Link, Diabetes Care Link, Cardiac Care Link and Spine Care Link programs. Because of the extensive reporting requirements, purchasers and their employees have the information they need to make better health care decisions while also obtaining cost-effective care.
The American Board of Internal Medicine, with support from B2E, is developing a new program called the Comprehensive Care Practice Improvement Module. This initiative will allow as many as 180,000 internists seeking to maintain ABIM board certification to send performance data collected through that process to B2E and eventually to other payers. Under this new partnership with ABIM, participating internists will qualify for maintenance of board certification, continuing medical education credits, and bonus payments under B2E’s new Internal Medicine Care Link program.[iii]
EBM Benefits – More Than Anecdotal
An EBM approach to care has clear benefits in quality of care, reduction of medical errors and cost savings. In peer-reviewed studies published in the Annals of Family Medicine and the Journal of Managed Care Pharmacy, an e-prescribing clinical decision support solution showed a significant impact on the cost and quality of patient care. The studies demonstrated a 12 percent savings in the costs of new prescriptions and refills, compared to contemporaneous control groups. The participating payer, Affinity Health Systems, enjoyed ongoing savings of $1,270 per doctor per month, relative to the contemporaneous control group, in pharmacy costs. In fact, there was remarkable consistency among the largest groups using the software. Their generic prescribing rates had all climbed to about 75%.
Another third-party study, performed in Maine by Anthem, found a savings of $3.55 per prescription, or some $470 per physician per month. Because Anthem was the payer for approximately 30% of patients in Maine, the total savings for all payers could be estimated to be more than $1,500 per physician per month in just one quarter.
Clinical decision support integrated with evidence driven data can make major impacts of patient safety. Take for example what happened at Esse Health, a 60-doctor medical practice in St. Louis, Missouri. The FDA often approves drugs that are appropriate for a small number of patients. Then, as we all know, direct-to-consumer advertising creates outsized demand for the medication. Earlier this decade, the majority of prescriptions in the U.S. (61 percent) for nonsteroid anti-inflammatory drugs (NSAIDS) were for the newer COX-2 inhibitors such as Vioxx and Bextra.
Using its e-prescribing software, Esse Health had been messaging its patients and physician-customers about the risks and limitations of these drugs since the CLASS and VIGOR studies came out in 2000. The group then tracked the prescribing patterns of doctors using the EBM clinical decision support software and found that 25% of all prescriptions for NSAIDs were for COX-2 inhibitors.
In September 2004, Vioxx was withdrawn from the market after studies showed the medication quadrupled the risk of heart attacks. In April 2005, Bextra was withdrawn from the market because it doubled the risk of heart attacks and strokes. Many e-prescribing companies took pride in being able to rapidly notify physicians when these drugs were withdrawn from the market and/or enable them to quickly message their patients to stop taking these medications.
While both these items are good things, Esse Health felt they were taking things a step further by scouring the medical evidence and seamlessly incorporating it into physician workflow at the point of medical decision making. Esse was ahead of the FDA in limiting the exposure of their patients to these risky drugs. The physician group knew that incorporating clinical decision support at the point of thought could make a measurable impact on patient safety.
Considering that 70% of U.S. healthcare is delivered by small and medium-sized practices, To realize rapid adoption of EBM, clinical decision support solutions should ideally be tailored to small and medium-sized practices, considering that 70% of U.S. healthcare is delivered by these groups. This means several things: eliminating the cost barrier; ensuring that workflow is not disrupted by infusing EBM into the workflow; making evaluation transparent, Web-based and convenient; and achieving rapid and non-disruptive implementation.
Clinical decision support with EBM is a critical factor in the success of Regional Health Information Networks as well as a larger National Health Information Infrastructure. Physicians prepared and supported in this manner can shift their focus from reacting to acute illness toward using data to proactively manage patients with chronic disease, as well as populations with specific diseases. More aggressive reimbursement reform that rewards fiscally responsible, high quality, evidence based care will save money for payers, representing a win-win proposition for physicians, payer and patients.
Tom Doerr, MD
Founder and Chief Medical Officer
R. Brian Haynes. “Of studies, syntheses, synopses, summaries and systems: the “5S” evolution of information services for evidence-based healthcare decisions.” Evid. Based Med. 2006;11;162-164
Laura Landro, “To Get Doctors to Do Better,Health Plans Try Cash Bonuses” Wall Street Journal, September 17, 2004 http://www.bridgestoexcellence.org/programs/upcoming.mspx (Accessed August 18, 2007).