Author: MontereyBayForum

By JEROME GROOPMAN and PAMELA HARTZBAND

“Last week, President Barack Obama convened a health-care summit in Washington to identify programs that would improve quality and restrain burgeoning costs. He stated that all his policies would be based on rigorous scientific evidence of benefit.

The flagship proposal presented by the president at this gathering was the national adoption of electronic medical records — a computer-based system that would contain every patient’s clinical history, laboratory results, and treatments. This, he said, would save some $80 billion a year, safeguard against medical errors, reduce malpractice lawsuits, and greatly facilitate both preventive care and ongoing therapy of the chronically ill.

Following his announcement, we spoke with fellow physicians at the Harvard teaching hospitals, where electronic medical records have been in use for years. All of us were dumbfounded, wondering how such dramatic claims of cost-saving and quality improvement could be true.

The basis for the president’s proposal is a theoretical study published in 2005 by the RAND Corporation, funded by companies including Hewlett-Packard and Xerox that stand to financially benefit from such an electronic system. And, as the RAND policy analysts readily admit in their report, there was no compelling evidence at the time to support their theoretical claims. Moreover, in the four years since the report, considerable data have been obtained that undermine their claims. The RAND study and the Obama proposal it spawned appear to be an elegant exercise in wishful thinking.

To be sure, there are real benefits from electronic medical records. Physicians and nurses can readily access all the information on their patients from a single site. Particularly helpful are alerts in the system that warn of potential dangers in the prescribing of a certain drug for a patient on other therapies that could result in toxicity. But do these benefits translate into $80 billion annually in cost-savings? The cost-savings from avoiding medication errors are relatively small, amounting at most to a few billion dollars yearly, as the RAND consultants admit.

Other potential cost-savings are far from certain. The impact of medication errors on malpractice costs is likely to be minimal, since the vast majority of lawsuits arise not from technical mistakes like incorrect prescriptions but from diagnostic errors, where the physician makes a misdiagnosis and the correct therapy is delayed or never delivered. There is no evidence that electronic medical records lower the chances of diagnostic error.

All of us are conditioned to respect the printed word, particularly when it appears repeatedly on a hospital computer screen, and once a misdiagnosis enters into the electronic record, it is rapidly and virally propagated. A study of orthopedic surgeons, comparing handheld PDA electronic records to paper records, showed an increase in wrong and redundant diagnoses using the computer — 48 compared to seven in the paper-based cohort.

But the propagation of mistakes is not restricted to misdiagnoses. Once data are keyed in, they are rarely rechecked with respect to accuracy. For example, entering a patient’s weight incorrectly will result in a drug dose that is too low or too high, and the computer has no way to respond to such human error.

Throughout their report, the RAND researchers essentially ignore downsides to electronic medical records. Rather, they base their cost calculations on 100% compliance with the computer programs “adopted widely and used effectively.” The real-world use of electronic medical records is quite different from such an idealized vision.

Where do the RAND policy analysts posit major cost-savings? They imagine that the computer will guide doctors to deliver higher quality care, and that patients will better adhere to quality recommendations embedded in the computer programs. This would apply to both preventive interventions like vaccines and weight reduction, and to therapy of costly chronic maladies like diabetes and congestive heart failure. Over 15 years, the RAND analysts assert, more than $350 billion would be saved on inpatient care and nearly $150 billion on outpatient care. Unfortunately, data to support such an appealing scenario are lacking.

A 2008 study published in Circulation, a premier cardiology journal, assessed the influence of electronic medical records on the quality of care of more than 15,000 patients with heart failure. It concluded that “current use of electronic health records results in little improvement in the quality of heart failure care compared with paper-based systems.” Similarly, researchers from the Brigham and Women’s Hospital and Harvard Medical School, with colleagues from Stanford University, published an analysis in 2007 of some 1.8 billion ambulatory care visits. These experts concluded, “As implemented, electronic health records were not associated with better quality ambulatory care.” And just this past January, a group of Canadian researchers reviewed more than 3,700 published papers on the use of electronic medical records in primary care delivered in seven countries. They found no solid evidence of either benefits or drawbacks accruing to patients. This gap in knowledge, they concluded, “should be of concern to adopters, payers, and jurisdictions.”

What is clear is that electronic medical records facilitate documentation of services rendered by physicians and hospitals, which is used to justify billing. Doctors in particular are burdened with checking off scores of boxes on the computer screen to satisfy insurance requirements, so called “pay for performance.” But again, there are no compelling data to demonstrate that such voluminous documentation translates into better outcomes for their sick patients.

Even before these new data, there were studies casting doubt on the benefits of electronic medical records. In response, the RAND researchers boldly stated, “We choose to interpret reported evidence of negative or no effect of health information technology as likely being attributable to ineffective or not-yet-effective implementation.” This flies in the face of the scientific method, where an initial hypothesis needs to be modified or abandoned in the face of contradictory results. Rather than wrestle with contrary information, the report invokes the successes of computer-based systems in saving money in industries like banking, securities trading, and merchandizing, using ATM machines, online brokerage and bar-coded checkouts. Medical care of human beings — treatment of acute and chronic illnesses and the even more complex process of effecting lifestyle changes like smoking cessation and weight loss to prevent disease — is not analogous to buying bar-coded groceries and checking-account balances online.

Some have speculated that the patient data collected by the Obama administration in national electronic health records will be mined for research purposes to assess the cost effectiveness of different treatments. This analysis will then be used to dictate which drugs and devices doctors can provide to their patients in federally funded programs like Medicare.

Private insurers often follow the lead of the government in such payments. If this is part of the administration’s agenda, then it needs to be frankly stated as such. And Americans should decide whether they want to participate in such a national experiment only after learning about the nature of the analysis of their records and who will apply the results to their health care.

All agree skyrocketing health-care costs are a dangerous weight on the economic welfare of the nation. Much of the growing expense is due to the proliferation of new technology and costly treatments. Significant monies are spent for administrative overhead related to insurance billing and payments. The burden of the uninsured who use emergency rooms as their primary care providers, and extensive utilization of intensive care units at the end of life, further escalate costs.

The president and his health-care team have yet to address these difficult and pressing issues. Our culture adores technology, so it is not surprising that the electronic medical record has been touted as the first important step in curing the ills of our health-care system. But it is an overly simplistic and unsubstantiated part of the solution.

We both voted for President Obama, in part because of his pragmatic approach to problems, belief in empirical data, and openness to changing his mind when those data contradict his initial approach to a problem. We need the president to apply real scientific rigor to fix our health-care system rather than rely on elegant exercises in wishful thinking.

Drs. Groopman and Hartzband are on the staff of Beth Israel Deaconess Medical Center in Boston and on the faculty of Harvard Medical School.

By JEROME GROOPMAN and PAMELA HARTZBAND
The Obama administration is working with Congress to mandate that all Medicare payments be tied to “quality metrics.” But an analysis of this drive for better health care reveals a fundamental flaw in how quality is defined and metrics applied. In too many cases, the quality measures have been hastily adopted, only to be proven wrong and even potentially dangerous to patients.

Martin KozlowskiHealth-policy planners define quality as clinical practice that conforms to consensus guidelines written by experts. The guidelines present specific metrics for physicians to meet, thus “quality metrics.” Since 2003, the federal government has piloted Medicare projects at more than 260 hospitals to reward physicians and institutions that meet quality metrics. The program is called “pay-for-performance.” Many private insurers are following suit with similar incentive programs.

In Massachusetts, there are not only carrots but also sticks; physicians who fail to comply with quality guidelines from certain state-based insurers are publicly discredited and their patients required to pay up to three times as much out of pocket to see them. Unfortunately, many states are considering the Massachusetts model for their local insurance.

How did we get here? Initially, the quality improvement initiatives focused on patient safety and public-health measures. The hospital was seen as a large factory where systems needed to be standardized to prevent avoidable errors. A shocking degree of sloppiness existed with respect to hand washing, for example, and this largely has been remedied with implementation of standardized protocols. Similarly, the risk of infection when inserting an intravenous catheter has fallen sharply since doctors and nurses now abide by guidelines. Buoyed by these successes, governmental and private insurance regulators now have overreached. They’ve turned clinical guidelines for complex diseases into iron-clad rules, to deleterious effect.

One key quality measure in the ICU became the level of blood sugar in critically ill patients. Expert panels reviewed data on whether ICU patients should have insulin therapy adjusted to tightly control their blood sugar, keeping it within the normal range, or whether a more flexible approach, allowing some elevation of sugar, was permissible. Expert consensus endorsed tight control, and this approach was embedded in guidelines from the American Diabetes Association. The Joint Commission on Accreditation of Healthcare Organizations, which generates report cards on hospitals, and governmental and private insurers that pay for care, adopted as a suggested quality metric this tight control of blood sugar.

A colleague who works in an ICU in a medical center in our state told us how his care of the critically ill is closely monitored. If his patients have blood sugars that rise above the metric, he must attend what he calls “re-education sessions” where he is pointedly lectured on the need to adhere to the rule. If he does not strictly comply, his hospital will be downgraded on its quality rating and risks financial loss. His status on the faculty is also at risk should he be seen as delivering low-quality care.

But this coercive approach was turned on its head last month when the New England Journal of Medicine published a randomized study, by the Australian and New Zealand Intensive Care Society Clinical Trials Group and the Canadian Critical Care Trials Group, of more than 6,000 critically ill patients in the ICU. Half of the patients received insulin to tightly maintain their sugar in the normal range, and the other half were on a more flexible protocol, allowing higher sugar levels. More patients died in the tightly regulated group than those cared for with the flexible protocol.

Similarly, maintaining normal blood sugar in ambulatory diabetics with vascular problems has been a key quality metric in assessing physician performance. Yet largely due to two extensive studies published in the June 2008 issue of the New England Journal of Medicine, this is now in serious doubt. Indeed, in one study of more than 10,000 ambulatory diabetics with cardiovascular diseases conducted by a group of Canadian and American researchers (the “ACCORD” study) so many diabetics died in the group where sugar was tightly regulated that the researchers discontinued the trial 17 months before its scheduled end.

And just last month, another clinical trial contradicted the expert consensus guidelines that patients with kidney failure on dialysis should be given statin drugs to prevent heart attack and stroke.

These and other recent examples show why rigid and punitive rules to broadly standardize care for all patients often break down. Human beings are not uniform in their biology. A disease with many effects on multiple organs, like diabetes, acts differently in different people. Medicine is an imperfect science, and its study is also imperfect. Information evolves and changes. Rather than rigidity, flexibility is appropriate in applying evidence from clinical trials. To that end, a good doctor exercises sound clinical judgment by consulting expert guidelines and assessing ongoing research, but then decides what is quality care for the individual patient. And what is best sometimes deviates from the norms.

Yet too often quality metrics coerce doctors into rigid and ill-advised procedures. Orwell could have written about how the word “quality” became zealously defined by regulators, and then redefined with each change in consensus guidelines. And Kafka could detail the recent experience of a pediatrician featured in Vital Signs, the member publication of the Massachusetts Medical Society. Out of the blue, according to the article, Dr. Ann T. Nutt received a letter in February from the Massachusetts Group Insurance Commission on Clinical Performance Improvement informing her that she was no longer ranked as Tier 1 but had fallen to Tier 3. (Massachusetts and some private insurers use a three-tier ranking system to incentivize high-quality care.) She contacted the regulators and insisted that she be given details to explain her fall in rating.

After much effort, she discovered that in 127 opportunities to comply with quality metrics, she had met the standards 115 times. But the regulators refused to provide the names of patients who allegedly had received low quality care, so she had no way to assess their judgment for herself. The pediatrician fought back and ultimately learned which guidelines she had failed to follow. Despite her cogent rebuttal, the regulator denied the appeal and the doctor is still ranked as Tier 3. She continues to battle the state.

Doubts about the relevance of quality metrics to clinical reality are even emerging from the federal pilot programs launched in 2003. An analysis of Medicare pay-for-performance for hip and knee replacement by orthopedic surgeons at 260 hospitals in 38 states published in the most recent March/April issue of Health Affairs showed that conforming to or deviating from expert quality metrics had no relationship to the actual complications or clinical outcomes of the patients. Similarly, a study led by UCLA researchers of over 5,000 patients at 91 hospitals published in 2007 in the Journal of the American Medical Association found that the application of most federal quality process measures did not change mortality from heart failure.

State pay-for-performance programs also provide disturbing data on the unintended consequences of coercive regulation. Another report in the most recent Health Affairs evaluating some 35,000 physicians caring for 6.2 million patients in California revealed that doctors dropped noncompliant patients, or refused to treat people with complicated illnesses involving many organs, since their outcomes would make their statistics look bad. And research by the Brigham and Women’s Hospital published last month in the Journal of the American College of Cardiology indicates that report cards may be pushing Massachusetts cardiologists to deny lifesaving procedures on very sick heart patients out of fear of receiving a low grade if the outcome is poor.

Dr. David Sackett, a pioneer of “evidence-based medicine,” where results from clinical trials rather than anecdotes are used to guide physician practice, famously said, “Half of what you’ll learn in medical school will be shown to be either dead wrong or out of date within five years of your graduation; the trouble is that nobody can tell you which half — so the most important thing to learn is how to learn on your own.” Science depends upon such a sentiment, and honors the doubter and iconoclast who overturns false paradigms.

Before a surgeon begins an operation, he must stop and call a “time-out” to verify that he has all the correct information and instruments to safely proceed. We need a national time-out in the rush to mandate what policy makers term quality care to prevent doing more harm than good.

Dr. Groopman, a staff writer for the New Yorker, and Dr. Hartzband are on the staff of Beth Israel Deaconess Medical Center in Boston and on the faculty of Harvard Medical School.