By Harlan Krumholz
Friday, July 23, 2010; A21
Negative studies -- the ones that fail to find that one therapy is superior to another -- usually do not elicit much excitement. Scientific studies with breakthrough results establishing the effectiveness of something new tend to get all the attention, while those that fail to find the hoped-for results are often ignored.
But with high costs and patient overtreatment looming large in the health-care debate, negative studies may be getting a closer look.
Many drugs and treatments routinely used by physicians can have specific results, such as opening an artery. But these "improvements" don't always translate into the expected benefit for patients' overall and lasting health. That's why it is important to also conduct trials that focus on strategies that are in use but are not fully tested -- and to confirm that they improve patients' quality of life or survival.
Consider: In the past two years scientists have announced the results of a $300 million trial sponsored by the National Institutes of Health to evaluate common, seemingly sensible but unproven strategies to treat blood sugar, lipids and blood pressure in patients with diabetes. In each case the more intensive approach, which used more medication to improve a patient's risk-factor profile, failed to improve the health of patients. One finding, corroborated by two other trials, confirmed that the strategy of markedly lowering blood sugar was not effective for older patients with long-standing diabetes.
There are many other examples of trials in which anticipated benefits from drugs and procedures did not materialize. A study published last fall tested the value of darbepoetin alfa, a drug that increases the red blood cell count and can treat anemia. This industry-sponsored study showed that the drug had no benefit and increased the risk of stroke. Another recent study failed to show the benefit of a widely used surgical procedure to treat osteoporotic vertebral compression fractures. A study published last November showed that a common procedure to improve blood flow to the kidneys in patients with blockages failed to reveal any benefit and in fact indicated some harms.
These trials raised questions about strategies employed in everyday medical practice. In fact, the blood sugar strategy that failed to improve patient outcomes was endorsed by many organizations as a practice guideline and used by some to judge the quality of physician care. Some employers and payers even paid doctors more for implementing this strategy.
Rather than a letdown, the failure to find an advantage in an expensive strategy opens the door to doing less and spending less without worsening patient care -- and in some cases improving it. It's simply the case that many popular medical strategies have little or no rigorous scientific evidence of their effectiveness regarding patient outcomes.
Realizing the potential for negative studies to change practice remains a challenge. Because negative studies often contradict conventional wisdom, they can undermine corporate-marketing messages and the pet theories of academics. Although no study is perfect, the flaws of the negative studies are often magnified by detractors to deflect the message that we can achieve more by doing less -- or spending less.
It can also be difficult to convince patients that no therapy is actually the best therapy. But strong studies showing that strategies do not deliver on their promise could provide practitioners and patients with the information they need to choose to do less.
So, long-unheralded negative trials may be headed for a renaissance. The price tag for these studies of popular strategies may appear high, but the investment in the knowledge could be far outweighed by the potential savings and reduction in risk of avoiding ineffective treatments. For the right question in the right setting, a negative study may be the most positive result yet for patients and our health-care system.
The writer is a cardiologist and the Harold H. Hines Jr. professor of medicine at Yale University.