A nurse feeds a baby in the neonatal intensive care unit at the Virtua Health facility in Vorhees, New Jersey
A nurse feeds a baby in the neonatal intensive care unit at the Virtua Health facility in Vorhees, New Jersey, November 19, 2009. Virtua Health, a suburban Philadelphia hospital chain, had a problem. Scheduled caesarean sections were running behind schedule -- from one to two hours on a good day to as much as eight to 10 when things really got backed up. Enter General Electric Co. The largest U.S. conglomerate is bringing its corporate arsenal of efficiency tools, like Six Sigma with its "black belt" experts who make factories run more smoothly, to the world of white coats and stethoscopes. They are the ones who helped Virtua clean up its scheduling mess. Picture taken November 19, 2009. Reuters

Solid evidence is often missing from the practice guidelines used by obstetrician-gynecologists across the U.S., a new study shows.

Less than a third of the recommendations from the American College of Obstetricians and Gynecologists (ACOG) are based on gold-standard scientific experiments, researchers found.

The rest are based on anecdotal evidence or expert opinion, which is subject to personal biases, they reported.

That is often the fall-back when there is no data, said Dr. Andrew D. Auerbach, a professor of medicine at the University of California, San Francisco, who was not part of the study.

He said expert opinion is helpful in pointing out what we don't know, but might not always translate into what's best for patients.

Many, many times we have the best intentions and the best wishes and yet things don't work out the way we would have liked, Auerbach told Reuters Health.

Guidelines help doctors keep up with the latest developments in their fields and are widely perceived as a recipe for good patient care.

But there is often surprisingly little hard data behind them. Earlier this year, for instance, a group of researchers found that only one in seven treatment recommendations from the Infectious Diseases Society of America was based on high-quality data from clinical trials.

One such recommendation urged doctors to treat pneumonia with antibiotics right away. It ended up fueling overtreatment with no apparent benefits, possibly breeding resistant bacteria and exposing patients to side effects. (See Reuters Health story of January 10, 2011).

In the new study, Dr. Jason D. Wright of Columbia University in New York and colleagues went through 717 practice recommendations from ACOG, the nation's leading group of ob-gyns.

They found 30 percent of those were based on top-notch evidence, so-called randomized controlled trials. About 38 percent came from observational studies, whose value is limited, and 32 percent were purely expert opinion.

It is disappointing, but I don't think it reflects their process, said Auerbach, who was recently part of an Institute of Medicine committee to develop standards for better guidelines. It is really a call for better evidence.

In an editorial published along with the new study in ACOG's journal Obstetrics and Gynecology, Editor-in-Chief Dr. James R. Scott said the group's guidelines compared favorably with those in other areas of medicine.

He added that the guidelines panels conduct extensive reviews of the medical literature to find all relevant evidence and also take care to exclude experts with financial conflicts of interest.

According to Dr. Sheldon Greenfield, who chaired the Institute of Medicine guideline committee, those are two key elements in creating good guidelines.

A third component is to ensure that guideline panels represent all stakeholders -- including doctors from other specialties and patients, he told Reuters Health.

If those requirements are fulfilled, he added, there is room for opinion.

For many of the recommendations there simply is not enough data, or it is disputed, said Greenfield, a professor of medicine at the University of California, Irvine. So there has to be a role for expert opinion.

So far, however, few if any guideline panels meet all three requirements.

It's about as common as peace in the Middle East, Greenfield said. People are struggling with it.

SOURCE: bit.ly/pkPZu4 Obstetrics and Gynecology, September 2011.