Medicine mix-ups, accidental overdoses and bad drug reactions harm roughly one out of 15 hospitalized children, according to the first scientific test of a new detection method.

That number is far higher than earlier estimates and bolsters concerns already heightened by well-publicized cases such as the accidental drug overdose of actor Dennis Quaid's newborn twins last November.

"These data and the Dennis Quaid episode are telling us that these kinds of errors and experiencing harm as a result of your health care is much more common than people believe.

It's very concerning," said Dr. Charles Homer of the National Initiative for Children's Healthcare Quality. His group helped develop the detection tool used in the study. Researchers found a rate of 11 drug-related harmful events for every 100 hospitalized children.

That compares with an earlier estimate of two per 100 hospitalized children, based on traditional detection methods. The rate reflects the fact that some children experienced more than one drug-treatment mistake.

The new estimate translates to 7.3 percent of hospitalized children, or about 540,000 kids each year, a calculation based on government data. Linda Stoverock, chief nursing officer at Nationwide Children's Hospital, said the hospital doesn't calculate medication errors the way the study did.

At Children's, there are 1.8 errors per 1,000 doses of medication, she said. Less than 0.5 percent are considered severe, she said. Children's, and hospitals across the country, are changing the way they operate to flag mistakes and find ways to prevent them, Stoverock said. In Columbus, computers alert nurses and doctors to possible allergic reactions and overdoses, among other things.

And the hospital tracks the use of drugs commonly used as antidotes to medication errors, such as morphine overdose, she said.

"We are trying to put in every tool we can to get the appropriate medication to the patient," she said, adding that pediatric hospitals are working together on many initiatives to improve care.

"I think the more we pull together and learn from each other that it improves health care of children across the nation." Simply relying on hospital workers to report such problems had found less than 4 percent of the problems detected in the new study.

The new monitoring method developed for the study is a list of 15 "triggers" on young patients' charts that suggest possible drug-related harm. It includes use of specific antidotes for drug overdoses, suspicious side effects and certain lab tests.

By contrast, traditional methods include nonspecific patient-chart reviews and voluntary error reporting.

The researchers said their findings highlight the need for "aggressive, evidence-based prevention strategies to decrease the substantial risk for medication-related harm to our pediatric inpatient population."

The study was released in the April issue of the journal Pediatrics. It involved a review of randomly selected medical charts for 960 children treated at 12 freestanding children's hospitals nationwide in 2002.

Triggers mentioned in the charts promoted an in-depth review of the patients' care. Patient-safety experts said the problem is likely even bigger than the study suggests because it involved only a review of selected charts.

Study author Dr. Paul Sharek said evidence is needed to show whether a big push to prevent medical errors in recent years has put a dent in the problem since 2002, when the data were gathered.

Homer, of the children's health-care initiative, said some hospitals have started using trigger methods similar to those in the study. But, he added, "We still have a long way to go." Quaid's twins got accidental life-threatening heparin overdoses in a Los Angeles hospital shortly after they were born last November.

The babies recovered, and the actor said "they appear to be normal kids, very happy and healthy."