News extra

Study finds US paediatric medical errors kill 4500 children a year

New Jersey Anne Harding

Medical errors in hospital are responsible for the deaths of nearly 4500 children in the United States every year, says a study examining the effect of lapses in patient safety on children in hospitals throughout the nation (Pediatrics 2004;113:1741-6).

"The bottom line is that none of these events should have happened," said Dr Marlene R Miller, the study’s lead author and director of quality and safety initiatives at the Johns Hopkins Children’s Center in Baltimore.

Children less than 1 year old and those covered by Medicaid, the government’s insurance programme for poor people, were most likely to experience medical errors, Dr Miller and Dr Chunliu Zhan report (Pediatrics 2004;113:1741-6).

The study, which was funded by the Agency for Healthcare Research and Quality, looked at 5.7 million hospital discharge records for people under 19 years in 27 states in 2000. On the basis of their findings, the researchers estimate that paediatric patient safety errors cost the United States more than $1bn (£551m, €828m) annually.

The researchers used patient safety indicators recently created by the agency. These are based on administrative data, and the authors emphasise that they are "indicators," not definitive measures. Medication errors were not included.

While several experts in paediatric patient safety have praised Drs Miller and Zhan for taking a national look at this issue, one specialist said the study grossly overstated the impact of medical errors in children’s care because the numbers included deaths that could not unequivocally be attributed to mistakes.

Dr Erin Stucky, physician advisor for quality management at the Children’s Hospital and Health Center, San Diego, and director of inpatient teaching at the University of California San Diego’s Department of  Pediatrics, said that of the 20 patient safety indicators analysed, only a few—such as "foreign body left after procedure"—could without exception be attributed to error.

"If someone dies in a paediatric intensive care unit from shock and sepsis, you cannot state that that’s a patient safety issue" without additional information, Stucky added. Several other categories, including "postoperative physiologic/metabolic derangement" and "failure to rescue," were also questionable, she said.

It was impossible to say whether the findings overestimated or underestimated the rate of paediatric patient safety events, said Dr Rainu Kaushal, a specialist in internal medicine and paediatrics at Brigham and Women’s Hospital in Boston who studies patient safety and information technology.

Dr Kaushal thought it would be useful to implement known principles of patient safety to improve children’s care. Giving parents a medical passport with complete information on their child’s medications and medical problems would be a good first step, she said, as would more widespread use of computerised physician decision support systems, which she estimated were currently implemented in only 10% to 15% of US hospitals.

"Information technology is really lagging in our health care in the US in general, but particularly in paediatrics," Kaushal said.

While Stuckey, Kaushal, and the authors agree that the new study did not include enough data to offer targeted solutions to paediatric patient safety problems, Stucky calls for pulling data from patients’ charts to get a clearer picture of problems within paediatric health care. "Electronic systems help, but education and people are part of it," she said.
 

Related Article

NICE and its value judgments: Option of safe "understudy" treatments should be available

Woody Caan
BMJ 2004 329: 740. [Extract] [Full Text] [PDF]

This article has been cited by other articles:

• Lenzer, J. (2005). US teenager's parents sue school over depression screening test. BMJ 331: 714-714 [Full text]  
• Lenzer, J. (2005). Bush unveils mental health action plan. BMJ 331: 592-592 [Full text]  
• Lenzer, J. (2004). Bush's plan to screen for mental health meets opposition in Illinois. BMJ 329: 1065- [Full text]  
• Caan, W. (2004). NICE and its value judgments: Option of safe "understudy" treatments should be available. BMJ 329: 740-740 [Full text]  
• Caan, W. (2004). Bush plans to screen whole US population for mental illness: Is road of initiative paved with good intentions?. BMJ 329: 292-292 [Full text]  
• Broomfield, J. R (2004). Bush plans to screen whole US population for mental illness: An unlikely public healthcare initiative. BMJ 329: 292-292 [Full text]  
• Julien, A. W (2004). Bush plans to screen whole US population for mental illness: Bush's sanity test is revealed. BMJ 329: 292-292 [Full text]  
• (2004). Hit parade. BMJ 329: 298-298 [Full text]