A widely publicized study identifying medical error as the third-leading cause of death in the U.S., behind heart disease and cancer, may have misled readers with statistical guesswork and author bias.
“The science of safety has matured to describe how communication breakdowns, diagnostic errors, poor judgment, and inadequate skill can directly result in patient harm and death,” wrote Johns Hopkins University School of Medicine Professor Martin Makary and Research Fellow Michael Daniel on May 3 in “Medical error—the third leading cause of death in the US,” published in The BMJ (formerly British Medical Journal).
To create its annual list of the most common causes of death, the Centers for Disease Control (CDC) relies on death certificates, which require physicians to list a cause of death associated with the International Classification Disease (ICD) code, Makary and Daniel write. Because the ICD has no classification for “human and system factors,” causes of death resulting from medical error are not captured, the authors write.
Dr. Gerard Gianoli, president of The Ear and Balance Institute and clinical associate professor at Tulane University School of Medicine, says The BMJ misled readers with the non-scientific, non-representative paper.
“The actual number of patient deaths analyzed, over a period of 10 years was—35,” Gianoli wrote in an op-ed published by the American Association of Physicians and Surgeons (AAPS) on May 20. “Shame on The BMJ for allowing this inflammatory, sensationalist article to be published and reported in the general news media without noting to the public that this was not a scientific study at all, but only an opinion piece. It presented no new research, and it was not peer-reviewed.”
Dr. Jane Orient, executive director of AAPS, says the paper relies on ambiguous data that fails to represent most hospitals in the United States.
“The claim is a statistical guess of what the medical error is and whether it actually caused the death, based on a couple of small studies that don’t represent the entire scope of medical care in U.S. hospitals,” Orient said.
Treating patients with serious illnesses requires physicians to attend to countless data points difficult to quantify in a study, Orient says.
“So much depends on the individual variability of a human being: biochemical, emotional, responses, social circumstances, and other factors,” Orient said. “It is different from flying a plane, which is much more mechanical and where the laws of physics apply.”
Practicing medicine requires doctors to treat patients in worse condition than clinical guidelines assume, Orient says.
“[Clinical] study subjects are fairly healthy and representative of the average patient,” Orient said. “This means the guidelines being applied may actually be way off the mark for your individual patient who has multiple chronic conditions and is being treated for cancer.”
Dr. John Dunn, instructor in emergency medicine at Fort Hood, Texas, and an advisor at the American Council on Science and Health, says the study’s findings could be tainted by biased research and possibly ulterior motives.
“These results are heavily based on a methodology of screening for bad outcomes, which introduces outcome bias to research,” Dunn said. “The outcome bias is also created by the nature of the researchers, who are motivated to find fault that will get them notoriety and give them influence in patient safety programs and consulting for hospitals and medical groups anxious to look cooperative and constructive in addressing patient safety.”
Nothing New Here
The authors rehashed previously published data without adding a meaningful contribution of their own, Dunn says.
“The authors admit they accepted research previously published for their projection of negligent deaths and did not make an independent effort to assess the negligence numbers for reliability,” Dunn said. “They simply applied the patient safety claims of other researchers to the previously reported death statistics and then made projections based on population growth. They don’t offer any original research, assume too much, and provide a superficial treatment to get so much publicity.”
The study’s findings conflict with Dunn’s research published by the Texas Medical Association Journal Texas Medicine in 2000, which analyzed a database of more than 300,000 chart reviews. Dunn found only a 0.15 percent rate of negligence or preventable injury and death among Medicare-aged patients, who typically are sicker and more exposed to riskier medical care than the overall population. A 1984 Harvard review of 4,980 charts in New York had shown a rate of negligence or preventable injuries of 1.9 percent. Harvard found that rate to be 6.7 percent in a 1992 review of 3,966 charts in Utah and Colorado.
‘We Need Serious Research’
Shoddy studies motivated by a desire to increase government control over people’s health care won’t help providers give better care, Dunn says.
“We need serious research, not shaky claims to get publicity by patient safety ‘experts,’ so that patient safety projects and solutions can be based on reliable science,” Dunn said. “At the present time, the patient safety crusade has the taint of research energized and motivated by a political agenda to denigrate hospitals, physicians, and nurses in order to increase government ‘management’ to prevent the negligence claimed.”
Orient says the best way to reduce medical errors is to unburden practitioners from regulatory demands that distract providers from treating patients.
“Clinicians need to exercise discretion and focus on what is really important for each patient, instead of managing the growth in bureaucracy,” Orient said. “Doctors and nurses are getting so burned out and are under threat of damage to careers if they slip up on measures, and they can’t wait to retire. Errors occur, but efforts to eliminate them entirely are counterproductive. Instead, let’s focus on empowering clinicians to act when they see something wrong.”
Christina Thielst ([email protected]) writes from Santa Barbara, California.
John Dale Dunn, M.D., J.D., “The Myths of Managing Health Care,” American Thinker, April 11, 2010: https://www.heartland.org/policy-documents/myths-managing-health-care
John Dale Dunn, M.D., J.D., “Patient Safety Research: Creating Crisis,” American Council on Science and Health, January 10, 2005: https://www.heartland.org/policy-documents/patient-safety-research-creating-crisis
John Dale Dunn, M.D., J.D., “Patient Safety in America: Comparison and Analysis of National and Texas Patient Safety Research,” Texas Medicine, Texas Medical Association, October 2000: https://www.heartland.org/policy-documents/patient-safety-america-comparison-and-analysis-national-and-texas-patient-safety-re
Image via Thinkstock