(Un)acceptable quality levels in health care

In the bad old days of American manufacturing aka the 1970s, terms like  "acceptable quality level" were widely used. The idea was that defects and errors were inevitable and that customers expected them. Only later did concepts such as six sigma emerge, reflecting the radical idea that defects can be extremely rare, i.e., a few per million.Unfortunately health care delivery is still in the defects per thousand or even defects per hundred mode. Some errors are trivial and almost laughable. Others are deadly and inexcusable. Here are three examples I heard about within the last 24 hours, in escalating order of concern.Last night a friend told me the story of his nine year old, who hurt her pinkie. He took his daughter to the emergency room of a local, highly-regarded hospital, where she was x-rayed and told there was a hairline fracture at the tip. Discharge instructions were unclear, but a few days later he brought her to a hand surgeon at the same hospital. Despite the hospital's multi-million dollar EHR and PACS system the hand surgeon could not retrieve the original x-ray. (The dad is an MIT grad and told me he had actually made a point of taking his daughter to the same hospital because he assumed the doc would have access to the image.) A new x-ray was ordered, but was inconclusive. Through palpation, the surgeon concluded there was a fracture on the growth plate. In any case the treatment recommendation was the same. But wouldn't it have been nice for the hand surgeon to have seen the original x-ray and for the patient to have avoided an extra zap of radiation, however small? Of course there's the duplication of expense, too...Overhearing this story, an acquaintance told us about his recent experience. He had what he described as a bad stomach ache that didn't go away over a period of days. He called his internist at a world-renowned hospital here in Boston, and left several voicemails over the course of almost a week. At one point he managed to speak to a physician on call, who told him to go to the emergency room. He went and found out he had appendicitis. After he recovered and visited his internist, the internist asked, "Why didn't you call me?" The patient was flabbergasted, and told the doc he could tell him exactly what his outgoing voicemail message says since he'd heard it so many times.Finally, on MedPage Today I read High Error Rates Found with Outpatient Cancer Therapy, which really concerns me:

Drug-related errors occurred in 7.1% of adult visits (95% CI 5.7% to 8.6%) and 18.8% of pediatric visits (95% CI 12.5% to 26.9%), according to a review of nearly 1,400 visits at four geographically diverse outpatient clinics, reported Kathleen E. Walsh, M.D., M.Sc., of the University of Massachusetts, and colleagues.Of the 112 errors identified, 64 had "potential to cause injury," the researchers said online in the Journal of Clinical Oncology. Actual injury occurred in 15 cases, they added...Per 1,000 medication orders, the researchers counted 8.2 errors associated with adult visits (95% CI 6.5 to 9.9) and 24.1 errors for pediatric patients (95% CI 14.2 to 34.1).The higher rate of pediatric errors in the current study was also reflected in potentially injurious errors and for those actually causing harm.Dr. Walsh and colleagues found rates of 9.9 potentially injurious errors per 1,000 medication orders for pediatric patients, versus 5.0 per 1,000 orders in adults.For actual injuries, the rates were 4.3 per 1,000 orders in children and 1.0 per 1,000 orders in adults (P values not reported)...Dr. Walsh and colleagues said their estimates were probably conservative. They said that direct observation of drug administration and interviews with patients and providers would likely have revealed additional errors not picked up by the chart reviews.

A lot of the errors are made by adults self-administering the drugs inappropriately or parents administering drugs inappropriately to their kids. This could be corrected by better communication and training.It's pretty shocking that people struggling with cancer end up as the victim of so many errors in their treatment. McDonald's would be out of business if they screwed up so many orders. So would Fedex. American consumers don't typically accept this level of error in other service industries and shouldn't accept it in health care either.