A recent story in the AMA's American Medical News (amednews.com) was headlined "Hospitals stumble in preventing 'never events.'" It reported that the Minnesota Department of Health said, "the number of patient disabilities attributed to the mistakes rose from 84 to 89, while related patient deaths jumped from five in 2011 to 14 in 2012."
Here’s a chart from that Minnesota DOH report. Deaths are in red and disabilities (harms) are in blue.
What do you think? Have the hospitals really stumbled?
Let’s take a closer look.
The report consists of all of the patient harms and deaths voluntarily reported to the state by hospitals for the years in question.
It is well-known that voluntary reporting captures less than 10% of all adverse patient events.
Minnesota has about 150 hospitals. Here are some numbers for the last reporting year (10/11-10/12).
Number of wrong patients operated on? None Number of intra-operative/postoperative deaths? None Number of misuse or malfunction of devices? None Number of contaminated drugs, devices or biologics? None Death or disability due to medication error? 2 (both disability)
Do you believe those numbers? I don’t.
And of course, the good news, if it’s true, was not mentioned in the article
The number of pressure ulcers declined by 8%. This is the first decline of this magnitude in the nine years of reporting. This year’s total of 130 is down from an all-time high of 141 last year. If you’re a skeptic, you might just question that figure as it means that the incidence of pressure ulcers averages less than 1 per hospital per year.
Retained foreign objects declined by 16 percent, the first drop in this category in five years.
Medication errors dropped by 75% from the previous year and were at the lowest level in all nine years of reporting.
Now here's the real problem with the report and the amednews.com story.
Here are charts of the deaths and harms with trend lines clearly showing that if anything, both harms and deaths are decreasing, albeit not significantly. They certainly aren’t getting worse.
As I have written many times, a bit of knowledge about statistics can be quite useful if you are doing research or reporting on it.
So yes, "hospitals stumble" but maybe not in preventing "never events." The stumble is in the reporting of "never events" by hospitals. I think amednews.com stumbled a bit there too.