Friday, January 24, 2014

Just how many hospital deaths through medical errors?

A few days ago, the German association of public health insurers lobbed a startling figure at the media: 19 000 avoidable deaths caused by medical errors in Germany in 2012.

Here at the K-Landnews, we blinked as we saw the headlines.

We did something uncharacteristic in our 24 minute news cycle: wait.

Numbers don't lie, but the way you get to a number may well be less rigorous than they teach in school.

The wait paid off.

The German hospital association counter attacked. Incidentally, we hope nobody got hurt in the attack and needed hospitalization because....

The hospitals came up with this number: 122 deaths through medical error in German hospitals in 2012.

Now what?

Both numbers are estimates. The health insurers use this formula: total number of hospitalizations, medical error rate of 1% of that total, then apply a 0.1% death rate on the number of errors. 19 million hospital stays results in the 19 000.

The hospital association gets its number like this: take the number of medical errors certified as errors by multiple experts, take an estimated 70% as in-hospital errors, apply an estimated fatality rate of 3.6%, and voila you are down to 122. They graciously concede unreported cases without giving an estimate for them.

Conclusion:
It is pretty damn obvious that the hospital association number is bunk. Very few miracles happen in Germany or elsewhere. 122 avoidable deaths on 19 million hospitalizations is a miracle.

The health insurers number, even if it were double the real number, is certainly better and would be more realistic if compared to figures from other countries.

Because, well, governments and associations in other countries do statistics, too. Methodologies vary, of course.
 
Do fewer people die due to medical errors in hospitals than through aspirin overdose?


No comments:

Post a Comment