Saturday, August 18, 2007

99% Accurate. Is it Good Enough?

I would like to comment on something I just read on KevinMD. Here is the Title:

Retail clinics: 99.15% proper treatment rate?

And here is Kevin's input: No doctor can match the perfection that retail health clinics offer.

Why don't we examine what 99% means, and then we can decide if Kevin is right in his assessment and more importantly, if the retail clinics' accuracy rate is acceptable.
  • If a surgeon removes 99% of the tumor, the patient dies of cancer
  • If the antibiotic I prescribe is 99% effective against the bacteria, the infection will recur.
  • If my sterilization technique for cystoscopy is 99% effective, I'll infect 2 people per year, and a busy urology group will infect 10-20 per year, potentially, with an infectious disease!
  • If my specimen labeling process is 99% effective, I'll mix-up specimens at least 26 times per year (100 per 2 weeks x 26 bi-weeks per year).
  • If a busy internist is given the correct chart with 99% accuracy, he will make his notation in the wrong chart on 52 patients per year, on average.
  • If a lab filing system is 99% accurate, labs will be misfiled 100 times per year, on average, in a moderately busy doctors' office.
  • If an established medical office has an EMR that uses back-up that is 99% accurate, a crash could cause the permanent loss of 100-200 charts (10,000 to 20,000 patient practices, respectively).
  • If surgeons operated on the correct side/site only 99% of the time, in my little community hospital, we would see 1 wrong side/site surgery per week, and the hospital would be shut down by the state.
  • If you drove with 99% accuracy, you'd have a car accident every 4 to 5 weeks.

You see, 99% sounds great, but in actuality, is only good if you are taking a college final exam. In life, big number enterprises, such as medicine, 99% is not very good at all.

I disagree with Kevin. "Proper treatments" are given by doctors at a far more accurate rate than 99%, or we'd all be out of business. And I'd think twice about going to a retail clinic.


The IU.