- Joined
- Nov 19, 2014
- Messages
- 623
- Reaction score
- 804
So AANA has made a press release that "proves" that AANA care is essentially equivalent to anesthesiologist care. First of all, their usage of the word "prove" annoys me because you don't really prove anything in medicine, you can give very convincing evidence but proof is a bit more complicated. Anyway, I digress. An M2 (I know, its pretty incredible how good it is for their level...way way better than I would do) on reddit did a very thorough and thoughtful critique on why its not a great study. I am posting it here so you can read it, share it with a congressional representative, or a colleague or start some sort of weird discussion as ya'll are wont to do. Anyway, here it is an before that a link to the AANA press release. Also if you share it with a congressional representative you may want to edit the naughty language...or not you're adults do whatever.
Their entire study is a misuse of statistics.
Their entire study is a misuse of statistics.
- It seems like their null hypothesis is set up in the experiment to be that there is no difference between doctors and NPPs, which would mean that their alternative hypothesis is that there is a difference. If they did not detect a difference, that doesn't PROVE the null hypothesis - they only fail to reject it.
- They didn't randomize - and there are so many confounding factors (which the ASA mentions in their rebuttal). For instance, there should definitely be a trend towards experts getting more complex cases (which they try to address - but I will get to that in #3). Also, though, there is likely a difference in which NPPs attempt to pursue independent practice. I would wager that there are several CRNAs that are better than some MDs - but I'm sure that the variability of CRNAs is much greater AND the mean expertise is much lower. But, if you take the best of the best and allow them to roam free, it doesn't prove that all NPPs are good - only that they can potentially self-select a bit. Also, it doesn't address practice patterns of anesthesiologists - perhaps there is a pattern of which anesthesiologists practice in which conditions and which cases are displaced by independent NPPs.
- The method they use to standardize the pre-op risk is to link it to the ICD-9 code. I can't believe there wasn't more of a problem with this. Let's say I asked you to perform anesthesia on a patient and agreed to provide you with an ICD-9 code prior to the operation so that could determine the risk of anesthesia. I feel like most people would say to **** off because it's stupid. I am not an anesthesiologist and I can tell you that an ICD-9 code means about as much to me for risk stratification as a cup of warm piss. You better believe that the NPPs take one look at some of these patients, realize that they have fat necks or sleep apnea or a history of malignant hyperthermia or whatever and turf that **** straight to a real doctor. Having the balls to suggest that they accurately modeled the risk of anesthesia on an ICD-9 code...if they could do that, it would be an article worth publishing all on its own.
- Conflict of interest, the elephant in the room. The entire study is obviously funded by the AANA.
- They use QZ billing to determine “independent” CRNA cases when in fact a lot of physician supervised groups use this to maximize 1:4 ratio and ensure full billing.
- They attempted to determine patient complexity by the mean number of anesthesia base units billed. Not a great method. Regardless, using their method, solo anesthesiologists care for patients who were on average more complex than solo CRNAs.