Some folks on twitter asked my thoughts on AMIKINHAL trial, so I thought I'd jot them here.
AMIKINHAL is a multi-center RCT that evaluated the ability of inhaled amikacin to prevent VAP (ventilator-associated pneumonia) among patients who had been intubated for three days. It's available here at NEJM.
It was a positive trial, which is pretty uncommon among critical care trials. Specifically, inhaled amikacin reduced the risk of VAP within a month from 22% to 15% (p = 0.004).
Nonetheless, I don't think that this trial should lead to adoption of prophylactic amikacin, for four reasons:
(#1) The primary outcome of VAP is defined partially on the basis of positive sputum culture results. It's conceivable that nebulized amikacin could sterilize upper airway secretions, thereby directly reducing the rate of VAP diagnosis (rather than truly affecting the number of actual clinical pneumonias). VAP diagnosis is extremely challenging, with problems regarding both over-diagnosis and under-diagnosis. For example, it's possible that many patients in the control group were over-diagnosed with VAP (e.g., due to having atelectasis plus positive sputum cultures). In that scenario, amikacin might appear effective merely by reducing the rate of VAP over-diagnosis (i.e., amikacin is treating a phantom infection that doesn't exist).
(#2) The length of mechanical ventilation was exactly the same between both groups:
- Amikacin group: median 9 days, interquartile range 6-16 days.
- Control group: median of 9 days, interquartile range 6-15 days.
If inhaled amikacin was truly affecting the rate of VAP, there should be some difference in the duration of mechanical ventilation.
Ultimately, duration of mechanical ventilation is the most proximate patient-centered endpoint to this intervention. There was no hint of improvement.
(#3) The rate of VAP seemed very high! Among the control group, 22% developed VAP. That just seems really high to me (noting, of course, that VAP rates will vary between different ICUs depending on illness severity and other factors).
Let's suppose for a minute that we believe that amikacin truly reduces the rate of VAP by 32% (as is purported to be shown in this study). If your ICU has a baseline VAP rate of 22%, then you have a NNT (number needed to treat) of 14 patients to prevent one VAP.
But I think that many ICUs probably have a lower VAP rate, especially if the average duration of mechanical ventilation is lower. Let's imagine that your baseline VAP rate is 10%. In that scenario, the NNT to prevent one VAP would be 30. The intervention suddenly seems a lot less effective.
(#4) Broad utilization of antibiotics in ICU will eventually lead to problems with drug-resistance. It's often difficult to detect such issues over the time-frame of a single clinical trial (because drug resistance may accumulate over a slower time-frame). So a single trial may look terrific, but the reality of implementing a therapy over years and years may be much less rosy. This issue is discussed in more detail in a prior blog here.
That's all for now. Sorry to be a downer.
Image credit: Photo by Jeremy Bishop on Unsplash
- PulmCrit: ADAPT and SCREEN trials are full of sound and fury, signifying little - December 13, 2024
- PulmCrit: How to quickly create a useful professional account in BlueSky - November 28, 2024
- PulmCrit Wee: Why MedTwitter should move to Bluesky - November 15, 2024
Very thoughtful review.
I want just to emphasize the #1 and #3 by saying the VAP definiton didn’t only includes “confirmed VAP” but also “possible or definitive VAP”.
If VAP is already a difficult diagnosis, a possible VAP is an even more difficult, subjective and assessor-dependent outcome.
If we start giving meropenem to every admitted patient many of the short-term colonizations would probably be reduced.
Cheers
Bernardo
Thank you first for the nod to antibiotic stewardship. It warms this ID doc’s heart. Your first critique makes me think about the phrase “therapeutic validation” recently mentioned in Jeremy Faust’s InsideMedicine blog. He suggested that the existence of long COVID is validated by a clinical trial suggesting that metformin could reduce its incidence. Here, a significant reduction in VAP rates without an impact on associated outcomes (vent days or other) essentially validates the phenomenon of VAP over-diagnosis in general and may even quantitate more precisely how often it occurs. I don’t see it in the supplement, but wonder if… Read more »
can you elaborate on the first part about therapeutic validation? I dont understand what you mean.