Among the many overused tropes in medicine is the myth of the supremacy of intravenous antibiotics. In the appropriate clinical context, it’s just a waste.
This is a retrospective analysis of 36,405 patients hospitalized for community-acquired pneumonia, and for whom a fluoroquinolone was selected as therapy. The vast majority – 94% – received an intravenous dose, while the remaining 2,205 (6%) were treated orally. Unadjusted mortality favored the oral dose – unsurprisingly, as those patients also generally has fewer comorbid conditions. In their multivariate, propensity-matched analysis, there was no difference in mortality, intensive care unit escalation, or mechanical ventilation.
These results are wholly unsurprising, and the key feature is the class of antibiotic involved. Commonly used antibiotics in the fluoroquinolone class, trimethoprim-sulfamethoxazole, metronidazole, and clindamycin, among others, have excellent oral absorption. I have seen many a referral to the Emergency Department for “intravenous antibiotics” prior to an anticipated discharge to home therapy when any one of these choices could have obviated the entire encounter.
“Association Between Initial Route of Fluoroquinolone Administration and Outcomes in Patients Hospitalized for Community-acquired Pneumonia”
http://www.ncbi.nlm.nih.gov/pubmed/27048748
I think "iv antibiotics" ensures that 1. you demonstrate to the patient that you care enough to "do something", and 2. the pt gets watched longer, i.e. more data to estimate pace of the disease and 3. serves as a (veiled) second opinion request when initial doc is uncertain about dx and severity. It would be great if obs admissions would simply state as much, rather than pretending that abx are better IV than PO.
The mental process behind using IV antibiotics is based on this false syllogism:
I prescribe oral antibiotics for every cough, cold, sniffle, or rash in non-sick patients.
None of my patients get better any quicker.
Therefore oral antibiotics don't work.
and subsequently:
Sick patients need IV antibiotics.
IV is caring!
A lot of my patients would rather go home than sit around for another 2 hours in the ED for their IV abx.
A little hard to say what sort of response to therapy you would see – if you're resuscitating, say, pyelonephritis before discharge, presumably that's the fluids ….
Ryan… once again an excellent and informative post!
Keep up the good work.
Excellent review. One potential confinuder (which may have been addressed in the paper proper; haven't had a chance to read it yet) is Case Management.
I've gotten pushback on admissions based on comorbidities where the case managers say my patient doesn't meet Interqual (or similar) criteria… thus they get IV despite their fully functional GI tract in order to "buy the bed" so to speak.
Sometimes the medicine isn't about the medicine.
-tim
While I completely agree IVABs are seriously overused, I find these propensity matched studies fairly uninformative. If you take a seriously flawed data set (e.g. due to selection bias etc) and attempt to statistically adjust for only known confounders it is completely unsurprising that the effect size reverts to the null hypothesis. IMHO unsurprising and pretty meaningless. I think this places too much belief in the value of statistical adjustments and not enough on the value of quality data sets and research methodology. We've seen similar pretty uninformative attempts at disproving the nephrotoxic effects of contrast in the April 2013 Radiology journal.