Is this bad medicine??

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

Cadet133

Full Member
7+ Year Member
Joined
Jun 1, 2015
Messages
104
Reaction score
6
I see so many attendings basically see SIRS criteria met on labs without patient having a source of infection and basically get started on antibiotics until infection is ruled out usually with blood cx etx. Is this bad practice.
 
If they have a non-infectious cause of SIRS that has been identified, like pancreatitis, I would not start abx. If they look like they have an infection, and I haven't found the source, I would start abx, and modify based on culture results. If someone has SIRS due to infection, and they are not treated, it can progress to things like septic shock, and even death if treatment is delayed.
 
I see so many attendings basically see SIRS criteria met on labs without patient having a source of infection and basically get started on antibiotics until infection is ruled out usually with blood cx etx. Is this bad practice.

No this is now a controversial CMS guideline and your attendings are probably under pressure from admin to never miss a 'sepsis' antibiotic by 2 hours from presentation since the hospital gets a major financial penalty from CMS if it does this too often.
 
I see so many attendings basically see SIRS criteria met on labs without patient having a source of infection and basically get started on antibiotics until infection is ruled out usually with blood cx etx. Is this bad practice.

It really depends on how unobvious the source of infection and how great your suspicion I suppose.

Patients who are not infected and get an initial round of unnecessary antibiotics don't have big morbidity and mortality, but those patients who are infected and miss getting timely empiric antibiotics can have big morbidity and mortality.
 
It really depends on how unobvious the source of infection and how great your suspicion I suppose.

Patients who are not infected and get an initial round of unnecessary antibiotics don't have big morbidity and mortality, but those patients who are infected and miss getting timely empiric antibiotics can have big morbidity and mortality.

Another reason why algorithm medicine is crap.
 
It is terrible. Someone already hinted at one of the reasons (time "goals"). If you think about it, a large portion (if not the majority) of patients in the ED should meet SIRS. HR over 90? that's my resting HR! put in a bit of anxiety due to being nervous for being in the doctor and there goes a RR of 20. Virtually any asthma/COP should meet this. Throw in the perennial fear of doctors of being sued for missing something and you have a recipe for disaster.
It is not uncommon for me to get a call for an admission, I chart review the patient for only 3-5 mins and see that patient already received ceftriaxone in the ED for "code sepsis", 10mins later, they received cefepime because patient came from a nursing home with UTI-sepsis and they were too quick previously with ceftriaxone and now they want to cover healthcare-associated organism. I review previous admission and patient had grown ESBL in the past, and I start meropenem... 3 antibiotics given within the first hour....
Honestly, I don't blame them either, I know my IM service does plenty of things that in turn annoy other specialties/consultants/etc. (Length of stay for instance).
I face this situation much much less often than the ED physicians at my institution, but when I do face it what I do is document very clearly that I do not think there is an infection and period.
If the picture is not clear and/or the patient is unstable enough that me missing an infection could be deadly for them, I wouldn't take the risk, I can always discontinue later. But I have saved a few dozen of my CHF patients with "bilateral cellulitis" that is clearly stasis dermatitis from antibiotics in this manner.
 
It’s a judgment call. If someone is sick enough to be in the hospital -and- I think it might be because of infection, yes I would rather get the cultures and treat empirically until stabilized or culture neg based on most likely source. The downsides of not treating an infection that is making someone qSOFA+ type sick outweigh the risks of overtreatment for a dose or two in my judgment, a lot of the time. If I’ve got a solid other diagnosis to explain the illness and the findings, and no real suspicion of infection, probably no antibiotics.

It’s really bad practice to let people get sicker and die from untreated infections under your care in the hospital, so.
 
I've seen this as well and really it comes down tot he clinical picture. If you don't have a source of infection yet but you have a high clinical suscpion start them on abx and re-evaluate in 24-48 hours.
 
On the ID service, I didn't see alot of FUO but it happened a few times. Each time the hospitalist would consider potential areas of sources. It usually ended up being malignancy. Either way, by the time ID was consulted they were on vancomycin and zosyn and they gave ID the reigns to stop or start abx.
 
It’s a judgment call. If someone is sick enough to be in the hospital -and- I think it might be because of infection, yes I would rather get the cultures and treat empirically until stabilized or culture neg based on most likely source. The downsides of not treating an infection that is making someone qSOFA+ type sick outweigh the risks of overtreatment for a dose or two in my judgment, a lot of the time. If I’ve got a solid other diagnosis to explain the illness and the findings, and no real suspicion of infection, probably no antibiotics.

It’s really bad practice to let people get sicker and die from untreated infections under your care in the hospital, so.


Good medicine

On the ID service, I didn't see alot of FUO but it happened a few times. Each time the hospitalist would consider potential areas of sources. It usually ended up being malignancy. Either way, by the time ID was consulted they were on vancomycin and zosyn and they gave ID the reigns to stop or start abx.

Bad hospitalist
 
It’s a judgment call. If someone is sick enough to be in the hospital -and- I think it might be because of infection, yes I would rather get the cultures and treat empirically until stabilized or culture neg based on most likely source. The downsides of not treating an infection that is making someone qSOFA+ type sick outweigh the risks of overtreatment for a dose or two in my judgment, a lot of the time. If I’ve got a solid other diagnosis to explain the illness and the findings, and no real suspicion of infection, probably no antibiotics.

It’s really bad practice to let people get sicker and die from untreated infections under your care in the hospital, so.
I'd be cautious with this kind of attitude. This is the sort of thing that leads to "pan-scan" "broadspectrum cover" and all sort of shotgun approach. There are specific cases where I don't have a clue and you are right, I rather don't take the chance if the picture is not clear. But I don't think that's the sentiment of the OP. In my experience, the vast majority of these "SIRS + empiric coverage without clear source" is a whole bunch of nonsense and this is indeed bad practice, in fact, it is malpractice. NOTHING that we do in medicine is without a cost, the levofloxacin for fake pneumonia that ruptures the tendon of a patient could just as easily lead to a sue as a missed septic shock that received no antibiotics. We seem to be biased towards the "forgot to do this" mentality and completely ignore the "did this dangerous med/test for a non-existing illness".
I am honestly tired of seeing asymptomatic patients end up getting cardiac cath! after an improperly ordered stress test by cardiology, after an improper consult for "pre-op clearance" by a surgeon for a low-risk procedure.
 
Believe me friend, I’m nothing if not cautious with regard to overtreatment and undertreatment, and gauging some balance between the two and taking responsibility for that decision is what they pay me for. I’m honestly not sure what scenarios OP is seeing...cases with a clear and defensible non-infectious diagnosis, with no real clinical suspicion of any infectious process? OP? If you’re not clear on why the plan is what it is it’s always reasonable to talk it over with your attending.

It IS bad practice imo if people aren’t aware of Sepsis-3 and are still treating/teaching SIRS + infection=sepsis like it’s 2014. If they have thought it over and have a reasonable critique of qSOFA and a judgment-based approach that works for them, that’s not what I’m talking about. As a screening measure SIRS is more sensitive, but qSOFA far more specific for who might die if you don’t treat their infection right, even if you’re not 100% sure what the infection is yet. I take account of both and make a decision. Either set of criteria is dependent on a clinical suspicion of infection. The mortality impact of time to antibiotics when it is sepsis is pretty compelling. Even if there was no such thing as a lawsuit, having someone die a preventable death under your care is horrible for the family and you in a way that a tendon rupture is not.
 
siliso you make a good point but I think we have a miscommunication issue. Let me be clear, if a patient is gravely ill of a likely infectious pathogen there is no discussion, antibiotics, antibiotics, antibiotics + source control if appropriate. The picture becomes clear after day 2 and we can narrow down a definitive source or rule it out completely, then we de-escalate as appropriate. But this is not the scenario that I have been picturing based on the OP. In the institutions that I have worked/visited, it has become common practice to administer "sepsis bundle", ask questions later. Seeing "2.3L bolus" followed by "60ml furosemide IV" in the chart within 1h of each other is not uncommon where I am working. Or the patient that comes into the ED with an asthma exacerbation, they are tachycardic, tachypneic and they do UA and voila, "trace Leucocyte esterase" code sepsis, ceftriaxone 1g in the ED... Did they ask the patient if they had symptoms? hell no. I have been in a half dozen codes this year in which a patient has been spiking fevers for 3 days, nurses trigger code sepsis and then they are rushing me to order ABx and fluids within 1h to fulfill their quality measures. This is not putting the patient first, this is simply playing the numbers. It leads to sloppy work, bizarre antibiotic selection, overtreatment (and not necessarily high-quality treatment), etc.
In short. Critically ill patient with a reasonable chance of having an infection and poor reserves to survive an infection if not treated? Go ahead, blast them with meropenem and vancomycin, etc. Completely stable patient with a very unlikely infection that "technically" fulfills criteria but your analysis does not agree? I think it is reasonable to step back and analyze it, rather than jumping the gun to give ABx, fluids, imagining, etc within a certain amount of time.
One day, not too far away, computers/robots/AI will take over our job, that day will come much much much sooner if we stop thinking.
 
Top