Blood Cx when discharging?

[engineeredout]

Advertisement - Members don't see this ad

Sorry to create a thread that doesn't include "IM IN 9TH GRADE AND GOT A 91% IN ENGLISH CAN I MATCH EM?!?!?" or "Hey you emergency doctors are stupid and admitted a not appendicitis" but...

Got into a discussion the other day.

Had a patient, 60 something, lung cancer currently on chemo, last chemo one week prior. Came to ED for DOE somewhat worse than baseline, felt nebs at home weren't making him feel better. Patient afebrile, normal vitals, baseline SpO2. Patient really doesn't look bad, no respiratory distress, hes awake, alert, joking with us.

Checked labs, only thing abnormal was CBC: WBC 6, 28% bands. CXR unremarkable. Gave hour long duoneb, steroids, Patient feels much better. Ambulating in ED, feels back to baseline.Called patient's oncologist, who says the patient is on one of the GM-CSF medications and that is why he would be expected to have such a significant bandemia. He is okay with the patient going home if he is feeling better and seeing him in the next few days. Patient wants to go home too. He is (understandably) sick of the hospital.

My attending is okay with dc as well, but she orders a set of blood cultures to be drawn prior to DC.


So what are your thoughts on this for these borderline patients? Patients look good, think they're probably going to be okay at home, but significant comorbidities. Little extra insurance to make sure that if something bad is developing maybe we can get them back sooner?

Discussed this thought process with another attending. He was against the practice. His reasoning was from a medicolegal standpoint - that if we are sending BCs prior to d/c it looks like we are saying that we are still reasonably concerned about this patient going home and that if we have that concern we should be hospitalizing them for IV abx and waiting for the cultures to come back.

The downsides I see to this are: false positives - how often do you get 1 out of 4 bottles with G+ cocci that ends up being coag negative staph but now you're stuck with it. Also, inability to actually reach the patient should they come back positive. I am also aware of the trend away from routine blood cultures on every patient with an infection as they end up being low yield a majority of the time, costly, have a lot of false positives, and don't actually impact care or outcome.

Upsides: Extra insurance/protection against some hidden badness in a baseline sick patient who otherwise looks well.

Thoughts? Anyone do this?

 

[ccfccp]

It would depend somewhat on the situation for me. If I knew and had spoken with the patient's oncologist (as it sounds like this was done) and was reasonably sure it would be followed up on, I might do it. The downside of the false positive to me is better than the downside of missing significant bacteremia and sepsis in a patient who might not have a typical presentation being one week out from chemo. I would basically always offer this patient admission for cultures and observation, but if patient and patient's doctor want them to go home, I would just document I discussed the risks with them (death or permanent disability from overwhelming bacterial infection) and and that they voiced back return instructions (with a plus if they have a family member in the room). I would document stable vitals and a repeat exam and wish them well. We have a "follow-up nurse" in our department whose whole job is to go through all these results and notify the patient if anything comes back positive, so that makes it easier at our shop.

 

[Doctor Bob]

Yea, I don't like this practice.

My question is why are you drawing the culture? Do you think they have occult bacteremia? If so, it's ok to draw the culture and discharge them, but be consistent in your approach. If you were worried enough about occult bacteremia to draw the culture, then you should start them on empiric abx. This can be PO but you should have them start it and continue it until the results come back. If you aren't worried about occult bacteremia... then don't draw the culture.
Let's replace the test. Let's say you have a patient with a benign abdomen, who was going to get follow up with their PCP in a few days. Would you get an abdominal CT on them, then send them out the door without getting the results (or even looking at the images yourself)? Relying on the callback system to alert them if there's something bad on the film?

It's a bad idea from an MDM standpoint.

 

[gman33]

In the environments I have worked, this is a bad idea.
Poor follow up. Absent/incorrect contact info.

Are you going to personally follow up on the results?
If not, this is a time bomb.
Maybe not so bad if you start a course of abx.

If I have enough concern for bld cx, they are getting admitted.
If not, I am not send blood cultures.

 

[RustedFox]

As per my department chair: "Dude, if they're sick enough for blood cultures - they're sick enough to be admitted."

 

[barnacle]

Cultures + d/c is common practice in the peds world for well-appearing infants w/ fever between 29-90 days old. If everything else is normal, shot of Rocephin & d/c home. (after pcp notified to f/u on cultures & parents deemed reliable to follow up w/ pcp)

You could make analogy that chemo pt is like newborn w/ no immune system. I agree if you are concerned about occult bacteremia/sepsis you should probably cover w/ abx until cultures result. The oncologist needs to be on board with this plan. They usually know their patients well & know which ones are reliable.

I also agree, general rule of thumb, if sick enough for cultures, you get admitted...

 

[LaBusqueda]

a rare benefit of EPIC is that if their doc is a system doc, we can order any test and have it routed to them and I can shoot them an epic message about their visit and the pending whatever.
Can't remember doing this for anything other than blood and/or urine cx's on very young infants. But the ability is there.
I know partners have routed urine cx on older pts.

 

[WilcoWorld]

28% bands.

Somebody call the New York Times.

 

[WilcoWorld]

Sorry to create a thread that doesn't include "IM IN 9TH GRADE AND GOT A 91% IN ENGLISH CAN I MATCH EM?!?!?" or "Hey you emergency doctors are stupid and admitted a not appendicitis" but...

Got into a discussion the other day.

Had a patient, 60 something, lung cancer currently on chemo, last chemo one week prior. Came to ED for DOE somewhat worse than baseline, felt nebs at home weren't making him feel better. Patient afebrile, normal vitals, baseline SpO2. Patient really doesn't look bad, no respiratory distress, hes awake, alert, joking with us.

Checked labs, only thing abnormal was CBC: WBC 6, 28% bands. CXR unremarkable. Gave hour long duoneb, steroids, Patient feels much better. Ambulating in ED, feels back to baseline.Called patient's oncologist, who says the patient is on one of the GM-CSF medications and that is why he would be expected to have such a significant bandemia. He is okay with the patient going home if he is feeling better and seeing him in the next few days. Patient wants to go home too. He is (understandably) sick of the hospital.

My attending is okay with dc as well, but she orders a set of blood cultures to be drawn prior to DC.


So what are your thoughts on this for these borderline patients? Patients look good, think they're probably going to be okay at home, but significant comorbidities. Little extra insurance to make sure that if something bad is developing maybe we can get them back sooner?

Discussed this thought process with another attending. He was against the practice. His reasoning was from a medicolegal standpoint - that if we are sending BCs prior to d/c it looks like we are saying that we are still reasonably concerned about this patient going home and that if we have that concern we should be hospitalizing them for IV abx and waiting for the cultures to come back.

The downsides I see to this are: false positives - how often do you get 1 out of 4 bottles with G+ cocci that ends up being coag negative staph but now you're stuck with it. Also, inability to actually reach the patient should they come back positive. I am also aware of the trend away from routine blood cultures on every patient with an infection as they end up being low yield a majority of the time, costly, have a lot of false positives, and don't actually impact care or outcome.

Upsides: Extra insurance/protection against some hidden badness in a baseline sick patient who otherwise looks well.

Thoughts? Anyone do this?

I do this sort of thing regularly. However, I see 3-5 chemo/transplant patients per shift and we have a good follow up system in place. Also, if I'm sending a chemo/transplant patient home, I'm doing so in consultation with their hematologist/transplant doc and with the patient's ascent. Your criticisms of blood cultures are worth considering, but they apply to hospitalized patients as well.

As for the medico legal risk, I don't think you have to worry about it so much in this situation (you have a patient who feels better and wants to be discharged as well as a specialist who is comfortable with discharge). Yes, your disposition should be commensurate with your ED care, but in a patient with stable vitals, good social support and close follow up, discharge to await cultures is probably safer than hospitalization + broad spectrum antibiotics.

In the right setting, I think blood cultures prior to discharge makes plenty of sense. But when I worked in a community ED where I saw a transplant patient once in a blue moon I did not do this sort of thing.

In conclusion, it depends.

 

[Janders]

Good discussion here.

I was initially taught this was a "never do it" move.

But I think as people note above, there are a special sliver or two of patients (well appearing infants, chemo/Xplant) who have potential bacteremia despite looking very well. IF YOU HAVE VERY TIGHT FOLLOWUP SYSTEMS I think it is reasonable to send them home after drawing cultures IF YOU HAVE A DOCUMENTED RISK:BENEFIT discussion to beef up your chart.

I will say, I almost never see +blood cx from pneumonia unless they are actively dying. Contrast this with UTI/Pyelo, where every once in a while we call back a young person who was feeling better @ home due to +bld cx drawn in the setting of their febrile UTI.

 

[WilcoWorld]

I will say, I almost never see +blood cx from pneumonia unless they are actively dying. Contrast this with UTI/Pyelo, where every once in a while we call back a young person who was feeling better @ home due to +bld cx drawn in the setting of their febrile UTI.

That's my experience too. Those who are called back for + blood Cx seem to almost always be gram negatives in a pyelo patient OR contaminants.

 

[coachB]

Good discussion here.

I was initially taught this was a "never do it" move.

But I think as people note above, there are a special sliver or two of patients (well appearing infants, chemo/Xplant) who have potential bacteremia despite looking very well. IF YOU HAVE VERY TIGHT FOLLOWUP SYSTEMS I think it is reasonable to send them home after drawing cultures IF YOU HAVE A DOCUMENTED RISK:BENEFIT discussion to beef up your chart.

I will say, I almost never see +blood cx from pneumonia unless they are actively dying. Contrast this with UTI/Pyelo, where every once in a while we call back a young person who was feeling better @ home due to +bld cx drawn in the setting of their febrile UTI.

1. As others have said, if you have a patient with an infection (presumably you can diagnose this by symptoms and UA), and you are concerned enough about bacteremia to order blood cultures, why on earth would you send them home?

2. If you have a source (urine), why on earth would you order blood cultures?

 

[SoCuteMD]

I hate sending cultures (ANY CULTURES) on a patient I'm discharging. We have no follow up nurse/PA/doc. Positive blood cultures actually get called to whoever is working, and that is a nightmare to deal with in the middle of a shift as many of the phone numbers for our patients don't work. I sent a urine culture on Saturday and I've had to check it every day since. No big deal on Sunday (working). Yesterday I checked and there were no sensitivities back, then today sensitivities showed up. I then had to call the patient's urologist (whom I had tried to reach on Saturday, but, you know, Saturday...so my pages weren't returned) and discuss with the PA, who said he was going to discuss with someone. I asked that he touch base with me afterwards so that I would know it was all covered. He did and everything's good, but it just took up more head space than I would like on my "day off."

 

[pulmoblast]

Either get culture and give empiric antibiotics and follow up results within a reasonable time

OR

Don't culture and don't do antibiotics

Not giving antibiotics and culturing is ok for inpatients where you can evaluate and change rx quickly

 

[Raryn]

1. As others have said, if you have a patient with an infection (presumably you can diagnose this by symptoms and UA), and you are concerned enough about bacteremia to order blood cultures, why on earth would you send them home?

2. If you have a source (urine), why on earth would you order blood cultures?

2. Because bacteremia will affect duration of treatment. Mind you, I don't blood culture my outpatients with UTIs... but most certainly the ones getting admitted should get it.

 

[deleted245139]

I like the IDEA of this but see the utility in a REALLY limited set of patients...
Those with immunosuppression, transplant patients, etc... So long as this is done in conjunction with their doc, the whole crew is on board, and you have reliable follow up.
The false positive risk is a real limiter on a large scale... the loss to follow-up rate combined with a few skin flora contaminants and BOOM... bad news when one of those people looked amazing when you discharged turns around and returns septic...
That said, clever thought... May use this sometime

 

[Tiger26]

Agree with Wilco--I do this somewhat regularly, provided that they aren't neutropenic and are well appearing with no other etiology after a thorough workup.

The caveat is that I work in a highly collaborative large group practice where I can get next day followup with primary care and most specialists in a well-insured patient population that's overwhelmingly elderly (i.e. not uncommon to have multiple 90+ year olds in the department).

I regularly follow up on labs and often call back discharged patient's for followup when I can.

 

[coachB]

2. Because bacteremia will affect duration of treatment. Mind you, I don't blood culture my outpatients with UTIs... but most certainly the ones getting admitted should get it.

Actually, this is not true.

 

[Janders]

1. As others have said, if you have a patient with an infection (presumably you can diagnose this by symptoms and UA), and you are concerned enough about bacteremia to order blood cultures, why on earth would you send them home?

2. If you have a source (urine), why on earth would you order blood cultures?

Eh, sometimes its due to partners ordering bld cx when I wouldn't; sometimes its due to "shotgun" ordering of multiple tests up front (i.e. I'm getting my blood cx with the first stick, not waiting 2hr to get U/A results and THEN getting blood cx...). But most
often these blood cx are part of a sepsis order set for people with set criteria at triage, etc. So a 25F who pops in with temp 101.7, HR 120, and otherwise normal vitals/exam and UTI sx might get a reflex lactate/cx. Then after two L NS and tylenol and, say, ceftriaxone, she has completely normal vitals, its playing on instagram while eating doritos and begging to leave.

Then 2 days later she grows E Coli.

I'm not saying its best practice to send these bld cx and discharge the patient home, but speaking from experience of followup up a handful of these over the past few years, they all tend to be doing much better on their oral abx.



This is going to be more common with the roll out of the new set of sepsis quality measures in October, one of which demands blood cx before abx on sepsis patients (relatively loosely defined).

 

[Raryn]

Actually, this is not true.

*shrug*. Assuming a patient has defervesced, treatment of an uncomplicated UTI is fine for 3 days, or pyelo for 5-7 days. If there's bacteremia, we treat for 7-14 days. May just be an institutional preference, but it does affect our management.

I suppose one could just treat everyone for 7 days (given that there's overlap there), but (at least for inpatients) if someone is more ill, we frequently prefer the longer course. I'm not saying you should blood culture all your outpatient cystitis though, that would be silly.

 

[Rendar5]

An 80year old woman comes in with a 102 fever and that's it or maybe uri symptoms, blood cultures are getting ordered. I get yelled at if they don't. If every lab and a imaging test is normal on a sepsis work up and she wants to go home, she ended up getting blood cultures and a discharge ( with a properly documented discussion esp. With family there )e. Not sure it's a huge deal. She gets a call a couple days later if gram negatives show up. Antibiotics are not without consequence so maybe she gets them maybe not. My old hospital had a slow turn around so they probably did Due to their sepsis guidelines. Current hospital has a 30 min turn around time so I can see really quickly if they're truly septic or not.