Fever in the ICU

67M in ICU (13d) for chronic resp failure s/p trach. Consult for 5d fevers & Q: WBC-tagged or PET/CT useful?

WBC 17k, CMP wnl, BCx -, UCx <50k CFU E.Coli, Trach Cx -. CT sinus/C/A/P unrevealing.

Do you recommend WBC-tagged or PET/CT scan?

Let’s review the approach to fever in the ICU while answering some of your questions.

#IDtweetorial 1 of 2

  1. Perform an ID-worthy HPI.

Careful, some clues can be misleading.
🤔What is signal vs. noise?

For example, in a study of FUO, there were ~ 15 potential diagnostic clues/pt, of which 81% proved to be misleading.


Briefly, pt w/DM found down under horse carriage unconcious for unknown period of⏲️. No sx prior. Lived in farm (dogs, horses, cattle) w/wife. (-) travel, other exposures

On admission: CT head w/ SAH. L PTX s/p CT, now removed. Ribs & L pelvis fx, medically managed.

PE: T38.9(102), HR110, BP137/80. unresponsive. PEG/trach site ok, rectal tube w/soft stool. (-) neck stiff, pressure ulc, ext edema, phlebitis (PIVs). No foley. Rest u/r.

C. diff -. UA 10-20 WBC. PCT n/a. CRP not done. Rest u/r or as 1st tweet.

Med review, unrevealing

For the ICU,
2. Head-to-toe approach, nicely mentioned by @TxID_Edu

Recommended read: Persistent fever in the ICU


2.a Drug fever
-Almost any drug. Hypersensitivity most common
-Pt “inappropriately well” for degree of fever.
-Usually 2nd wk (hrs-months)
-Some w/eos, rash, relative bradycardia
-Dx of exclusion. Stop➡️monitor. Rechallenge can be diagnostic


2.b DVT/PE. Can it cause fever?

Great tweetorial by @JonathanRyderMD addressing this:

Even if found, do not anchor to this diagnosis as cause of fever

2.c Neurogenic fever
-Thought 2/2 hypothalamic injury➡️disruption thermoregulation
-Usually no diurnal variation, resistant 2 antipyretics
-Dx of Exclusion
-SAH, intraventricular bleed, onset w/i 72h, & longer fever duration➡️predict neurogenic fever


After providing more information,

Repeating poll #IDtwitter #IDfellows

Do you recommend WBC-tagged or PET/CT scan?

Case had no dx after > 3d of investigations➡️met definition of Nosocomial FUO.

Excellent review of FUO here: https://bit.ly/2F5JhcM

A structured diagnostic work-up approach should be done before PET/CT scan. Repeat history and exam looking for any diagnostic clues

You can see WBC-tagged scan is not included. It localizes inflamm via neutrophils ➡️bacterial infx.

In a meta-analysis, Sn 33% & Sp 83% in FUO. https://bit.ly/2DOp6Qa

Found more useful w/suspected bacterial infx➡️OM (70%) or Vasc graft infx (67%).


Why is FDG-PET/CT included?

Fluorodeoxyglucose (FDG) from FDG-PET is used avidly (⬆️glucose consumers) by tumor and ALL inflammatory cells (macrophages, lymphocytes, neutrophils) ➡️functional modality for cancer & infx

CT adds anatomical info to accurately localize source

In a meta-analysis, FDG-PET/CT had a diagnostic yield of 56%, however when analyzing only studies that had already prior imaging studies (n=5), diagnostic yield ⬇️to 32%.


So, what about fever w/o source in the ICU?

In a retrospective study w/ventilated patients & suspected infx (fever, SIRS, ?septic emboli, ?mediastinitis, ?infected intraabdominal fluid collection), an overall accuracy of 91% was found for PET/CT.


Our pt underwent a PET/CT scan which showed increased uptake of the prostate. His PSA was elevated. He had a UCx positive for E. coli <50k CFU and patient received antibiotics for prostatitis with resolution of his fever.

In summary:

✅Head-toe-approach for Fever in ICU
✅Use a structured dx work-up approach
✅WBC-tagged scan may be useful when bacterial infx suspected (OM, vascular graft)
✅FDG-PET/CT scan can be considered after comprehensive HPI and testing has been done w/o diagnosis

Thank you all for reading and participating. Feedback always welcome.

Originally tweeted by Infectious Diseases Fellows Network (@ID_fellows) on 14 August, 2020.

Leave a Reply