Monday, April 26, 2010

April 20th - MHH - Fever in the MICU

At morning report a case was presented of a patient that developed mucormycosis after being in the ICU.

Here is a general overview/approach to patients that fever either in the MICU, or after being transferred out of the MICU. This may be hard to piece out, as some studies have suggested that approximately 70% of MICU admissions have a documented fever on admission, and 55%-90% of those admitted without a fever develop one sometime during their MICU stay.

Why is this important? Well, besides finding a cause - mortality seems to be affected. Studies disagree - some say that "any fever" increases mortality, and others state that only prolonged fever increases mortality. Nonetheless, causes must be investigated.

We can divide the categories into infectious and non-infectious causes. We commonly think of common infectious causes first (i.e. PNA, UTI etc) - but when this does not give an answer, we must turn to the less common infectious causes and the non-infectious causes.

Infectious Causes

Nosocomial Pneumonia - Consider in a patient with fever, new infiltrates, increased sputum production, leukocytosis and increased oxygen requirements. Have a high suspicion with a recent hospital admission on a patient on the vent.  Common pathogens implicated are Pseudomonas and Staph aureus, but instances of Stenotrophomonas and Acinetobacter are non uncommon in our ICUs. Viral causes need to be considered as well. Aspiration pneumonitis, which is usually not infectious, should also be on the differential. Perform trach cultures and then BAL if needed.

Bloodstream Infections - Staph is a common cause, as well as Gram Negatives. Consider Fungal infections (esp. Candida) with long-term indwelling catheters or immunosuppressed patients. You must think of catheter-related blood stream infections in all ICU patients. This is especially true when ANY line is present for >48 hours. If there is purulence at the site of catheter insertion, it usually needs to be removed and the cath tip sent for cultures (>15 CFU is suggestive of catheter related infection). Otherwise, you can draw blood cultures (from the catheter and from a distant peripheral site) - and check time to positivity. There are both quantitative (i.e. >5x colony counts from catheter compared to peripheral) and qualitative guidelines (time to positivity in lab is quicker from catheter than peripheral site) for diagnosis.

Urinary Tract Infections - Suspect in all patients with a Foley. Besides a positive culture, remember that you need both a positive Ua (with WBCs) to help diagnose this. This is because patients with chronic foleys can become colonized, and have positive cultures WITHOUT signs of inflammation on the urinalysis

Wound infections - Make sure to check the skin for infected decubitus ulcers and any other sites of soft tissue infection.

Clostrium Difficile Infection - Suspect with high WBC count, fever, diarrhea in a patient that has received antibiotics or chemotherapy in the weeks prior to the fever. Common antibiotics implicated are cephalosporings, clindamycin and the fluoroquinolones. In some patients fever can be the only initial manifestation of pseudomembranous colitis, so have a high index of suspicion. Send C. diff toxin x 2. If negative and diagnosis is high on the differential, flex sig can be done. Since this is such a severe infection, empiric antibiotics with metronidazole may be warranted while pending culture results.

Sinusitis - Cause of new-onset fever in about 5% of MICU patients. Suspect this in a patient that has a long-term nasogastric or nasotracheal tube. Also seen with nasal packing, facial trauma, and in patients on steroids. Most commonly present in the Maxillary sinuses. Workup with a plan-film of the sinuses and then do CT if the plain films are non-revealing or not definitive. Radiographic evidence alone does not give a definitive answer, so drainage with culture is needed. Most pathogens are Gram Negative Rods (esp pseudomonas, e coli and proteus), but Staph, Candida,  are not uncommon.

Abscess - Consider an abscess in a patient that has undergone any surgery or abdominal intervention prior to the surgery. This is not always precedent with these infections, so have a high likelihood for this if no other causes are found. Patients commonly "spike" fevers with these abscesses at about the same time daily. Imaging of the abdomen and pelvis should be considered in the workup

Endocarditis - This can occur without positive blood cultures, especially if the patient is partially treated or has a "culture-negative" cause (i.e. HACEK, psittacosis etc). Especially consider with poor dentition, prosthetic value, or IV Drug use. A TTE may be performed as the initial diagnostic test.

Other infections to consider - Does the patient have meningitis? Do they have osteomyelitis? Do they have peritonitis? Consider these infectious causes as well, even though they might be further down the list on the differential

Non-Infectious Causes

Thrombosis - DVT and PE should be considered in patients. Don't think that being on prophylaxis excludes this. Use the Well's Score and clinical predication rules to determine the possibility. Also, very extensive superficial thrombophlebitis can cause fever - so check all prior and current access sites.

Drug-Induced Fever - Can be from multiple agents. Common drugs are penicillins, cephalosporins, phenytoin, procainamide and quinidine. Usually patients with "drug" fever do not feel "feverish" but can manifest chills and myalgias. Drug rash , leukocytosis and eosinophillia can clue to in to this, but are frequently absent (less than 10-15% of cases!). This is not a common cause, but should be considered when other workup is non-revealing. You should, however consider drug fever in a patient that has relative bradycardia in relation to the temperature!!! Especially consider this with the scenario of an infectious source that is found with temperature initially improving during the admission - then recurring. Treatment is stopping the agent. Usually the fever will resolve in a few days if this is the cause. If the infection is improving, some will actually continue the antibiotics until repeat cultures are negative despite the presence of drug-induced fever. Also consider causes of hyperthermia (in prior blog) - i.e. serotonin syndrome, neuroleptic malignant syndrome and malignant hyperthermia. Consider these when commonly implicated agents are being given, or were recently administered. However, these conditions usually have other findings evident on exam or labs.

Ischemic Disease - Ischemic Bowel Disease or Ischemic Colitis Can produce fever. Not uncommonly, temperature elevations can be seen with myocardial ischemia or CVA. Related to the cardiac system - post MI pericarditis (Dressler's) can also cause low grade fevers.

CVA - Okay, just mentioned above, but should be another category. This is especially common when the posterior fossa or hypothalamus are involved. Subarachnoid hemorrages are also notorious for causing fever in the MICU. Have a high suspicion of this post-trauma, in the elderly, in alcoholics, or in patients with coagulopathy

Acalculous Cholecysitits - Think in our medical patients that are either septic, on a vent, or is getting parental nutrition. Perform U/S as first test. May actually need a CT scan If non-conclusive since HIDAAs in this setting give false positive rates >50%.

Pancreatitis - In the ICU setting, pancreatitis causes by procedures (ERCP), infections and medications should be considered. 

Acute Liver Injury - Unless there is an infectious agent present, this is usually due to ischemic causes or drugs. High transaminase levels can clue this into the cause of fever

Endocrine Diseases - Adrenal Insufficiency and Thyrotoxicosis are possible causes. Adrenal Insufficiency in the MICU is more common. Usually the adrenal insufficiency which is severe enough to cause fever is not due to primary causes (i.e. Addisons) or withdrawal of steroids - but from SIRS/Sepsis, adrenal hemorrhage with DIC or other causes, or adrenal infiltration (i.e. TB). Keep Pheochomocytomas in mind as well.

TTP/HUS - Consider this as a cause of fever especially from drug effect (TTP) or infections (HUS). Fever is usually a LATE manifestation of TTP, so be on the lookout for these conditions in the setting of thrombocytopenia or microangiopathic hemolytic anemia. It is not uncommon for MICU patients to have thrombocytopenia, so have a high suspicion and sent a blood smear to look for schistocytes.

Post-Op State/Atelectasis - Fever is not uncommon after surgery, but usually do not produce very high temperatures, and rarely last >72 hours. There is some "controversy" that atelectasis actually causes fever. Some believe that the post-operative state, which causes atelectasis, is the actual cause of fever. Still, this is up for debate...

Alcholol Withdrawal and Delerium Tremens - Consider this could be the cause of fever and a hyperdynamic state in a patient with history of significant EtOH use prior to their MICU admission. Not uncommonly seen with benzo withdrawal, and even fevers reports with opiate or barbiturate withdrawal.

Iatrogenic Causes - There are reports of fevers being reported in the MICU due to faulty equipment (i.e. heated mattresses) - this is not a common cause, but it's certainly a cost-effective element to look for. Sometimes there is a febrile response in patients post-procedure - after endoscopy, bronchoscopy or even post endotracheal suctioning. Make sure there is no reaction due to blood product administration as well.

And as always, with any differential regarding fever don't forget the vast array of rheumatic and malignant causes.

The HIV/Cancer Patient:
Although the above applies to the HIV patient (or any immunocompromised patient), some other considerations include Acute HIV Infection, Higher Suspicion for Meningitis/Encephalitis and Consideration of Typhilitis (necrotizing infection of colon) in cancer patients.

This is an approach to workup in the article by Dimopoulos and Falagas:


First the Obvious:
1) Find the cause and treat it specifically
2) Until a cause is found, or unless a non-infectious source is obvious - give broad-spectrum antibiotics to those patients that are unstable (which would be the norm in the MICU)

Now the controversy:
Should you just lower the temperature.  There are considerable arguments for and against the use of antipyretics.  There is no doubt that they should be used in two circumstances: When the temperature is extremely elevated (i.e >103.5 degrees) or when any fever is present in a patient where the state would be detrimental (i.e in the setting of MI or CVA). There is also some evidence in patients with needs for high FiO2, that lowering the fever decreases oxygen requirements.  Some will treat routinely to simply lower the temperature - there is evidence in the surgical ICUs that this either increases mortality or makes no difference.  Similar evidence is lacking in the MICU patients.

References -
Dimopoulos, G & Falagas, M. Approach to the Febrile Patient in the ICU. Infect Dis Clin N Am 23 (2009) 471–484

Laupland, K. Fever in the critically ill medical patient. Crit Care Med. 2009; 37[Suppl.]:S273–S278

Marino, P. The ICU Book. Lipincott Williams and Wilkins. 2007

Riga, M et al.  Rhinosinusitis in the intensive care unit patients: A review of the possible underlying mechanisms and proposals for the investigation of their potential role in functional treatment interventions.  Journal of Critical Care (2010) 25, 171.e9–171.e14

Rizoli, S & Marshall, J. Saturday night fever: finding and controlling the source of sepsis in critical illness. Lancet Infectious Diseases 2001; 2: 137–44

1 comment:

Anonymous said...