Fever workups in a surgical ICU: an opportunity for diagnostic stewardship

Author(s):
Candice Preslaski; Kevin Harrell; Chelsea Horwood; Kimberly Hardin; Elizabeth Perkins-Pride; Barry Platnick; Nicole Werner

Background:

Fever is common in surgical patients and often prompts an infectious investigation. Delay in the diagnosis and treatment of infection has a measurable mortality. However, in this population fevers may be caused by transfusions, alcohol withdrawal, drug reactions, or inflammation secondary to trauma or surgery.

Hypothesis:

We hypothesize this extensive workup rarely finds infection as the cause and can add to patient morbidity and hospital cost through unnecessary testing, longer lengths of stay, and antimicrobial overuse.

Methods:

A retrospective analysis of the fever workups in a surgical ICU was performed. Patients that had blood cultures obtained were identified and then further evaluated for baseline characteristics, risk factors for hospital acquired infections, studies ordered, and infections identified. A workup was considered positive if an infection was identified and treated with antibiotics.

Results:

A total of 361 workups from 167 patients were reviewed. In addition to blood cultures, most workups included radiographic imaging, 298 (83%) chest x-rays and 146 (40%) CT scans. The most common microbiologic studies were from urine (133, 37%) and respiratory tract (127, 35%). Less than half (164, 45%) of workups resulted in finding an infection with pneumonia identified most often in 70 patients (19%). Patients with a positive workup had longer median hospital (27 vs. 17.5 days, p<0.05) and ICU (15 vs. 8 days, p<0.05) lengths of stay and were more likely to require mechanical ventilation (OR 4.3, 1.9-9.9) for longer (13 vs. 4.5 days, p<0.05). In a multivariate regression leukocytosis was associated with an increased likelihood of a positive workup for all patients (OR 1.1, 1.0-1.1), while temperature was associated with a positive workup for patients with traumatic brain injury (OR 2.5, 1.3-4.9). Surgeries involving head, face, and chest were all associated with negative workups (OR 0.3, 0.4, 0.4 respectively, all p<0.05) suggesting these procedures in particular cause a non-infectious inflammatory response. Patients on antibiotics at the time of the workup were also more likely to have a negative workup (OR 0.6, 0.4-0.9). While respiratory cultures were most likely to be positive, there was a high rate of potential overtreatment with 23 (33%) patients receiving antibiotics but not having clinical or radiologic findings.

Conclusions:

Critically ill surgical patients undergo many diagnostic tests when they experience a fever. However, fever remains a non-specific indicator of infection with less than half of workups identifying an infectious cause. Opportunities exist to focus testing within fever workups in the surgical ICU.