The onset of a new fever or presence of a new or increasing leukocytosis is a problem that is encountered frequently in any intensive care unit (ICU) and can lead to an extensive workup, which subjects patients to many tests and often produces inconclusive results. Fevers may be a generalized response to a noninfectious inflammatory state such as those related to postoperative changes, alcohol or drug withdrawal, transfusion of blood products, fever, pancreatitis, adrenal insuffiency, deep venous thrombosis, and various other etiologies. Infections, however, are a much more common cause of fever with the prevalence of nosocomial infections in ICUs ranging from 3% to 31%. The most commonly reported infections in the ICU are ventilator-associated pneumonia, bloodstream infection, and catheter-related infections. The onsent of new fever will usually result in a workup obtaining tests that are directed to the most common and likely etiologies. This will often include cultures obtained from blood, urine, and sputum and a chest radiograph to assess for the presence of atelectasis or in-filtrates. If the etiology is not obvious based on preliminary studies, further workup such computed tomography (CT) scans to rule out fluid collections will be done based on the next most likely sources. By this point noninfectious etiologies are considered, as are less common infectious causes such as central nervous system (CNS) infections and infected pleural effusion (parapneumonic effusions or empyemas).

During routine daily testing of ICU patients, the presence of a pleural effusion is often noted on chest radiograph. In a recent study of medical ICU patients, it was noted that most of the ICU patients were admitted for conditions other than pleural disease, but the presence of pleural effusions was common secondary to pleural effects of pulmonary parnenchymal disorders and dysfunction of other organ systems. In this study, pleural effusions resulted from noninfectious causes in 82% of patients found to have an effusion. Of these patients the most common cause of pleural effusions was heart failure diagnosed as either a primary condition or developing after aggressive fluid resuscitation. Only 11% of patients in this study were determined to have an infected pleural effusion. The suspicion for infection was
raised when a patient remained toxic despite antiobiotic coverage. Of all the study patients with an effusion, 21% of the patients underwent a thoracentesis to rule out a malignant effusion or an infection. Only three of eight with a suspected infection had a successful thoracentesis, with others being unsuccessful secondary to small size of effusion or termination of procedure secondary to instability. The patients who did not undergo a thoracentesis had resolution of their pleural effusion after initiation or changes in antibiotics. Another study looking specifically at febrile medical ICU patients and assessing the utility of ultrasound for diagnosis of empyemas found that 62% of patients with fevers and a pleural effusion did have an infectious exudate. The prevalence of empyemas (a potentially life-threatening condition), however, was only 16%. In this particular study, the specific findings such as complex, septated patterns on ultrasound increased the likelihood that an effusion was truly an empyema that would require drainage. The remainder of pleural effusions were found to be parapneumonic effusions or transudative (noninfectious/noninflammatory collections), which have been found to resolve with appropriate antibiotic coverage and observation, respectively.