Journal of Intensive Care Medicine

The increasing complexity of the intensive care patient combined with the recent advances in imaging technology has generated a new perspective on intensive care radiology. The purpose of this 2-part review article is to describe the contribution of radiology to the management of these critically ill patients. The first article will discuss the impact of picture archiving and communication system (PACS) on critical care management and utility of the portable chest radiograph in the detection and evaluation of pulmonary disease with correlation to computed tomography (CT). The second article describes in more detail the increasing role of CT in diagnosis and therapeutic procedures. In particular, the implementation of CT pulmonary angiography in the evaluation of pulmonary emboli and the introduction of the new multislice detector CT scanners that allow even the most dyspneic patient to be evaluated. Pleural complications in the intensive care unit and image-guided intervention will also be discussed.

B-type natriuretic peptide (BNP) has diagnostic, therapeutic, and prognostic utility in critically ill patients. For severe sepsis and septic shock patients in particular, similar clinical utility from the most proximal aspects of hospital presentation to the intensive care unit has not been examined. BNP levels were measured at 0, 3, 6, 12, 24, 36, 48, 60, and 72 hours in 252 patients presenting to the emergency department with severe sepsis and septic shock. The clinicians were blinded to the BNP levels. Elevated BNP levels (>100 pg/mL) were seen in 42% and 69% of patients on presentation and at 24 hours, respectively. Elevated BNP ranges (>230 pg/mL) were significantly associated with myocardial dysfunction and severity of global tissue hypoxia. When adjusted for age, gender, history of heart failure, renal function, organ dysfunction, and mean arterial pressure, a BNP greater than 210 pg/mL at 24 hours was the most significant independent indicator of increased mortality: odds ratio 1.061 (1.026-1.097), P < .001, 95% confidence interval. Patients with severe sepsis and septic shock often have elevated BNP levels, which are significantly associated with organ and myocardial dysfunction, global tissue hypoxia, and mortality. Serial BNP levels may be a useful adjunct in the early detection, stratification, treatment, and prognostication of high-risk patients.

The aim of this study was to evaluate and compare brain natriuretic peptide (BNP) and cardiac troponin I (cTnI) levels as mortality prognosticator and predictor for myocardial dysfunction in severe sepsis and septic shock. Baseline clinical and biological variables were collected from 47 patients with severe sepsis or septic shock. Ventricular systolic function assessed by echocardiography was measured over a 5-day period. Both cTnI and BNP plasmatic levels were determined at intensive care unit (ICU) admission and during the following 15 days. At admission, cTnI and BNP levels were compared to those of 12 control critically ill nonseptic patients. The plasma levels of BNP and cTnI in patients with sepsis were elevated at admission and significantly higher than in the controls. Among patients with sepsis, BNP levels were significantly more elevated in nonsurvivors compared to survivors at admission and 1 day later. The cTnI levels were also significantly more elevated in nonsurvivors compared to survivors, but only at admission. From admission to day 5, patients with sepsis with left ventricular systolic dysfunction had higher BNP plasmatic concentrations than those without; differences were significant at days 3 and 4. In contrast, plasma cTnI levels were similar between the 2 groups. In critically ill patients, sepsis induces significant increase in BNP and cTnI levels. High BNP and cTnI plasma levels during ICU admission appear to be associated with poor outcome of sepsis. Time course of BNP levels seems helpful to discriminate between surviving and nonsurviving patients with sepsis and to detect myocardial dysfunction where troponin levels fail to do so.

Fever is a relatively common occurrence among patients in the intensive care setting. Although the most obvious and concerning etiology is sepsis, drug reactions, venous thromboembolism, and postsurgical fevers are all on the differential diagnosis. There is abundant evidence that fever is detrimental in acute neurologic injury. Worse outcomes are reported in acute stroke, subarachnoid hemorrhage, and traumatic brain injury. In addition to the various etiologies of fever in the intensive care setting, neurologic illness is a risk factor for neurogenic fevers. This primarily occurs in subarachnoid hemorrhage and traumatic brain injury, with hypothalamic injury being the proposed mechanism. Paroxysmal sympathetic hyperactivity is another source of hyperthermia commonly seen in the population with traumatic brain injury. This review focuses on the detrimental effects of fever on the neurologically injured as well as the risk factors and diagnosis of neurogenic fever.

Delirium is commonly described in critically ill patients as 1 factor contributing to increased length of intensive care unit and hospital stay, secondary complications, and increased mortality. Initial screening tools for delirium in hospitalized patients are generally easy to use; however, many centers have struggled with implementing these tools in a consistent and systematic manner. Haloperidol has traditionally been prescribed as the primary agent of choice for the treatment of delirium in critically ill patients. Clinicians have been challenged to consider alternative agents due to adverse effects such as extrapyramidal symptoms, QTc prolongation, and possible torsades de pointes with haloperidol use. The atypical antipsychotics are attractive alternatives to haloperidol with improved safety profiles but are flawed by limited data to support dosing and efficacy in this patient population. Future studies that provide large, prospective, double-blinded, placebo-controlled data to support the implementation of these agents as standard therapy over haloperidol are needed.

Massive hemoptysis is a life-threatening complication of many pulmonary disorders, and it occurs most com monly in association with longstanding inactive tuber culosis, bronchiectasis, lung abscess, bronchogenic car cinoma, and fungal disease. It is generally a neovascular change or local erosive effect of chronic pulmonary dis ease and may originate from either bronchial or pulmo nary circulation. Recurrent bleeding is unpredictable; therefore, diagnostic and therapeutic intervention must be undertaken with urgency. The immediate priorities must be protection of the airway to the nonbleeding lung and localization of the site of hemorrhage prefera bly by bronchoscopy, which has a high yield when per formed during active hemorrhage. Immediate control of bleeding may be obtained by local tamponade with a balloon-tipped Fogarty catheter, use of a double-lumen endotracheal tube, or angiographically guided emboliza tion. Surgical resection is the preferred definitive treat ment for those who meet operative criteria; those who lack adequate pulmonary reserve are candidates for em bolization of sites with persistent bleeding. The high mortality of conservatively treated massive hemoptysis and the current inability to predict which patients will have fatal hemorrhage mandate rapid assessment and intervention.

The objective of this retrospective study was to determine the optimal initial antibiotic regimen for hospital-acquired pneumonia using the frequency and sensitivity of Gram negative microorganisms found in sputum cultures. An antibiogram was constructed and compared with the hospital intensive care unit (ICU) antibiogram. The results yielded 191 microorganisms. The study-generated antibiogram showed that the highest percent susceptible antibiotics for all Gram-negative microorganisms were imipenem (75%) and amikacin (84%). Considering only Pseudomonas aeruginosa, the study-generated antibiogram and the hospital ICU antibiogram showed similar results, piperacillin and amikacin (86% and 82%, respectively, vs 91% and 85%, P = nonsignificant for both). The optimal empiric antibiotic regimen in the surgical ICU is different if directed against all possible microorganisms as opposed to the most prevalent microorganism P aeruginosa. Determining initial empiric antibiotic therapy using an ICU and culture-type specific antibiogram would result in a g reater likelihood that more patients would receive adequate initial antibiotic therapy.

Research in the intensive care unit (ICU) raises a number of scientific and ethical challenges. Potential participants in critical care studies are likely to be considered particularly vulnerable—they may lack sufficient capacity to make informed decisions about trial participation, their health care proxies may lack legal authority to enroll them in research trials or may not know their true intent, and the life-threatening nature of the illness may make them or their surrogates more susceptible to therapeutic misconception. Because of this, critical care investigators must exercise extreme caution when designing and conducting studies in the ICU. In this article, we review the key literature addressing the various scientific and ethical issues raised by critical care research, including questions of equipoise and the selection of control groups, informed consent, therapeutic misconception, conflict of interest, and quality improvement projects. We also describe the current status of key policy or regulatory initiatives designed to address these issues, particularly in light of recent controversies involving critical care studies like the ARDSNet trial.