Body temperature is a fundamental physiologic parameter that is altered by several pathologic processes. Abnormal or altered body temperature is a key to aid in diagnosis and is often a component of therapeutic strategy. Fever is particularly common in critically ill patients who are admitted to the intensive care unit (ICU).
However, it is unclear if alterations in temperature are themselves pathologic or if they represent an adaptive response to pathology. Thus, before reviewing the evidence and making recommendations for fever control in the critically ill, it is fundamentally important to understand the mechanisms of temperature control and fever generation.
Infection such as that induced by gram-negative bacteria induces several important changes in host physiology by generating pathogen-associated molecular patterns (PAMPs) that bind to pattern recognition receptors (PRRs) such as toll-like receptor 4 (TLR-4). This process activates host fixed-tissue macrophages to release circulating substances such as prostanoids that alter activity in thermoregulatory regions of the preoptic area (POA) of the rostral hypothalamus ( Fig. 19-1 ). Inflammatory cytokines released from immune cells also induce expression of prostaglandin E 2 (PGE 2 ) in endothelial cells around the POA that further signal the hypothalamic temperature control center. PGE 2 reduces cyclic adenosine monophosphate (cAMP) levels in the POA, and reduced activity of cAMP causes fever. Thus, temperature and generation of fever are controlled predominantly via a PGE 2 –dependent pathway; however, it is possible that cytokines or the hepatic branch of the vagus contribute to fever, although such suggestions are controversial.
Data suggest that fever accompanies infection or sterile noninfective inflammation (postoperative patients, pancreatitis, trauma, inflammatory disorders such as lupus) and drug reactions. In infection, fever may modulate the host defense response, affecting cellular innate immunity, CD4 T cells, or B cells. Fever also decreases inflammatory cytokines such as tumor necrosis factor-α and delays expression of interferon-γ expression, rapidly increases expression of cytoprotective heat shock proteins, and inhibits nuclear factor-κB. These changes can limit the deleterious effects of infection, reducing organ dysfunction and improving survival. Fever may also affect the pathogen by inhibiting growth or increasing antibiotic susceptibility. These effects suggest that fever is adaptive.
Conversely, an elevated body temperature has several detrimental effects that need to be considered. Fever increases energy expenditure and oxygen demand and decreases vascular tone, resulting in an imbalance between oxygen demand and supply. Thus, controlling fever may decrease (1) global oxygen demand when oxygen supply is limited ; (2) the risk of oxygen supply/demand mismatch; (3) the metabolic rate, and with it catabolism; and (4) the risk of septic encephalopathy, thereby decreasing the need for sedatives and antipsychotic drugs. These possibilities provide a rationale for treating fever.
The association between fever and the clinical outcome of critically ill patients has been assessed in several observational cohort studies. A retrospective observational cohort study of 24,204 adult patients in medical-surgical ICUs demonstrated that the presence of fever greater than 39.5° C was associated with significantly increased mortality rates. The effect of fever on outcome may be particularly critical in patients with neurologic disorders. A prospective cohort study of patients emergently admitted to a neurological ICU revealed that elevated body temperature was independently associated with increased length of ICU stay, increased mortality, and worse hospital disposition. Controlling body temperature to avoid fever in patients who had cardiac arrest outside of the hospital has been widely studied and appears to improve neurologic outcomes and reduce mortality ( Table 19-1 ), but a randomized controlled trial (RCT) showed that hypothermia in severe bacterial meningitis did not improve outcome and may even have been harmful ( Table 19-1 ).
| Study, Year | Setting | Population | Patients (n) | Intervention | Primary Outcome | Main Findings |
|---|---|---|---|---|---|---|
| Bernard, 2002 | RCT | Out-of-hospital cardiac arrest | 77 | Target temperature 33° C within 2 hours after the ROSC and maintained for 12 hours | Survival to hospital discharge with good neurologic function | Patients treated with hypothermia had a good outcome compared with those with normothermia (49% vs. 26%, P = .046). |
| Hypothermia After Cardiac Arrest study group, 2002 | RCT | Postcardiac arrest | 136 | Target temperature 32 to 34° C for 24 hours | Favorable neurologic outcome within 6 months after cardiac arrest | Patients in the hypothermia group had a favorable neurologic outcome compared with the normothermia group (55% vs. 39%, RR 1.40, 95% CI 1.08-1.81). |
| Nielsen, 2013 | RCT | Out-of-hospital cardiac arrest | 939 | Targeted temperature 33 or 36° C | Mortality | There were no significant differences between the two groups in neurologic function or death at 180 days. |
| Mourvillier, 2013 | RCT | Severe bacterial meningitis | 130 | Targeted temperature 32 to 34° C for 48 hours using cold IV fluids | Glasgow outcome scale score at 3 months | Moderate hypothermia did not improve outcome in patients with severe bacterial meningitis (unfavorable outcome, 86% (hypothermia) vs. 74% (control), RR, 2.17, 95% CI, 0.78–6.01, P = 0.13). |
The benefits of actively preventing fever in ICU patients who do not have a neurologic disorder are unknown. As such, guidelines generated by the American College of Critical Care Medicine (ACCM) and the Infectious Diseases Society of America (IDSA) make no recommendations regarding temperature management in adult patients. We therefore review recent evidence from RCTs.
Evidence from Randomized Controlled Trials
The mechanism of cooling to control fever is likely important. Interventions to control fever that were assessed in RCTs include external cooling and nonsteroidal anti-inflammatory drugs (NSAIDs) such as ibuprofen and acetaminophen (paracetamol).
Although fever is inconsistently defined, a core body temperature of 38.3° C or greater is most commonly used in the literature and in the ACCM/IDSA guidelines for fever in critically ill adult patients. These guidelines recommended that an intravascular, esophageal, or bladder thermistor most accurately measures temperature. In addition, the guidelines state that oral or infrared ear thermometry are accurate and acceptable for use in the ICU, whereas temporal artery, axillary, or chemical dot temperatures are not.
External Cooling
There have been two RCTs of external cooling for non-neurologic critically ill patients ( Table 19-2 ). The first RCT (n = 38) assessed the efficacy of external cooling in febrile patients admitted to a surgical ICU without neurologic factors, and it failed to demonstrate a significant effect on length of hospital stay or mortality. This study was limited by a sample size that left it severely underpowered and ineffective external cooling—patients who received external cooling had similar body temperature compared with the control group. The second RCT was conducted to evaluate the safety and efficacy of fever control by external cooling in febrile patients who had septic shock. This RCT enrolled 200 critically ill patients who had infection with a core body temperature greater than 38.3° C and required vasopressor infusion, ventilator support, and sedation. The intervention group (n = 101) had external cooling to normothermia for 48 hours whereas controls (n = 99) had usual care without cooling. Because there was no difference in adverse events between the two groups, fever control by the external cooling was demonstrated to be safe. This study successfully attained rapid, significant fever control by external cooling. However, the primary outcome variable, the number of patients reaching a 50% decrease in the dose of vasopressors within 48 hours, was not significantly different between the two groups. The patients who had external cooling required significantly lower vasopressor dosage at 12 hours and had significantly decreased 14-day mortality compared with patients without external cooling (19% vs. 34 %; P = .013). External cooling can cause shivering and thus increase the need for sedation and even neuromuscular paralysis. Although the use of sedation and neuromuscular paralysis was similar between groups, it is unclear if sedation of neuromuscular blockade was managed by a protocol. Importantly, the groups differed at baseline; initial vasopressor doses in patients without external cooling were higher than the patients with external coolings. The adjusted analyses confirmed the efficacy of cooling; nonetheless, post hoc statistical adjustment of baseline differences in the primary outcome variable (vasopressor dose) between study groups is always a concern in the interpretation of RCTs.
| Study, Year | Setting | Population | Patients (n) | Intervention | Duration | Main Findings |
|---|---|---|---|---|---|---|
| Gozzoli, 2001 | Surgical ICU | Postoperative | 38 | External cooling | 24 hours | Treatment of fever had no effect on length of ICU stay, duration of hospital stay, and ICU mortality |
| Schortgen, 2012 | Medical ICUs | Septic shock | 200 | External cooling | 48 hours | Decreased vasopressor requirement and decreased mortality at 14 days |
| Bernard, 1997 | Surgical, trauma, and medical centers | Severe sepsis | 455 | IV ibuprofen | Every 6 hours for 48 hours | No significant improvement of mortality rate at 30 days |
| Morris, 2010 | Medical hospitals | Critically ill | 53 | IV ibuprofen | Every 4 hours for 24 hours | No significant differences in renal function, bleeding, or mortality |
| Schulman, 2005 | Trauma ICU | Trauma | 82 | Enteral acetaminophen (paracetamol) every 6 hours for >38.5° C External cooling added for >39.5° C | Febrile episode | Terminated early because of strong trend to higher mortality rate ( P = .06) in intervention group |
Although additional trials need to be performed, external cooling is simple, easily implemented, and lacks some of the potential adverse effects of antipyretics. Subsequent validation in larger RCTs is required; however, we suggest that cooling of febrile patients who have septic shock should be considered in patients who are on high doses of vasopressor, who require inotropic agents (e.g., dobutamine), or who have marked tachycardia.
Ibuprofen
A large (n = 450) RCT funded by the National Institutes of Health published in 1997 compared ibuprofen, an NSAID, with placebo in patients with severe sepsis. Ibuprofen significantly decreased core temperature, heart rate, oxygen consumption, and lactate level. In addition, potential drug-induced adverse events, including renal dysfunction and gastrointestinal bleeding, did not differ between the two groups. There was, however, no significant difference in organ failure or 30-day mortality between study groups (ibuprofen vs. placebo, 37% vs. 40%, respectively).
A smaller RCT (n = 53) evaluated the efficacy and safety of intravenous ibuprofen (100, 200, or 400 milligrams every 4 hours for six doses) in critically ill patients with fever. Intravenous ibuprofen effectively controlled fever and was not associated with an increase in adverse events, but there was no difference in mortality between ibuprofen and placebo. This was a very small RCT, though, that was not even close to being of adequate power for mortality assessment. Taken together, ibuprofen can safely reduce fever in critically ill patients, but there is no reported evidence to support a beneficial effect on clinical outcomes.
Acetaminophen (Paracetamol)
To our knowledge, there are no RCTs of acetaminophen (paracetamol) for fever control in sepsis or septic shock. One small RCT ( Table 19-2 ) compared enteral acetaminophen (paracetamol) (650 mg every 6 hours for the febrile episode) for temperatures greater than 38.5° C with additional external cooling for temperatures greater than 40° C to a specified fever control regimen in 82 trauma patients without traumatic brain injury. There were no significant differences in infections between the two groups; however, a trend to increased mortality in the aggressive fever control group (mortality: aggressive vs. permissive, 16% vs. 3%; P = .06) resulted in premature termination of the RCT at an interim analysis.
Related posts:
What Lessons Have Intensivists Learned During the Evidence-Based Medicine Era?
What Is the Role of Invasive Hemodynamic Monitoring in Critical Care?
What Fluids Should I Give to the Critically Ill Patient? What Fluids Should I Avoid?
Should Exacerbations of COPD Be Managed in the Intensive Care Unit?
Are Anti-inflammatory Therapies in ARDS Effective?
Has Outcome in Sepsis Improved? What Has Worked? What Has Not Worked?
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
