HEAT

Published

Young P, et al. "Acetaminophen for Fever in Critically Ill Patients with Suspected Infection". The New England Journal of Medicine. 2015. 373(23):2215-2224.
PubMed • Full text • PDF

Clinical Question

In adult ICU patients with fever and probable infection, does administration of acetaminophen result in more ICU-free days compared to no treatment of the fever?

Bottom Line

Among ICU patients with probable infection, treating fevers with acetaminophen does not improve mortality or decrease ICU days. Among ICU survivors, acetaminophen shortened ICU length of stay and delayed death in non-survivors.

Major Points

Antipyretic agents are commonly administered to ICU patients with fever from a known or suspected infection, on the assumption that restoration of a euthermic state may reduce physiological stress. Yet at the same time fever may impair microorganism growth, enhance the efficacy of antibiotics, and may have additional favorable effects, and thus the routine use of antipyretic agents such as acetaminophen has been called into question.

The 2015 HEAT trial randomized 700 patients with fevers due to documented or suspected infection to either acetaminophen or placebo, continued until cessation of antibiotics, resolution of fevers, ICU discharge, or death. At 28 days, there was no significant difference in mortality, ICU length of stay, or days requiring inotropes, ventilators, or RRT. Prespecified analysis revealed that among those that ultimately survived the trial, acetaminophen use was associated with a shorter ICU length of stay; among non-survivors, acetaminophen use was associated with a longer ICU length of stay.

These findings are in line with previous studies and meta-analyses, which have found no benefit to acetaminophen use for this indication. Though a randomized study in 2010 did demonstrate a decrease in early mortality and requirement of vasopressors a subsequent meta-analysis failed to find mortality benefits for treating fevers infectious with acetaminophen. Similarly, a 2015 meta-analysis of 6 trials and 819 patients failed to find any benefit of acetaminophen use among patients with sepsis.

Taken as a whole, these findings suggest that outside of hyperthermic syndromes (e.g. heat stroke), there are no clinically significant benefits demonstrated from using acetaminophen to treat fevers in critically ill patients with sepsis.

Guidelines

As of August 2017, no guidelines have been published that reflect the results of this trial.

Design

• Multicenter, prospective, blinded, randomized, controlled trial
• N=700 febrile ICU patients
  • Intensive (n=346 acetaminophen)
  • Standard (n=344 placebo)
• Setting: 23 medical-surgical ICUs in Australia and New Zealand
• Enrollment: February 2013 - July 2014
• Follow-up: 28 days
• Analysis: Intention-to-treat
• Primary outcome: ICU-free days at day 28

Population

Inclusion Criteria

• Age ≥16 years
• Fever ≥38°C
• Receiving antimicrobial therapy for known or suspected infection
• Enrolled within first 12 hours of fever

Exclusion Criteria

• Acute brain disorders
• Severe liver dysfunction or other contraindications to acetaminophen or NSAIDs
• Admission following therapeutic hypothermia or where need for therapeutic hypothermia is anticipated
• Hyperthermic syndromes (eg, heat stroke, thyrotoxicosis, malignant hyperthermia, neuroleptic malignant syndrome, drug-induced hyperthermia)
• Death perceived to be imminent in <24 hours
• Rhabdomyolysis deemed clinically significant
• Pregnancy
• <12 hour stay in ICU

Baseline Characteristics

From the acetaminophen group.

• Demographics: Age 59 years, female 35%
• Weight: 86.2 kg
• APACHE-II Scores: 19.1
• Requiring invasive mechanical ventilation: 50.7%
• Requiring non-invasive mechanical ventilation: 6.1%
• PMH: Diabetes 26.2%, cancer 21.3%, ischemic heart disease 15.0%, COPD 11.8%, CHF 3.7%, ESRD 2.9%, HIV 2.0%
• Sepsis diagnosis: Sepsis 99.7%, severe sepsis 83.3%, septic shock 18.7%
• Site of infection: Pulmonary 50.7%, intra-abdominal 12.7%, UTI 9.2%, skin/soft tissue 7.2%, bone/joint 3.5%, bloodstream 4.6%, neutropenic sepsis 2.6%, endocarditis 0.9%, unclear 5.2%
• Infection data:
  • Organism identified: 62.5%
  • Positive blood culture: 28%
  • Type of microbe:
    • Gram positive (n): 122
      • S. aureus: 12.7%
      • MRSA: 2.6%
      • S. pneumoniae: 4.9%
      • Enterococcus: 3.7%
    • Gram negative (n): 112
      • E. coli: 12.1%
      • Haemophilus spp.: 4%
      • Pseudomonas spp. : 4.3%
      • Klebsiella spp.: 2.6%
      • Enterobacter spp.: 2.0%
    • Others (n): 51
      • Influenza: 6.3% (vs 2.9% in placebo group)
      • Legionella spp.: 1.4%
      • Candida spp.: 2.0%

Interventions

• Randomization to acetaminophen or placebo:
  • Acetaminophen 1 g IV q6h
  • D5 water IV q6h
• Temperature measured at 8am on day 2 of study and if >37.5°C, study medication given regularly, then reassessed each morning at 8am
• Study treatment stopped once temp <37.5°C for 24 hours
• If fevers of 38°C return in next 48 hours, then study treatment restarted every 6 hours
• Given in identical 100cc glass bottles
• Continued to treat up to 28 days after enrollment

Outcomes

Comparisons are acetaminophen vs. placebo.

Primary Outcomes

ICU-free days at day 28

23 vs. 22 days (P=0.07)

Secondary Outcomes

28-Day Mortality

13.9% vs. 13.7% (RR=1.02; P=0.94)

90-Day Mortality

15.9% vs. 16.5% (RR=0.96; P=0.84)

No significant differences in ICU length of stay (4.1 vs. 4.2 days), hospital length of stay (13.7 vs. 13.8 days), ventilator-free days (27 vs. 26 days), inotrope-free days (27 vs. 27 days), or RRT-free days (28 vs. 28 days)

Survivors

Median ICU length of stay: 3.5 vs. 4.3 days (P=0.01)

Median hospital length of stay: 13.2 vs. 14.1 days (P=0.13)

Non-survivors

Median ICU length of stay: 10.4 vs. 4.0 days (P<0.001)

Median hospital length of stay: 13.9 vs. 7.7 days (P<0.001)

Subgroup Analysis

• No significant differences in the primary outcome when analyzed by (1) Presence or absence of septic shock, (2) Use or non-use of aspirin, (3) Presence or absence of high fever (T ≥39.0°C), (4) Location of infection acquisition (community, hospital, or ICU).

Adverse Events

• There was 1 death in the placebo group. This patient developed significant hyperthermia on day 5, with core body temp of 42.7°C.

Criticisms

• All patients were from the Australia/New Zealand health systems and may not be generalizable to that of other countries.
• Antibiotic exposure days were not measured in this study, which might have been an important improved outcome.
• Patient comfort was not measured in this study, which might have been an important improved outcome.
• Unlike common clinical practice, acetaminophen was given around-the-clock in this trial (not as-needed), which may have lead to excessive acetaminophen administration.

Funding

• Funded by the Health Research Council of New Zealand and others