Acetaminophen (paracetamol) against fever in infections: at best ineffective and may even contribute to mortality
Verbeek, E.M.E., Arntzenius, A.B., Wattel-Louis, G.H.
Locatie(s): Auditorium 1
Categorie(ën): Plenaire sessie; Presentations "Schop de Heilige Huisjes omver!"
Moderator:
Mw. dr. H.A.H. Kaasjager, Arnhem
Panelleden
Mw. dr. P.C. Oldenburg, Amersfoort
Mw. W.E.M. Schouten, Amsterdam
Dr. F.L.J. Visseren, Utrecht
Introduction: The symptomatic treatment of fever is nowadays a widespread practice. One of the most common methods is the administration of antipyretics. Many observational studies concern the prescription of acetaminophen. Although this practice is employed commonly in febrile patients because of the feared intrinsic harm of fever and patients discomfort, controversy over the value of fever remains.
Preclinical studies suggested that active cooling confers benefit in acute brain trauma patients, but clinical trials have been inconclusive so far.
In patients with other types of diagnosis, the actual support for the necessity of fever treatment is lacking. In fact, several clinical data suggest that fever has a beneficial effect on the outcome of many infectious diseases and mortality would be higher in hypothermic patients. Other studies show that antipyretics increase the duration of chickenpox illness in children as well as the inhibition of antibody responses in adults.
Experimental evidence shows that fever is a natural host respons against infectious diseases which causes an increase of cytokines and heat shock proteins, T-cell proliferation and the synthesis of antibodies. These processes are essential to prevent organ damage by oxidative injury in situations of stress. Fever also plays a key role in the inhibition of microbial growth. So fever could be in advantage of the host, and suppression of this natural reaction has never been proven to be safe in general.
The aim of the study is to be able to give answer to the question whether the standard use of paracetamol in infectious fever management is in the advantage of the patient or not.
Methods: Two searches were conducted throughout PubMed and Cochrane (a PICO design):
- ‘critical care or critical illness or ICU or Emergency department’ AND ‘paracetamol or acetaminophen’ and ‘fever or pyrexia’
- ‘paracetamol or acetaminophen’ AND ‘fever or pyrexia’ AND ‘treatment outcome’ (a mesh term covering efficacy and safety)
Search results: The combined search resulted in the discovery of: 1 review, 1 retrospective study, 2 randomised prospective studies and 3 prospective observational studies. All studies excluded patients with neurological illness or liver diseases.
The only conducted review with meta-analysis about antipyretics (including NSAIDs) and mortality, unfortunately could not draw a conclusion because of the heterogeneity of the studies.[1].
In a retrospective study about the efficacy of paracetamol on fever reduction two cohorts of fever episodes in SIRS patients were formed: a paracetamol group versus a group receiving no medication. The absolute temperature drop in patients who received paracetamol was quite unpredictable as well as variable but in the end significantly greater than in non treated patients (mean drop of 0.86 °C vs. 0.56 °C for treated vs. untreated) and the cooling rate was slightly more rapid (a significant temperature reduction in time of 0.20 °C/h vs. 0.13 °C/h for treated vs. untreated).[2]
Although statistically significance was reached, the differences were minor and normalisation of body temperature did not occur more often in treated patients (mean 56% vs. 43.5% treated vs untreated p = 0.352).
The only two randomised studies conducted, confirm the result of modest temperature reduction.
The first study showed a significant temperature fall of 0.5 °C in a small population (30 patients).[3]
The second randomised study (not double blinded, 82 patients) by Schulman showed that the daily mean temperature was 1.09 °C (significant) lower in the aggressive (acetaminophen when temperature > 38.5 °C, and a cooling blanket when temperature > 39.5 °C) treatment group compared to the daily mean temperature in the permissive (acetaminophen and a cooling blanket when temperature > 40 °C) group.[4] But aggressive treatment did not have an influence on the average daily maximum temperature. Taking this into account and the fact that the mean of all temperatures in the permissive group was not high at all (mean of 37.8 °C), one may question the potential benefit of prescribing paracetamol. A striking result in the aggressive group was that antibiotics were more frequently given (77.2 vs. 70.9%; aggressive versus permissive) and SIRS- score during the study was higher (while APACHE score at baseline was similar in each group). This study was stopped prematurely due to a trend noticed by interim analysis towards a higher mortality in the aggressive treatment group (p = 0.06, Fisher Exact Test). Definite conclusions may not be drawn but these findings raise the concern whether interrupting the natural actions of fever may be harmful to some patients.
The most recent and largest study is a prospective, observational multicentre study which only adds to the concern mentioned above.[5] A significantly higher mortality rate was found in non-septic patients with a fever > 39.5 °C, and in septic patients administration of NSAIDs and acetaminophen was independently associated with mortality (after multivariable adjustment, adjusted odds ratio: NSAIDs 2.61, p = 0.028; acetaminophen 2.05, p = 0.01).
Finally two other smaller studies conducted focused on the safety of paracetamol administration in critically ill patients.[6,7] Both observational studies concluded that paracetamol (or propacetamol: a bio precursor) caused a significant drop in systolic blood pressure, sometimes even requiring an intervention with noradrenaline.
Conclusion: There is evidence that paracetamol lowers temperature but the degree to which is not quite convincing and is less than assumed. Because large randomised, double blinded studies are lacking it might be premature to make firm conclusions about the causality between symptomatic reduction of infectious fever and mortality. Conversely, we can conclude that adverse effects do occur, and that there is not enough evidence that fever management is safe in all patients.
Currently underway is the HEAT trial (a New Zealand and Australian trial) in which 700 patients are being randomised between receiving 1 gr. of intravenous paracetamol or placebo, and safety and efficacy will be studied.
While awaiting the results of the HEAT trial, we advise restraint in the symptomatic treatment of fever with acetaminophen in adult patients.
