Are antipsychotics effective for the prevention of postoperative delirium?

Castro, Viviana; Guinguis, Rami; Carrasco, Marcela

doi:10.5867/medwave.2018.02.7195

Living FRIendly Summaries of the Body of Evidence using Epistemonikos (FRISBEE)

Medwave 2018 Mar-Abr;18(2):e7195 doi: 10.5867/medwave.2018.02.7195

Are antipsychotics effective for the prevention of postoperative delirium?

Viviana Castro, Rami Guinguis, Marcela Carrasco

References | Download PDF | Print | A(+) A(-) | Easy read

Abstract

INTRODUCTION
Antipsychotics have been proposed as a pharmacological alternative to prevent postoperative delirium. Nonetheless, their actual clinical benefits and harms are a matter of debate.

METHODS
To answer this question, we used Epistemonikos, the largest database of systematic reviews in health, which is maintained by screening multiple information sources, including MEDLINE, EMBASE, Cochrane, among others. We extracted data from the chosen systematic reviews, reanalyzed the data from the primary studies, conducted a meta-analysis and generated a summary of findings table using the GRADE approach.

RESULTS AND CONCLUSIONS
We identified 16 systematic reviews that included eight primary studies, all were randomized trials. We concluded the use of prophylactic antipsychotics reduces the incidence of postoperative delirium, but has no effect on the duration of hospital stay and might increase mortality.

Problem

Postoperative delirium is associated to an increase in morbidity, mortality, prolongation of hospital stay and increased costs, especially in elderly patients.

Several non-pharmacological interventions have been proposed in order to reduce the incidence of this complication, but the role of pharmacological alternatives is still controversial. It is known antipsychotics are effective in the prevention of delirium. However, their efficacy and safety in the postoperative context are not yet clearly established.

Methods

To answer the question, we used Epistemonikos, the largest database of systematic reviews in health, which is maintained by screening multiple information sources, including MEDLINE, EMBASE, Cochrane, among others, to identify systematic reviews and their included primary studies. We extracted data from the identified reviews and reanalyzed data from primary studies included in those reviews. With this information, we generated a structured summary denominated FRISBEE (Friendly Summary of Body of Evidence using Epistemonikos) using a pre-established format, which includes key messages, a summary of the body of evidence (presented as an evidence matrix in Epistemonikos), meta-analysis of the total of studies when it is possible, a summary of findings table following the GRADE approach and a table of other considerations for decision-making.

Key messages

Prophylactic antipsychotics reduce the incidence of postoperative delirium, but they have no effect on length of hospital stay in elderly patients.
Perioperative use of antipsychotics might increase mortality, but the certainty of this evidence is low.

About the body of evidence for this question

What is the evidence. See evidence matrix in Epistemonikos later	We found 16 systematic reviews [1],[2],[3],[4],[5],[6], [7],[8],[9],[10],[11],[12],[13],[14],[15],[16], that include eight primary studies [17],[18],[19],[20],[21],[22],[23],[24] answering the question of interest, all were randomized trials.
What types of patients were included*	Four trials included patients undergoing orthopedic surgery [17],[18],[20],[21], two included patients undergoing cardiovascular surgery [19],[22], two included patients undergoing gastrointestinal surgery [18],[23], and one trial mentioned the participants were not submitted to cardiovascular surgery [24]. Average age was reported in six trials [17],[20],[21],[22],[23],[24], ranging between 61 and 87 years, with an average of 74.6 years. All the trials reported the gender of included participants. The average proportion of men was 50.4%, with a range from 20.2% to 68.3%.
What types of interventions were included*	Typical (haloperidol) and atypical (risperidone, olanzapine) antipsychotics were used in the trials. Five trials used haloperidol, between 0.5 and 5 mg intravenously [17],[18],[20],[23],[24]. Two trials used risperidone 0.5 mg [20] and 1 mg [23]. One trial used olanzapine 5 mg [22]. In most of the trials the intervention was postoperative [17],[18],[19], and in two trials [20],[21] it was administered before and after surgery. All trials compared against placebo. It is important to note the trials did not make any distinction according to psychomotor activity of delirium, except for one trial [19] that applied the intervention to ‘subsyndromal delirium’.
What types of outcomes were measured	The main outcomes evaluated by the systematic reviews were: incidence of postoperative delirium, duration of delirium, days of hospitalization, length of stay in intensive care unit, severity of delirium, safety of haloperidol, and mortality.

* The information about primary studies is extracted from the systematic reviews identified, unless otherwise specified.

The main outcomes evaluated by the systematic reviews were: incidence of postoperative delirium, duration of delirium, days of hospitalization, lenght of stay in intensive care unit, severity of delirium, safety of haloperidol, and mortality.

Summary of Results

The information about the effects of antipsychotics for the prevention of postoperative delirium is based on eight randomized trials [17],[18],[19],[20],[21],[22],[23],[24] that included 1932 patients overall. Seven trials reported the incidence of delirium [17],[18],[19],[20],[21],[22],[23],[24]; four trials reported the duration of delirium [19],[20],[21],[24]; four trials reported the length of hospital stay [19],[20],[22],[24]; and only one trial reported mortality [19]. The summary of findings is the following:

The use of prophylactic antipsychotics reduces the incidence of postoperative delirium in elderly population. The certainty of the evidence is high.
It is not clear if the use of prophylactic antipsychotics reduces the duration of postoperative delirium, because the certainty of the evidence is very low.
The use of perioperative prophylactic antipsychotics has little or no effect in the length of hospital stay in elderly patients. The certainty of the evidence is high.
It is not clear whether prophylactic antipsychotics increase or reduce postoperative mortality in the elderly population, because the certainty of the evidence is low.

Other considerations for decision-making

To whom this evidence does and does not apply

The results of this summary can be applied to older adults at risk of delirium that will be submitted to orthopedic, cardiovascular or abdominal surgery.
This evidence does not apply to patients that develop delirium in the postoperative period.

About the outcomes included in this summary

The outcomes presented in the summary of findings are those considered critical for decision-making by the authors of this article, and in general agree with the outcomes selected by the systematic reviews identified.

Balance between benefits and risks, and certainty of the evidence

Even though the intervention decreases the incidence of delirium, this does not translate into a shorter length of hospital stay, which makes this benefit less relevant from a clinical point of view.
The possibility of this intervention increasing mortality, while based on low certainty evidence, counterbalance any possible benefit. It is important to consider there is a clear pathophysiological mechanism that would explain this risk, mediated by cardiovascular effects, like QT-interval prolongation. The evidence does not allow to answer if it is safe to use this intervention in patients with a normal electrocardiogram, as some experts suggest [25].

Resource considerations

Antipsychotics have a relatively low cost, but their potential increase in mortality would hamper any observed benefit.

What would patients and their doctors think about this intervention

Faced with the evidence presented in this summary, most clinicians and patients should lean against its use, or restrict it to selected cases.

Differences between this summary and other sources

The conclusion of this summary agrees with most systematic reviews, in terms of the effect on the incidence of delirium. One exception is a Cochrane review [8], which concludes prophylactic antipsychotics do not have an effect on the incidence of delirium, but on the duration and severity. However, this review is out of date, and includes only one of the eight randomized trials.
The conclusions of this summary partially agree with the American Geriatrics Society [26] guideline, that states there is not enough evidence to recommend this intervention, due to the low quality of the existing trials.

Could this evidence change in the future?

The probability that future evidence changes the conclusions in this summary is high, considering the uncertainty in relation to the risk of death.
We identified at least four ongoing trials [27],[28],[29],[30] in the International Clinical Trials Registry Platform of the World Health Organization, that could add relevant information.
The identified systematic reviews have limitations and do not include all of the trials. Also, there may exist new trials, making a new systematic review urgently needed. However, we did not identify any ongoing review in the PROSPERO database.

How we conducted this summary

Using automated and collaborative means, we compiled all the relevant evidence for the question of interest and we present it as a matrix of evidence.

Follow the link to access the interactive version: Antipsychotics for the prevention of postoperative delirium

Notes

The upper portion of the matrix of evidence will display a warning of “new evidence” if new systematic reviews are published after the publication of this summary. Even though the project considers the periodical update of these summaries, users are invited to comment in Medwave or to contact the authors through email if they find new evidence and the summary should be updated earlier.

After creating an account in Epistemonikos, users will be able to save the matrixes and to receive automated notifications any time new evidence potentially relevant for the question appears.

This article is part of the Epistemonikos Evidence Synthesis project. It is elaborated with a pre-established methodology, following rigorous methodological standards and internal peer review process. Each of these articles corresponds to a summary, denominated FRISBEE (Friendly Summary of Body of Evidence using Epistemonikos), whose main objective is to synthesize the body of evidence for a specific question, with a friendly format to clinical professionals. Its main resources are based on the evidence matrix of Epistemonikos and analysis of results using GRADE methodology. Further details of the methods for developing this FRISBEE are described here (http://dx.doi.org/10.5867/medwave.2014.06.5997)

Epistemonikos foundation is a non-for-profit organization aiming to bring information closer to health decision-makers with technology. Its main development is Epistemonikos database (www.epistemonikos.org).

Potential conflicts of interest

The authors do not have relevant interests to declare.

Esta obra de Medwave est� bajo una licencia Creative Commons Atribuci�n-NoComercial 3.0 Unported. Esta licencia permite el uso, distribuci�n y reproducci�n del art�culo en cualquier medio, siempre y cuando se otorgue el cr�dito correspondiente al autor del art�culo y al medio en que se publica, en este caso, Medwave.