The effectiveness and harms of pharmacological interventions for the treatment of delirium in adults admitted into the intensive care unit after cardiac surgery: a systematic review

Abstract

Background: Patients who undergo cardiac surgery are at high risk of delirium (incidence: 50-90%), increasing the risk of death and adversely affecting recovery. Clinical interventional trials have been conducted to prevent and treat postoperative delirium pharmacologically including antipsychotics and sedatives. These trials have provided some evidence about efficacy and influenced clinical decision making. However, much reporting is incomplete and provides biased assessments of efficacy; benefits are emphasised while harms are inadequately reported. The purpose of this study was to undertake a systematic review using the Joanna Briggs Institute (JBI) methodology that aimed at identifying and synthesising the best available evidence about the effectiveness and harms of pharmacological interventions in the treatment of delirium in adult intensive care patients after cardiac surgery. Inclusion Criteria: Types of participants: Participants were ≥ 16 years, any gender or ethnicity, who were treated postoperatively in a cardiothoracic intensive care unit (ICU) following cardiac surgery and identified as having delirium. Types of interventions: Any pharmacological intervention for the treatment of delirium was included, regardless of drug classification, dosage or frequency of administration. Types of comparators: Studies that compared any pharmacological interventions for the treatment of delirium in patients who were admitted in the ICU after cardiac surgery. No limitations were placed on drug classification, dosage of the medications or frequency of administration. Types of outcomes: This systematic review examined eleven primary and five secondary outcomes of interest. The primary outcomes of interest included: mortality, duration and severity of delirium, use of physical restraints, quality of life, family members satisfaction with delirium management, duration/severity of the aggressive episode, associated falls, severity of accidental self-harm, pharmacological harms, and harms related to over-sedation. Types of studies: Randomised controlled trials (RCTs) were considered first and in their absence, non-RCTs and quasi-experimental would have been considered followed by analytical observational studies. Search Strategy: A comprehensive search was conducted across seven databases, three clinical trial registers and a database for dissertations and theses as well as a hand search for published primary studies. Methodological quality: Two reviewers assessed the methodological quality of the included studies using standardised critical appraisal instruments from JBI and McMaster University. Data extraction: Quantitative data were extracted using the standardised JBI data extraction tool. A meta-analysis was not performed as there was too much clinical and methodological heterogeneity in the included studies. Results have been presented in a narrative form. Standard GRADE (Grading of Recommendations, Assessment, Development, and Evaluation) evidence assessment of outcomes has been reported. Results: Three RCTs investigating morphine vs haloperidol (n=53), ondansetron vs haloperidol (n=72) and dexmedetomidine vs midazolam (n=80) were included. Overall the methodological quality of these studies was found to be low. There is currently insufficient evidence to confirm or refute the effectiveness of morphine compared with haloperidol, ondansetron compared with haloperidol or dexmedetomidine compared with midazolam for reducing the duration or severity of hyperactive delirium in the postoperative cardiac surgical patient treated in the ICU. Additionally, this review found reporting of harms to be inadequate for all three studies and did not meet the required standards for harms reporting. Conclusions: This review was unable to draw any valid conclusions regarding the effectiveness of the included pharmacological interventions in treating delirium after cardiac surgery. This is due to the low number of studies, the poor methodological quality in conducting and reporting and the heterogeneity between the studies. Implications for practice: There is insufficient evidence to support the use of morphine, ondansetron or dexmedetomidine as effective pharmacological agents in treating delirium. It is imperative that clinicians remain vigilant to the known indications, contraindications and harms of the pharmacological agents that are being administered and to understand the implications of such drugs on cardiac performance in the initial postoperative recovery phase after cardiac surgery