Regional Cerebral Oxygen Saturation and Risk of Delirium: A Systematic Review and Meta-Analysis.

Background: Delirium onset is associated with increased comorbidity and mortality. Identifying reliable delirium biomarkers remains challenging. Regional cerebral oxygen saturation (rSO2) offers an objective, easily obtainable measure suitable for hospital monitoring. Objective: We aimed to analyse the relationship between regional cerebral oxygen saturation (rSO2) values obtained by near-infrared spectroscopy (NIRS) and the subsequent development of delirium. Methods: Studies eligible for inclusion in our systematic review evaluated rSO2 values obtained by NIRS or a used a similar method to study hospitalised patients aged 18 years or older, some of whom subsequently developed delirium. We searched MEDLINE, Scopus and Web of Science without restrictions to 24 March 2024. Two review authors independently assessed the methodological quality of the included studies using Joanna Briggs Institute Critical Appraisal tools. Using a random-effects model in RevMan v 5.4.0 (Cochrane Collaboration, Oxford, UK), we analysed baseline and minimum rSO2 values. Results were presented as means and mean differences (MDs) with their 95% confidence intervals (CIs). We followed PRISMA guidelines and registered our review protocol in PROSPERO (CRD42024523573). Results (or Findings): We included 22 studies (20 in the meta-analysis) published between 2009 and 2024 and involving 5757 participants. The delirium group had a lower mean baseline rSO2 value (62.47%, 95% CI 58.40 to 66.55) compared with the non-delirium group (64.24%, 95% CI 61.33 to 67.15). Meta-analysis of effect estimates confirmed this result (MD -2.92%, 95% CI -4.38 to -1.47). The MD between the delirium and non-delirium group was larger among patients assessed with the INVOS device and patients who underwent cardiac surgery. Studies that analysed baseline values according to sensor location showed a larger MD in rSO2 values obtained via a right-sided sensor. Conclusions: Our results show lower baseline and minimum rSO2 in hospitalised patients who subsequently developed delirium. The difference varies according to the type of surgery and type of NIRS monitor.