Delirium, Cerebral Perfusion, and High-Frequency Vital-Sign Monitoring in the Critically Ill. The CONFOCAL-2 Feasibility Study
Jasmine M Khan,Michael D Wood,Kevin F H Lee,David Maslove,John Muscedere,Shane W English,Ian Ball,Marat Slessarev,J Gordon Boyd
DOI: https://doi.org/10.1513/AnnalsATS.202002-093OC
Abstract:Rationale: Studies suggest that reduced cerebral perfusion may contribute to delirium development in the intensive care unit (ICU). However, evidence is limited because of factors including small sample size and limited inclusion of covariates.Objectives: To assess the feasibility of a multicenter prospective observational study using a multimodal data collection platform. Feasibility was assessed by enrollment, data-capture, and follow-up rates. The full study will aim to assess the association between noninvasively derived surrogate markers of cerebral perfusion, delirium development, and long-term cognitive outcomes in critically ill patients.Methods: Adult patients in the ICU were enrolled if they had shock and/or respiratory failure requiring invasive mechanical ventilation for >24 hours. For the first 72 hours, a near-infrared spectroscopic sensor was placed on the forehead to continuously monitor regional cerebral oxygenation (rSo2) and high-frequency (1 Hz) vital signs were concurrently captured via an arterial line. Cerebral perfusion was estimated using three variables, including mean rSo2, duration of disturbed autoregulation, and time/magnitude away from optimal mean arterial pressure (MAP). Patients were screened for delirium in the ICU and ward daily for up to 30 days. Cognitive function was assessed 3 and 12 months after ICU admission to identify cognitive impairment.Results: Fifty-nine patients were enrolled across four sites in 1 year. Data-capture rates varied across modalities but exceeded 80% for rSo2, blood gas, and delirium data capture. Vital-sign capture and 3-month follow-up rates were lower at 53% and 55%, respectively. Eighty-three percent (49 of 59) of patients experienced delirium, with a median severity of 0.56 in the ICU. Mean physiological (±standard deviation) values were: rSo2 (70.4% ± 7.0%), heart rate (83.9 ± 16.45 beats/min), MAP (76.4 ± 12.8 mm Hg), peripheral oxygenation saturation (96.5% ± 2.1%), proportion of recording time spent with disturbed autoregulation (10.1% ± 7.3%) and proportion of area under the curve outside optimal MAP (39.6% ± 22.4%). Thirty-two (54%) individuals had cerebral autoregulation curves where a targeted optimal MAP was identified. Barriers to data collection included missing vital-sign data and low follow-up rates.Conclusions: Given our current protocol, a multicenter study examining the association between cerebral oxygenation, delirium, and long-term cognitive impairment is not feasible. However, by performing an early assessment of feasibility, we identified strategies to increase capture rates to ensure success as the study begins the next phase of study recruitment.Clinical trial registered with clinicaltrials.gov (NCT03141619).