Shinozaki, Gen

Abstract Delirium is a common yet underdiagnosed medical emergency involving acute deficits in cognition and attention. It affects over 20% of hospitalized patients and up to 80% of ICU patients, with increased prevalence in older adults. Delirium is associated with increased mortality, prolonged hospitalization, long-term cognitive decline, and substantial healthcare costs. However, current screening tools are subjective and difficult to implement consistently. Traditional EEGs show diagnostic utility; however, they are costly and require specialized interpretation. We conducted a prospective pilot study evaluating a point-of-care, single-channel EEG device for delirium detection in hospitalized older adults. EEG spectra were utilized to calculate a ‘bispectral EEG’ (BSEEG) score reflecting the ratio of slow to fast brain waves. Twenty-two postoperative patients (mean age 75; 59.09% female) were assessed daily for up to seven days using both the BSEEG and the gold standard 3-Minute Diagnostic Interview for Confusion Assessment Method (3D-CAM). ROC analysis was utilized to determine an optimal threshold BSEEG score to indicate delirium positivity. BSEEG scores were significantly higher in the 3D-CAM-positive group than in the 3D-CAM-negative group (1.60 and 1.45, respectively; p < 0.05). A BSEEG threshold of 1.5 demonstrated 77.78% sensitivity and 56.10% specificity for delirium. In contrast, route clinical screenings (Delirium Triage Screen and Brief CAM) showed 0.00% sensitivity and 96.30% specificity, and failed to detect all true delirium cases. Although further validation in larger cohorts is needed, these findings suggest that the point-of-care single-channel EEG is a promising, non-invasive screening tool for delirium screening, particularly in fast-paced clinical settings.

Recently, a FDA Safety Communication warned of a dose-dependent risk for QTc prolongation with citalopram, which is metabolized by CYP2C19 of the cytochrome P450 system. We investigate associations between citalopram and escitalopram dose, serum concentration, CYP2C19 phenotype, and QTc. We undertook a retrospective chart review of citalopram or escitalopram patients with the inclusion criteria of consistent medication dose, CYP2C19 phenotype (extensive metabolizers [EM], intermediate metabolizers [IM], poor metabolizers [PM]), and QTc interval on ECG. We further identified 42 citalopram users with citalopram serum concentration measurements and ECG. Regression and one-way ANOVA were used to examine the relationship between citalopram dose, citalopram serum concentration, CYP2C19 phenotype, and QTc interval. Of 75 citalopram patients, the EM group had significantly shorter QTc intervals than a combined IM+PM group (427.1±23.6 ms vs. 440.1±26.6 ms, one-tailed t-test, p=0.029). In the 80 escitalopram cohort, there was no significant difference in QTc between phenotype groups. There was no statistical correlation between citalopram ( p=0.62) or escitalopram ( p=0.30) dose and QTc. QTc was not associated with citalopram serum level ( p=0.45). In contrast to the FDA warning, this study found no association between citalopram/escitalopram dose and QTc. However, PM of the drug tended to have longer QTc intervals. Our findings suggest cytochrome P450 genotyping in select patients may be helpful to guide medication optimization while limiting harmful effects.