
Zou et al. developed and temporally validated a prospective multicentre model combining quadriceps ultrasound and clinical variables to predict intensive care unit-acquired weakness. We focus on its intended clinical use: outcome verification depends on patients later becoming assessable by volitional strength testing, whereas some predictors are collected throughout the first 24 h. The later cohort appropriately evaluates performance across time within the participating centre network. Clarifying the verification population, prediction landmark, threshold selection, and imputation pipeline would support reproducible use; subsequent evaluation in independent hospitals would address geographic transportability.
Jouffroy et al. reported lower 30-day mortality associated with prehospital antibiotics and suggested that the association varied by presumed septic shock origin. However, treatment assignment and covariate timing require clarification; formal treatment-by-origin interaction testing was not reported; and several effect measures and subgroup counts are inconsistent. Resolving these issues is central to interpreting the findings. Confirmed digestive diagnoses, antimicrobial appropriateness, and source-control timing were not described in the published report. These factors warrant future investigation, whereas current evidence does not establish that infection origin modifies the treatment effect.
Frailty has been associated with adverse outcomes during mechanical ventilation (MV). We evaluated the association between frailty severity, assessed using the Clinical Frailty Scale (CFS), and survival to intensive care unit (ICU) discharge without ongoing invasive MV. We analyzed adults with sepsis enrolled in the prospective, multicenter Korean Sepsis Alliance registry between September 2019 and December 2023 who received invasive MV on ICU day 1. Patients with CFS 9 were excluded. The primary outcome was survival to ICU discharge without ongoing invasive MV. Frailty was defined as CFS 5–8 versus CFS 1–4 and was examined across increasing CFS severity using categorical and restricted cubic spline analyses. Mixed-effects logistic regression with a participating-center random intercept was adjusted for age, sex, Charlson comorbidity index, ICU day 1 Sequential Organ Failure Assessment score, and infection site. Among 2,204 patients, 1,458 (66.2%) had CFS 5–8. The primary outcome was achieved by 1,120 patients (50.8%); crude rates were 52.9% in patients with CFS 1–4 and 49.7% in those with CFS 5–8 ( P = 0.152). Frailty was associated with lower adjusted odds of the primary outcome (adjusted odds ratio [OR], 0.73; 95% CI 0.59–0.90). Compared with CFS 1–4, the estimate for CFS 5 was close to the null (OR, 1.08; 95% CI 0.78–1.49), while the odds were lower for CFS 6, 7, and 8 (ORs, 0.68, 0.73, and 0.59, respectively; global P < 0.001). In spline analysis, the adjusted probability of the primary outcome decreased from 58.6% at CFS 1 to 43.3% at CFS 8 ( P overall < 0.001; P for nonlinearity = 0.55). In pairwise analyses of ICU discharge status, frailty was associated with higher odds of survival with ongoing invasive MV versus survival without ongoing invasive MV (adjusted OR, 1.94; 95% CI 1.26–2.98); the corresponding estimate for ICU death versus survival without ongoing invasive MV was 1.22 (95% CI 0.97–1.53). Among invasively ventilated patients with sepsis, baseline frailty was associated with lower adjusted odds of survival to ICU discharge without ongoing invasive MV. These findings support considering frailty severity when discussing recovery and care needs, but do not establish a treatment threshold or a causal effect of frailty.
Autonomic dysfunction and reduced signal complexity (SC) may occur after aneurysmal subarachnoid haemorrhage (aSAH) and reflect impaired physiological adaptability. However, the relationships with cerebral physiology, delayed cerebral ischaemia (DCI), and long-term functional outcome remain incompletely understood. This study explores heart rate variability (HRV), baroreflex sensitivity (BRS), and systemic and cerebral SC in aSAH patients during neurointensive care (NIC). This observational, single-centre cohort study included 690 aSAH patients with intracranial pressure (ICP) monitoring during the first 10 days post-ictus, treated at Uppsala University Hospital, Sweden (2008–2024). HRV, BRS, and multiscale entropy derived SC of each of heart rate (HR), mean arterial pressure (MAP), ICP, ICP pulse amplitude (AMP), and cerebral perfusion pressure (CPP) were analysed using high-frequency monitoring data. Associations with NIC physiology (ICP, CPP, pressure reactivity index [PRx]), DCI, and 12-month Glasgow Outcome Scale-Extended (GOSE) were assessed using multivariable logistic regression, two-variable outcome heatmaps, and Spearman's correlations. DCI was defined as new focal neurological deficits or declining consciousness after exclusion of other aetiologies. In multivariable regression adjusted for demographic, clinical, radiological, and physiological predictors, reduced SC of each of HR, MAP, ICP, AMP, and CPP independently correlated with worse functional outcome, while reduced cerebral SC (of ICP, AMP, and CPP) the first 3 days post-ictus was independently associated with DCI. Reduced SC generally correlated with higher ICP, lower CPP, and higher PRx. Reduced SC combined with high ICP and PRx and low CPP was particularly associated with poor recovery in outcome heatmaps. Reduced SC was independently associated with DCI and worse functional outcome after aSAH. SC may serve as a prognostic marker and a bedside indicator of DCI and vulnerability to secondary insults, with potential to guide more individualised NIC. Prospective, multi-centre studies are needed to validate these findings.
We respond to the comment on longitudinal lactate indices after traumatic brain injury. Full-stay time-weighted average lactate characterizes longitudinal lactate exposure over the intensive care unit stay but is unavailable at an earlier prediction time. Additional post hoc 72 h landmark analyses showed modest discrimination of subsequent 14-day intensive care mortality among eligible patients. Paired comparisons demonstrated similar full-stay discrimination for time-weighted and arithmetic mean lactate. Both analyses yielded similar estimates in matched and full eligible cohorts. These findings support the prognostic relevance of longitudinal lactate measurements; however, prospective validation at prespecified landmarks is necessary before clinical implementation.