Daily Sepsis Research Analysis
An updated Bayesian network meta-analysis of randomized trials suggests PMX and HA330 hemoperfusion may reduce mortality in severe septic shock, while convective therapies could increase mortality. A large nationwide cohort developed and validated the STeP score to predict early tracheostomy in ventilated sepsis patients, and a MIMIC-IV study shows the Advanced Lung cancer Inflammation index (ALI) independently predicts 28- and 90-day mortality in sepsis-associated AKI.
Summary
An updated Bayesian network meta-analysis of randomized trials suggests PMX and HA330 hemoperfusion may reduce mortality in severe septic shock, while convective therapies could increase mortality. A large nationwide cohort developed and validated the STeP score to predict early tracheostomy in ventilated sepsis patients, and a MIMIC-IV study shows the Advanced Lung cancer Inflammation index (ALI) independently predicts 28- and 90-day mortality in sepsis-associated AKI.
Research Themes
- Extracorporeal blood purification strategies in severe septic shock
- Early risk stratification and decision support in ICU sepsis care
- Inflammation–nutrition composite indices predicting outcomes in SA-AKI
Selected Articles
1. Effects on mortality of blood purification techniques in severe septic shock patients. An updated Bayesian network meta-analysis of randomized controlled trials.
Across 31 RCTs (n=2678), a Bayesian NMA found that PMX and HA330 hemoperfusion significantly reduced mortality versus standard care in severe septic shock, while convective therapies were associated with higher mortality. Findings were supported by SUCRA rankings, with risk of bias assessed per Cochrane guidance and protocol preregistered.
Impact: This synthesis directly compares blood purification modalities and identifies devices linked to mortality reduction, challenging prevailing uncertainty and informing trial design and clinical choices.
Clinical Implications: Consider PMX or HA330 hemoperfusion in severe septic shock where available and expertise exists, while exercising caution with convective therapies until confirmatory head-to-head RCTs clarify benefit–risk.
Key Findings
- PMX and HA330 hemoperfusion significantly reduced mortality versus standard treatment in severe septic shock.
- Convective therapies were associated with higher mortality compared with standard care across trials.
- Bayesian NMA of 31 RCTs (n=2678) with SUCRA rankings and PROSPERO registration supports robustness.
Methodological Strengths
- Bayesian network meta-analysis restricted to RCTs with Cochrane risk-of-bias assessment
- Protocol preregistration and SUCRA-based ranking across 15 interventions
Limitations
- Heterogeneity in intervention protocols and outcome timeframes across trials
- Potential publication bias and limited head-to-head comparisons for some devices
Future Directions: Conduct adequately powered, head-to-head RCTs comparing PMX, HA330, and standard care with harmonized endpoints and patient phenotyping to confirm mortality effects and identify responders.
2. Development and validation of the STeP score for predicting tracheostomy in patients with sepsis using a nationwide ICU database: a retrospective observational study.
Using JIPAD data (n=7357), the STeP model with 8 LASSO-selected variables predicted tracheostomy within 14 days of ICU admission in sepsis with AUC 0.74 (validation). The simplified 0–17 point STeP score achieved AUC 0.73 and stratified patients into low, moderate, and high risk with tracheostomy rates of 4.0%, 13.6%, and 27.1%.
Impact: Provides a validated, early bedside tool to anticipate tracheostomy needs in ventilated sepsis patients, supporting communication and resource planning.
Clinical Implications: Incorporate the STeP score within 24 hours of ICU admission to inform shared decision-making about tracheostomy timing, anticipating prolonged ventilation and resource needs.
Key Findings
- STeP model (8 variables) achieved AUC 0.74 in validation; simplified 0–17 point score achieved AUC 0.73.
- Risk strata: ≤2 points (4.0%), 3–6 points (13.6%), ≥7 points (27.1%) tracheostomy within 14 days.
- Large nationwide cohort of ventilated sepsis patients from JIPAD (n=7357) with temporal validation and web calculator.
Methodological Strengths
- Large, nationwide cohort with temporally independent validation
- LASSO feature selection with tenfold cross-validation and simplified bedside score
Limitations
- Retrospective design with potential residual confounding
- Generalizability beyond Japan and to viral pneumonia excluded populations is uncertain
Future Directions: External validation across diverse health systems and prospective impact studies to assess whether STeP-guided planning improves outcomes and resource utilization.
3. Prognostic value of the advanced lung cancer inflammation index for 28 day mortality in sepsis associated acute kidney injury.
In 5,565 SA-AKI patients from MIMIC-IV, higher ALI was independently associated with lower 28- and 90-day mortality, with the highest quartile showing significantly reduced risk (e.g., HR≈0.72 for 28-day mortality). ALI provided discriminatory capability and incremental value versus other indices across multiple sensitivity analyses.
Impact: Introduces a simple composite inflammation–nutrition index for robust mortality risk stratification in SA-AKI, facilitating pragmatic prognostication using routinely available data.
Clinical Implications: ALI can be integrated into ICU risk stratification for SA-AKI to identify high-risk patients and tailor monitoring and supportive care, pending external validation.
Key Findings
- Among 5,565 SA-AKI patients, higher ALI quartiles were associated with significantly lower 28- and 90-day mortality (e.g., HR≈0.72 for top quartile at 28 days).
- Multivariable Cox, Kaplan–Meier, and spline analyses demonstrated consistent inverse associations between ALI and mortality.
- ROC analyses suggested discriminatory ability and incremental value of ALI compared with other indices.
Methodological Strengths
- Large sample from a high-quality ICU database with extensive multivariable and sensitivity analyses
- Use of complementary statistical approaches (Cox, KM, restricted cubic splines, ROC comparisons)
Limitations
- Retrospective single-database analysis with potential residual confounding
- External validation and clinical impact assessment are needed before routine use
Future Directions: Prospective external validation and integration of ALI into dynamic prediction models to assess whether ALI-guided management improves SA-AKI outcomes.