Daily Anesthesiology Research Analysis
Analyzed 140 papers and selected 3 impactful papers.
Summary
Three perioperative and critical care anesthesia studies stand out today: a multicenter prospective cohort links preoperative ACE inhibitor/ARB use to higher rates of moderate-to-severe cardiac surgery–associated AKI; an obstetric anesthesia study disentangles epidural-related maternal fever from occult infection, showing no adverse neonatal outcomes after excluding histologic chorioamnionitis; and a multicenter machine learning model predicts stage ≥II histologic chorioamnionitis at the onset of epidural-related fever with strong external validation.
Research Themes
- Perioperative medication management and renal risk
- Obstetric anesthesia: differentiating fever etiologies and neonatal safety
- Machine learning for intrapartum infection risk stratification
Selected Articles
1. Preoperative angiotensin-converting enzyme inhibitor or angiotensin receptor blocker use and acute kidney injury after high-risk cardiac surgery: a multicenter prospective cohort study.
In a 14-center prospective cohort of high-risk cardiac surgery, preoperative ACEI/ARB use was independently associated with higher incidence of stage 2–3 AKI (adjusted OR 2.79), with a number needed to harm of approximately 5. No significant differences were observed in in-hospital mortality, and withholding versus continuation strategies did not significantly alter outcomes.
Impact: This prospective multicenter analysis addresses a persistent perioperative controversy with robust causal-inference techniques and suggests actionable risk associated with ACEI/ARB exposure in a high-stakes population.
Clinical Implications: For high-risk cardiac surgery, teams should reassess blanket continuation of ACEI/ARB therapy and consider individualized risk–benefit discussions focused on AKI prevention strategies and renal monitoring.
Key Findings
- Stage 2–3 AKI occurred more frequently with preoperative ACEI/ARB use (38.8% vs 26.1%).
- Adjusted odds ratio for stage 2–3 AKI with ACEI/ARB was 2.79 (95% CI 1.47–5.30).
- Number needed to harm was approximately 5.3; results were robust across sensitivity analyses (E-value 2.73; AIPW consistent).
- No significant difference in in-hospital mortality; withholding vs continuation strategies showed no clear outcome differences.
Methodological Strengths
- Prospective multicenter cohort with predefined high-risk criteria (Cleveland Clinic Score ≥4).
- Robust causal-inference methods including augmented inverse probability weighting and E-value analysis.
Limitations
- Observational design cannot fully eliminate residual confounding.
- Mortality and withholding-versus-continuation comparisons may have been underpowered.
Future Directions: Randomized trials or pragmatic cluster trials testing standardized perioperative ACEI/ARB strategies with renal-protective bundles are warranted.
BACKGROUND: The impact of preoperative Angiotensin-Converting Enzyme Inhibitors or Angiotensin Receptor Blockers (ACEI/ARB) on cardiac surgery-associated Acute Kidney Injury (AKI) remains elusive. We sought to evaluate the association of ACEI/ARB use with stage 2 or 3 AKI and in-hospital mortality, while evaluating whether preoperative withholding versus continuation strategies modulate clinical outcomes. METHODS: Multicenter prospective cohort study involving 14 Spanish and British hospitals. The study population comprised high-risk cardiac surgery patients (Cleveland Clinic Score ≥ 4). RESULTS: Among 249 patients (mean age: 69.5-years; 39% women; 53.8% ACEI/ARB users), the overall incidence of stage 2 or 3 AKI was 32.9%. ACEI/ARB use was associated with a significantly higher likelihood of stage 2 to 3 AKI (38.8% vs. 26.1%; adjusted OR = 2.79; 95% CI 1.47-5.30; p = 0.002). The adjusted risk difference was 0.188 (95% CI 0.079-0.298; p = 0.001), yielding a number needed to harm of 5.3. Sensitivity analyses confirmed the robustness of these findings, with an E-value of 2.73 (lower limit: 1.72) and consistent results using augmented inverse probability weighting (average treatment effect: 0.191; 95% CI 0.081-0.300; p = 0.001). No statistically significant associations were observed in in-hospital mortality or the impact of preoperative withholding versus continuation strategies on clinical outcomes. CONCLUSION: This study suggests preoperative ACEI/ARB use may be associated with an increased incidence of stage 2 or 3 AKI. While differences in in-hospital mortality and preoperative withholding versus continuation strategies did not reach statistical significance, these findings reveal clinical trends that warrant further investigation.
2. Machine learning for prediction of histologic chorioamnionitis (stage ≥II) in parturients receiving labor analgesia: a retrospective multicentre cohort study.
Using multicenter data from 2,715 laboring parturients with placental pathology, a six-feature random forest model predicted stage ≥II histologic chorioamnionitis at the onset of epidural-related maternal fever with AUC 0.945 (internal) and 0.849 (external). SHAP interpretation highlighted BMI as the leading contributor among retained features.
Impact: Provides an interpretable, externally validated tool for real-time risk stratification at the bedside when diagnostic uncertainty is high and management decisions (antibiotics, monitoring, delivery timing) are time-sensitive.
Clinical Implications: At ERMF onset, clinicians could apply the model to prioritize cultures, early antibiotics, or enhanced fetal surveillance for high-risk patients while avoiding overtreatment in low-risk cases.
Key Findings
- Random forest using six EHR features achieved AUC 0.945 (internal) and 0.849 (external) for predicting HCA stage ≥II.
- Sensitivity and specificity were balanced in external validation (0.957 and 0.867, respectively).
- SHAP analysis identified BMI as the top contributor among the retained predictors.
- Model designed for use at ERMF onset, before delivery and histopathologic confirmation.
Methodological Strengths
- Multicenter dataset with independent external validation.
- Model interpretability via SHAP supporting clinical reasoning.
Limitations
- Retrospective design with potential selection bias tied to availability of placental pathology.
- Generalizability may be limited outside institutions with similar EHR variables and workflows.
Future Directions: Prospective impact studies integrating the model into intrapartum workflows to assess clinical utility, calibration drift, and antibiotic stewardship outcomes.
BACKGROUND: Histological chorioamnionitis (HCA) is a serious pregnancy complication, but early diagnosis is challenging, especially in parturients receiving labor analgesia, where identification is even more difficult. Therefore, we developed and validated a machine learning model for early prediction of HCA (stage ≥II) intended for use at the time of epidural-related maternal fever (ERMF) onset, before delivery and confirmatory histopathology. METHODS: This study utilized a multicenter retrospective cohort dataset, including parturients receiving labor analgesia and completed placental pathological examination. Candidate features were extracted from electronic health records (EHR), followed by data preprocessing procedures such as addressing multicollinearity, Z-score standardization, and minority class weighting. Subsequently, three machine learning models-logistic regression (LR), random forest (RF), and extreme gradient boosting (XGBoost)-were developed and compared, with their performance evaluated through internal and independent external validation. Finally, the SHapley Additive exPlanations (SHAP) method was used to interpret the best-performing model, aiming to clarify the predictive contribution of each feature. RESULTS: A total of 2,715 parturients were included in this study, of which 676 cases (24.9%) were diagnosed with HCA (stage ≥II). After feature selection, the model retained 6 key features. The RF model exhibited the best performance, achieving an Area Under the Curve(AUC) of 0.945 in the internal validation set and an AUC of 0.849 in the independent external validation set, demonstrating balanced sensitivity (0.957) and specificity (0.867). SHAP analysis indicated that body mass index (BMI) was the most important predictive factor. CONCLUSIONS: The RF model performed the best in predicting the risk of HCA (stage ≥II) in parturients receiving labor analgesia. SHAP analysis further revealed that BMI was the most important predictive factor in the model.
3. Impact of Epidural-Related Maternal Fever on Neonatal Outcomes: A Single-Center Retrospective Case-Control Study Excluding Confirmed Histological Chorioamnionitis.
After excluding histological chorioamnionitis and balancing covariates via propensity matching (186 pairs), intrapartum maternal fever during labor analgesia was linked to longer labor stages and increased fetal tachycardia but not to adverse neonatal acid-base status, Apgar scores, NICU admission, or later developmental milestones.
Impact: By rigorously excluding occult infection, this study clarifies the neonatal safety profile of isolated ERMF and may reduce unnecessary neonatal sepsis workups and maternal/infant antibiotic exposure.
Clinical Implications: In the absence of histologic chorioamnionitis and clinical infection criteria, neonatal outcomes are reassuring despite ERMF; clinicians can focus on intrapartum monitoring while avoiding reflexive neonatal sepsis evaluations.
Key Findings
- Propensity-matched analysis of 186 pairs showed ERMF associated with longer labor and higher fetal tachycardia.
- No association between ERMF and adverse neonatal outcomes (umbilical artery pH <7.2, low Apgar, NICU admission).
- Long-term development assessed by standardized questionnaire was within normal range in both groups.
Methodological Strengths
- Explicit exclusion of histologic chorioamnionitis to reduce occult infection confounding.
- Propensity score matching achieving standardized mean differences <0.1 for balanced comparison.
Limitations
- Single-center retrospective design limits generalizability.
- Placental pathology performed based on predefined risk criteria, not universally, introducing selection bias.
Future Directions: Prospective multicenter studies standardizing placental pathology and neonatal follow-up are needed to confirm safety and refine fever management algorithms.
BACKGROUND: Maternal fever during labor analgesia may arise from infectious causes (such as chorioamnionitis) or noninfectious causes (such as epidural-related maternal fever [ERMF]). While chorioamnionitis is associated with neonatal outcomes, the impact of isolated ERMF remains controversial. This is due, in part, to the potential for occult intrauterine infection, which may not be clinically apparent during labor. Consequently, inadequate consideration of histological chorioamnionitis has limited the scope and validity of previous studies. This study aimed to evaluate the effects of ERMF on neonatal outcomes by excluding confirmed histological chorioamnionitis and minimizing suspected infection using predefined clinical criteria. METHODS: This retrospective study included women with singleton term deliveries under labor analgesia between January 2017 and July 2023. Cases with fetal anomalies or growth restriction were excluded. Placental pathological examination was performed when any of the predefined risk-based criteria were met, irrespective of maternal fever: clinical suspicion of chorioamnionitis, prolonged rupture of membranes, and neonatal asphyxia. Among febrile cases, short-term neonatal outcomes were first compared between those with and without histological chorioamnionitis. Subsequently, outcomes were compared between mothers with intrapartum fever (≥38 °C) and those without fever using propensity score matching. Long-term infant development was assessed using a Maternal and Child Health Handbook-based questionnaire. RESULTS: Overall, 186 matched pairs were included. All matched covariate standardized mean differences were <0.1, confirming acceptable balance. Compared with the nonfever group, the fever group had longer mean ± standard deviation durations from rupture of membranes to delivery (12.9 ± 9.4 vs 8.7 ± 10.9 hours; difference 4.2 hours; 95% confidence interval [CI], 2.17-6.33), first stage of labor (13.2 ± 6.7 vs 9.6 ± 6.3 hours; difference 3.6 hours; 95% CI, 2.32-4.99), and duration of labor analgesia (14.9 ± 9.6 vs 8.6 ± 6.7 hours; difference 6.3 hours; 95% CI, 4.62-8.01), along with a higher incidence of fetal tachycardia (36.0 vs 10.7%; absolute risk difference 25.3%; 95% CI, 12.3-37.2). In contrast, maternal fever was not associated with adverse neonatal outcomes, including umbilical artery pH <7.2, Apgar score <7 at 1 and 5 minutes, or neonatal intensive care unit admission. Growth and developmental milestones assessed at long-term follow-up using the Maternal and Child Health Handbook questionnaire were within the normal range in both groups. CONCLUSIONS: Following exclusion of histological chorioamnionitis, ERMF was associated with prolonged labor and fetal tachycardia but was not associated with adverse short-term neonatal outcomes or impaired long-term development.