Ashritha A. Udupa
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Ashritha A. Udupa has a total of 17 co-authors. Here are the top co-authors by number of shared publications:
- Shruthi Rao
Department of Critical Care Medicine, Kasturba Medical College, Manipal Academy of Higher Education, Manipal, India
5 shared publications - Souvik Chaudhuri
Department of Critical Care, Kasturba Medical College, Manipal Academy of Higher Education, Manipal, Karnataka, India
5 shared publications - Thejesh Srinivas
Department of Critical Care Medicine, Kasturba Medical College, Manipal Academy of Higher Education, Manipal, India
5 shared publications - Shwethapriya Rao
Department of Critical Care, Kasturba Medical College, Manipal Academy of Higher Education , Manipal
4 shared publications - Danavath Nagendra
Department of Critical Care Medicine, Kasturba Medical College, Manipal Academy of Higher Education, Manipal, India
3 shared publications - Gagana Hanumaiah
Department of Obstetrics and Gynaecology, Kasturba Medical College, Manipal Academy of Higher Education, Manipal, Karnataka, India
3 shared publications
Latest Publications
- Open Access
- 6 August 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Interpretable prediction of mortality in zinc phosphide poisoning using combined statistical and machine learning models: a retrospective observational study
Zinc phosphide poisoning is associated with high mortality due to rapidly evolving multi-organ dysfunction, yet reliable early risk stratification remains limited. This retrospective observational study evaluated the feasibility of combining conventional statistical modelling with interpretable machine learning (ML) for prediction of in-hospital mortality. A total of 290 adults with confirmed zinc phosphide poisoning admitted to a tertiary care hospital between January 2016 and November 2022 were included. In-hospital mortality occurred in 88 patients (30.3%). Demographic, physiological, and laboratory variables recorded within the first 48 h of hospitalization were analysed. Multivariable logistic regression was performed to identify independent predictors of mortality. Subsequently, ML models (logistic regression, Random Forest, and Extreme Gradient Boosting [XGBoost]) were developed using repeated stratified cross-validation, and model discrimination was assessed using the area under the receiver operating characteristic curve (AUROC), calibration analysis, and bootstrap-derived confidence intervals. Explainable artificial intelligence using SHapley Additive exPlanations (SHAP) was applied to the best-performing model. Serum lactate remained the only independent predictor of in-hospital mortality in the multivariable logistic regression analysis. Among the ML models, XGBoost achieved the highest numerical AUROC (0.745; 95% confidence interval, 0.680–0.807), although differences between models (DeLong’s test) were not statistically significant. SHAP analysis identified bicarbonate, lactate, potassium, creatinine, and activated partial thromboplastin time as the largest contributors to the mortality predictions generated by the optimized XGBoost model. These findings suggest that interpretable ML may complement conventional regression by providing transparent, individualized mortality prediction while preserving clinical interpretability. However, these exploratory findings require prospective multicentre validation before routine clinical implementation.
- 6 August 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Open AccessInterpretable prediction of mortality in zinc phosphide poisoning using combined statistical and machine learning models: a retrospective observational study
Zinc phosphide poisoning is associated with high mortality due to rapidly evolving multi-organ dysfunction, yet reliable early risk stratification remains limited. This retrospective observational study evaluated the feasibility of combining conventional statistical modelling with interpretable machine learning (ML) for prediction of in-hospital mortality. A total of 290 adults with confirmed zinc phosphide poisoning admitted to a tertiary care hospital between January 2016 and November 2022 were included. In-hospital mortality occurred in 88 patients (30.3%). Demographic, physiological, and laboratory variables recorded within the first 48 h of hospitalization were analysed. Multivariable logistic regression was performed to identify independent predictors of mortality. Subsequently, ML models (logistic regression, Random Forest, and Extreme Gradient Boosting [XGBoost]) were developed using repeated stratified cross-validation, and model discrimination was assessed using the area under the receiver operating characteristic curve (AUROC), calibration analysis, and bootstrap-derived confidence intervals. Explainable artificial intelligence using SHapley Additive exPlanations (SHAP) was applied to the best-performing model. Serum lactate remained the only independent predictor of in-hospital mortality in the multivariable logistic regression analysis. Among the ML models, XGBoost achieved the highest numerical AUROC (0.745; 95% confidence interval, 0.680–0.807), although differences between models (DeLong’s test) were not statistically significant. SHAP analysis identified bicarbonate, lactate, potassium, creatinine, and activated partial thromboplastin time as the largest contributors to the mortality predictions generated by the optimized XGBoost model. These findings suggest that interpretable ML may complement conventional regression by providing transparent, individualized mortality prediction while preserving clinical interpretability. However, these exploratory findings require prospective multicentre validation before routine clinical implementation.
- Open Access
- 23 July 2026
- Published by Springer Nature in BMC Pulmonary Medicine
- Vol. 26 (1)
Metabolomic signatures of early pathway disruptions and clinical outcomes in community-acquired pneumonia: a prospective case-control study
Background: The disruption of metabolic pathways in patients with community-acquired pneumonia (CAP) may provide insights into disease progression. We aim to identify early disruption in the serum metabolome of CAP patients using untargeted metabolic approaches. Methods: We conducted a single-centre, prospective case-control study involving 105 participants (70 with CAP and 35 healthy controls), between January 2025 and November 2025. We also performed a subgroup exploratory metabolomic analysis among the 70 CAP patients, comparing 60 patients with severe CAP requiring ICU admission with the 10 non-severe CAP patients who did not require ICU admission. Untargeted serum metabolomics was performed using gas chromatography-mass spectrometry (GC-MS) to investigate metabolic dysregulation in patients with CAP. The analysis of the clinical data was performed using SPSS version 25.0. The data of metabolomics was analyzed using MetaboAnalyst version 6.0. After the preprocessing, log 10 transformation, normalization and scaling, 93 GC-MS metabolites were retained from the initial 300 metabolites. Principal component analysis was done for unsupervised clustering, which was then followed by OPLS-DA for supervised group discrimination. Model performance was assessed using R²Y and cross-validation-derived Q² values, and model robustness and overfitting exclusion was further evaluated using permutation testing. Discriminatory metabolites were identified using VIP scores, fold change, log₂ fold change and univariate testing with false discovery rate correction where applicable. ROC analysis, heatmap-based clustering and pathway enrichment analysis were performed. Results: The variable importance in projection (VIP) scores showed that out of 93 untargeted metabolites studied using GC-MS, the greatest differences between the CAP patients and healthy controls were about malic acid (VIP:3.374, fold change 2.591), 4-hydroxybenzeneacetic acid (VIP:2.459, fold change 27.97), 2-hydroxybutyric acid (VIP:2.059, fold change 4.779), and 4-hydroxyphenyllactic acid (VIP:1.882, fold change 20.67). Key disrupted pathways in CAP patients included propanoate, glyoxylate, and dicarboxylate; tricarboxylic acid cycle; nitrogen; tyrosine, cysteine, and methionine; and amino acid pathways. Comparison between the severe CAP patients (ICU admission required) and non-severe CAP patients (no-ICU admission), VIP analysis identified boric acid, ethylene glycol, ethanimidic acid, linoleic acid, oleic acid, cholesterol, 3-hydroxybutyrate, and α-tocopherol as the principal discriminating metabolites, while pathway analysis revealed perturbations in glutathione metabolism, amino acid metabolism, the tricarboxylic acid cycle, pyruvate metabolism, and arachidonic acid metabolism. Conclusions: CAP patients demonstrated early disruptions of key metabolic pathways compared with healthy controls, spanning energy metabolism, amino acid pathways, redox balance, lipid inflammatory signalling, and gut–host metabolic axes. These findings highlight the potential utility of metabolomic profiling as an adjunctive clinical tool for early risk stratification and biochemical phenotyping in CAP. Trial registration: Clinical trials Registry of India (CTRI 2024/12/078631) registration was done 27.12.2024.
- 23 July 2026
- Published by Springer Nature in BMC Pulmonary Medicine
- Vol. 26 (1)
Open AccessMetabolomic signatures of early pathway disruptions and clinical outcomes in community-acquired pneumonia: a prospective case-control study
Background: The disruption of metabolic pathways in patients with community-acquired pneumonia (CAP) may provide insights into disease progression. We aim to identify early disruption in the serum metabolome of CAP patients using untargeted metabolic approaches. Methods: We conducted a single-centre, prospective case-control study involving 105 participants (70 with CAP and 35 healthy controls), between January 2025 and November 2025. We also performed a subgroup exploratory metabolomic analysis among the 70 CAP patients, comparing 60 patients with severe CAP requiring ICU admission with the 10 non-severe CAP patients who did not require ICU admission. Untargeted serum metabolomics was performed using gas chromatography-mass spectrometry (GC-MS) to investigate metabolic dysregulation in patients with CAP. The analysis of the clinical data was performed using SPSS version 25.0. The data of metabolomics was analyzed using MetaboAnalyst version 6.0. After the preprocessing, log 10 transformation, normalization and scaling, 93 GC-MS metabolites were retained from the initial 300 metabolites. Principal component analysis was done for unsupervised clustering, which was then followed by OPLS-DA for supervised group discrimination. Model performance was assessed using R²Y and cross-validation-derived Q² values, and model robustness and overfitting exclusion was further evaluated using permutation testing. Discriminatory metabolites were identified using VIP scores, fold change, log₂ fold change and univariate testing with false discovery rate correction where applicable. ROC analysis, heatmap-based clustering and pathway enrichment analysis were performed. Results: The variable importance in projection (VIP) scores showed that out of 93 untargeted metabolites studied using GC-MS, the greatest differences between the CAP patients and healthy controls were about malic acid (VIP:3.374, fold change 2.591), 4-hydroxybenzeneacetic acid (VIP:2.459, fold change 27.97), 2-hydroxybutyric acid (VIP:2.059, fold change 4.779), and 4-hydroxyphenyllactic acid (VIP:1.882, fold change 20.67). Key disrupted pathways in CAP patients included propanoate, glyoxylate, and dicarboxylate; tricarboxylic acid cycle; nitrogen; tyrosine, cysteine, and methionine; and amino acid pathways. Comparison between the severe CAP patients (ICU admission required) and non-severe CAP patients (no-ICU admission), VIP analysis identified boric acid, ethylene glycol, ethanimidic acid, linoleic acid, oleic acid, cholesterol, 3-hydroxybutyrate, and α-tocopherol as the principal discriminating metabolites, while pathway analysis revealed perturbations in glutathione metabolism, amino acid metabolism, the tricarboxylic acid cycle, pyruvate metabolism, and arachidonic acid metabolism. Conclusions: CAP patients demonstrated early disruptions of key metabolic pathways compared with healthy controls, spanning energy metabolism, amino acid pathways, redox balance, lipid inflammatory signalling, and gut–host metabolic axes. These findings highlight the potential utility of metabolomic profiling as an adjunctive clinical tool for early risk stratification and biochemical phenotyping in CAP. Trial registration: Clinical trials Registry of India (CTRI 2024/12/078631) registration was done 27.12.2024.
- Open Access
- 19 December 2025
- Published by Jaypee Brothers Medical Publishing in Indian Journal of Critical Care Medicine
- Vol. 29 (12) , 1010-1019
Utility of Extracellular Nicotinamide Phosphoribosyl Transferase as a Novel Biomarker in Predicting Early Severe Organ Dysfunction and Mortality in Acute Respiratory Distress Syndrome: A Prospective Observational Study
Nicotinamide phosphoribosyl transferase (NAMPT) is an upstream cytozyme (cytokine plus enzyme) with unique features, having intracellular NAMPT (iNAMPT) and extracellular NAMPT (eNAMPT) components. Genetic associations and therapeutic inhibition highlight its potential as both a biomarker and therapeutic target. We aimed to study the utility of eNAMPT as a predictor of severe organ dysfunction and its association with mortality in acute respiratory distress syndrome (ARDS). This is a single-center, prospective observational study involving 90 patients with ARDS. We noted plasma eNAMPT levels, oxygenation levels, inflammatory markers, lung ultrasound scores, driving pressures, and echocardiography parameters. Severity of organ dysfunction and ICU mortality were the outcomes. Plasma eNAMPT was found to be a predictor of severe organ dysfunction in ARDS [adjusted OR: 1.343, 95% CI (1.105–1.634), p-value 0.003]. The cut-off eNAMPT level of ≥ 4.38 ng/mL was used to predict early severe organ dysfunction, with an AUC of 0.752, 95% CI (0.647–0.857), p-value < 0.001, sensitivity of 71.4%, and specificity of 79.2%. Plasma eNAMPT was significantly higher in non-survivors [4.53 (4–9.98) ng/mL] as compared to survivors [3.76 (3.40–4.29) ng/mL] (p-value < 0.001). eNAMPT level ≥4.175 ng/mL was associated with higher mortality (hazard ratio: 3.82; 95% CI: 2.010–7.276, p-value < 0.001) and a shorter median survival time [5 days vs. 16 days (Log-rank (Mantel-Cox) p-value < 0.001]. Plasma eNAMPT (≥4.38 ng/mL) predicts early severe organ dysfunction in ARDS patients. It is also associated with mortality and a shorter median survival time. Rao S, Srinivas T, Parampalli V, Todur P, Udupa AA, Rao S, et al. Utility of Extracellular Nicotinamide Phosphoribosyl Transferase as a Novel Biomarker in Predicting Early Severe Organ Dysfunction and Mortality in Acute Respiratory Distress Syndrome: A Prospective Observational Study. Indian J Crit Care Med 2025;29(12):1010–1019.
- 19 December 2025
- Published by Jaypee Brothers Medical Publishing in Indian Journal of Critical Care Medicine
- Vol. 29 (12) , 1010-1019
Open AccessUtility of Extracellular Nicotinamide Phosphoribosyl Transferase as a Novel Biomarker in Predicting Early Severe Organ Dysfunction and Mortality in Acute Respiratory Distress Syndrome: A Prospective Observational Study
Nicotinamide phosphoribosyl transferase (NAMPT) is an upstream cytozyme (cytokine plus enzyme) with unique features, having intracellular NAMPT (iNAMPT) and extracellular NAMPT (eNAMPT) components. Genetic associations and therapeutic inhibition highlight its potential as both a biomarker and therapeutic target. We aimed to study the utility of eNAMPT as a predictor of severe organ dysfunction and its association with mortality in acute respiratory distress syndrome (ARDS). This is a single-center, prospective observational study involving 90 patients with ARDS. We noted plasma eNAMPT levels, oxygenation levels, inflammatory markers, lung ultrasound scores, driving pressures, and echocardiography parameters. Severity of organ dysfunction and ICU mortality were the outcomes. Plasma eNAMPT was found to be a predictor of severe organ dysfunction in ARDS [adjusted OR: 1.343, 95% CI (1.105–1.634), p-value 0.003]. The cut-off eNAMPT level of ≥ 4.38 ng/mL was used to predict early severe organ dysfunction, with an AUC of 0.752, 95% CI (0.647–0.857), p-value < 0.001, sensitivity of 71.4%, and specificity of 79.2%. Plasma eNAMPT was significantly higher in non-survivors [4.53 (4–9.98) ng/mL] as compared to survivors [3.76 (3.40–4.29) ng/mL] (p-value < 0.001). eNAMPT level ≥4.175 ng/mL was associated with higher mortality (hazard ratio: 3.82; 95% CI: 2.010–7.276, p-value < 0.001) and a shorter median survival time [5 days vs. 16 days (Log-rank (Mantel-Cox) p-value < 0.001]. Plasma eNAMPT (≥4.38 ng/mL) predicts early severe organ dysfunction in ARDS patients. It is also associated with mortality and a shorter median survival time. Rao S, Srinivas T, Parampalli V, Todur P, Udupa AA, Rao S, et al. Utility of Extracellular Nicotinamide Phosphoribosyl Transferase as a Novel Biomarker in Predicting Early Severe Organ Dysfunction and Mortality in Acute Respiratory Distress Syndrome: A Prospective Observational Study. Indian J Crit Care Med 2025;29(12):1010–1019.
- Open Access
- 8 December 2025
- Published by Taylor & Francis in Annals of Medicine
- Vol. 57 (1) , 2596535
The utility of angiopoietin-2 and blood cell-derived biomarker indices in differentiating rapidly improving acute respiratory distress syndrome (RIARDS) phenotype from persistent-ARDS: a prospective observational study
Rapidly improving acute respiratory distress syndrome (RIARDS) is defined as normalization of oxygenation (PaO2/FiO2 >300 mmHg) within 24 h of invasive mechanical ventilation. Angiopoietin-2 (Ang-2), a marker of endothelial dysfunction, may predict acute respiratory distress syndrome (ARDS) outcomes, but its role in RIARDS is unclear. This study evaluated Ang-2 and other biomarkers for predicting RIARDS and their association with survival. In this prospective observational study (June 2023-January 2025) at a 2032-bedded tertiary centre, 193 consecutively enrolled mechanically ventilated ARDS patients were assessed. Demographic, clinical, and laboratory data were recorded on Day 1. Plasma Ang-2 was measured using a high-sensitivity ELISA. Primary outcome was RIARDS; secondary outcome was ICU survival. RIARDS occurred in 18.6% of patients, highest in mild ARDS (52.9%) versus moderate (19.2%) and severe (10.4%). Multivariable regression identified Ang-2 as an independent predictor. A cut-off of 5896 pg/mL yielded an AUC of 0.731 (sensitivity 57.3%, specificity 88.9%). Patients with Ang-2 < 5896 pg/mL were more likely to develop RIARDS (86.1% vs. 13.9%, p-value < 0.001). Kaplan-Meier analysis showed significantly better ICU survival in RIARDS. Early Ang-2 measurement may help differentiate RIARDS from persistent ARDS, enabling prognostic enrichment and personalized management.
- 8 December 2025
- Published by Taylor & Francis in Annals of Medicine
- Vol. 57 (1) , 2596535
Open AccessThe utility of angiopoietin-2 and blood cell-derived biomarker indices in differentiating rapidly improving acute respiratory distress syndrome (RIARDS) phenotype from persistent-ARDS: a prospective observational study
Rapidly improving acute respiratory distress syndrome (RIARDS) is defined as normalization of oxygenation (PaO2/FiO2 >300 mmHg) within 24 h of invasive mechanical ventilation. Angiopoietin-2 (Ang-2), a marker of endothelial dysfunction, may predict acute respiratory distress syndrome (ARDS) outcomes, but its role in RIARDS is unclear. This study evaluated Ang-2 and other biomarkers for predicting RIARDS and their association with survival. In this prospective observational study (June 2023-January 2025) at a 2032-bedded tertiary centre, 193 consecutively enrolled mechanically ventilated ARDS patients were assessed. Demographic, clinical, and laboratory data were recorded on Day 1. Plasma Ang-2 was measured using a high-sensitivity ELISA. Primary outcome was RIARDS; secondary outcome was ICU survival. RIARDS occurred in 18.6% of patients, highest in mild ARDS (52.9%) versus moderate (19.2%) and severe (10.4%). Multivariable regression identified Ang-2 as an independent predictor. A cut-off of 5896 pg/mL yielded an AUC of 0.731 (sensitivity 57.3%, specificity 88.9%). Patients with Ang-2 < 5896 pg/mL were more likely to develop RIARDS (86.1% vs. 13.9%, p-value < 0.001). Kaplan-Meier analysis showed significantly better ICU survival in RIARDS. Early Ang-2 measurement may help differentiate RIARDS from persistent ARDS, enabling prognostic enrichment and personalized management.
- Open Access
- 18 November 2025
- Published by Jaypee Brothers Medical Publishing in Indian Journal of Critical Care Medicine
- Vol. 29 (11) , 942-953
Aviptadil Therapy in Acute Respiratory Distress Syndrome Patients: A Systematic Review and Meta-analysis
Acute respiratory distress syndrome (ARDS) is a life-threatening condition with a high mortality rate despite advances in supportive care. Aviptadil, a synthetic analogue of vasoactive intestinal peptide (VIP), exhibits anti-inflammatory potential and cytoprotective effects that may improve pulmonary function. However, its role in improving survival among ARDS patients remains uncertain. This systematic review and meta-analysis aimed to evaluate the effectiveness of aviptadil in improving survival and oxygenation outcomes in ARDS. A comprehensive search was conducted across six databases—PubMed, Scopus, Embase, Google Scholar, Cochrane Library, and Web of Science—up to October 2025. Two investigators independently screened eligible studies. Descriptive synthesis and meta-analysis were performed using a random-effects model. The risk of bias (RoB) was assessed with the RoB 2 tool for randomized controlled trials (RCTs) and the Joanna Briggs Institute (JBI) tool for case series. The systematic review included nine studies (two RCTs and seven case series) with a total of 665 patients, 361 of whom received aviptadil. Meta-analysis of the two RCTs yielded a pooled prevalence of survival of 0.71 [95% confidence interval (CI): 0.53–0.87]. The survival odds ratio (OR) comparing aviptadil to placebo was 1.01 (95% CI: 0.72–1.42, p = 0.93), indicating no significant benefit. Across the case series, 48 of 54 patients (88.9%) survived following aviptadil therapy, demonstrating consistent improvements in oxygenation and reductions in inflammatory markers. Aviptadil may have a physiological role in improving oxygenation and reducing inflammation in ARDS; however, current evidence does not indicate a significant survival benefit. Larger, well-designed RCTs are needed to clarify its therapeutic potential. Udupa AA, Todur P, Chaudhuri S, Gupta N, Bhat VR, Nagendra D, et al. Aviptadil Therapy in Acute Respiratory Distress Syndrome Patients: A Systematic Review and Meta-analysis. Indian J Crit Care Med 2025;29(11):942–953.
- 18 November 2025
- Published by Jaypee Brothers Medical Publishing in Indian Journal of Critical Care Medicine
- Vol. 29 (11) , 942-953
Open AccessAviptadil Therapy in Acute Respiratory Distress Syndrome Patients: A Systematic Review and Meta-analysis
Acute respiratory distress syndrome (ARDS) is a life-threatening condition with a high mortality rate despite advances in supportive care. Aviptadil, a synthetic analogue of vasoactive intestinal peptide (VIP), exhibits anti-inflammatory potential and cytoprotective effects that may improve pulmonary function. However, its role in improving survival among ARDS patients remains uncertain. This systematic review and meta-analysis aimed to evaluate the effectiveness of aviptadil in improving survival and oxygenation outcomes in ARDS. A comprehensive search was conducted across six databases—PubMed, Scopus, Embase, Google Scholar, Cochrane Library, and Web of Science—up to October 2025. Two investigators independently screened eligible studies. Descriptive synthesis and meta-analysis were performed using a random-effects model. The risk of bias (RoB) was assessed with the RoB 2 tool for randomized controlled trials (RCTs) and the Joanna Briggs Institute (JBI) tool for case series. The systematic review included nine studies (two RCTs and seven case series) with a total of 665 patients, 361 of whom received aviptadil. Meta-analysis of the two RCTs yielded a pooled prevalence of survival of 0.71 [95% confidence interval (CI): 0.53–0.87]. The survival odds ratio (OR) comparing aviptadil to placebo was 1.01 (95% CI: 0.72–1.42, p = 0.93), indicating no significant benefit. Across the case series, 48 of 54 patients (88.9%) survived following aviptadil therapy, demonstrating consistent improvements in oxygenation and reductions in inflammatory markers. Aviptadil may have a physiological role in improving oxygenation and reducing inflammation in ARDS; however, current evidence does not indicate a significant survival benefit. Larger, well-designed RCTs are needed to clarify its therapeutic potential. Udupa AA, Todur P, Chaudhuri S, Gupta N, Bhat VR, Nagendra D, et al. Aviptadil Therapy in Acute Respiratory Distress Syndrome Patients: A Systematic Review and Meta-analysis. Indian J Crit Care Med 2025;29(11):942–953.