Performance of Early Sepsis Prediction Models in a Tertiary Care Intensive Care Unit: A Retrospective Cohort Study.
- Jaskaranjeet kaur , Md Anesthesia Senior Resident Esic Medical Collage and hospital Faridabad India.
- Arpan Muniyal , Md Anesthesia Senior Resident Esic Medical Collage and hospital Faridabad India.
- Bikash , MBBS, MD -Internal Medicine, PGDCC, PGdip. Diabetes Associate Consultant epartment of Critical Care Amerix Superspecility Hospital, Dwarka, New Delhi India.
- Bandana Boro , MBBS, PG Trainee - Microbiology Shaheed Hasan Khan Mewati Government Medical College,Nalhar , Nuh, Haryana India.
Article Information:
Abstract:
Background: Early recognition of sepsis remains critical for improving intensive care unit (ICU) outcomes. Multiple clinical scoring systems have been proposed for early prediction. Objective: To compare the performance of Systemic Inflammatory Response Syndrome (SIRS), quick Sequential Organ Failure Assessment (qSOFA), and Modified Early Warning Score (MEWS)–based prediction model for early sepsis identification. Methods: Retrospective cohort study conducted between January 2022 and December 2024 in a tertiary ICU. Adult patients admitted with suspected infection were included. Sepsis was defined using Sepsis-3 criteria. Area under the receiver operating characteristic curve (AUROC), sensitivity, and specificity were calculated. Results: 268 patients were analyzed; 156 developed sepsis. AUROC was 0.68 for SIRS, 0.72 for qSOFA, and 0.81 for MEWS. MEWS identified patients earlier and was associated with lower 28-day mortality (21.4% vs 34.7%, p=0.018). Conclusion: The Modified Early Warning Score–based model demonstrated superior predictive performance and clinical utility for early sepsis detection.
Keywords:
Article :
INTRODUCTION:
Sepsis is a leading cause of intensive care unit mortality. Early detection tools such as Systemic Inflammatory Response Syndrome (SIRS), quick Sequential Organ Failure Assessment (qSOFA), and Modified Early Warning Score (MEWS) aim to improve outcomes through rapid recognition and timely intervention.
MATERIALS AND METHODS:
Retrospective observational study including adult ICU patients with suspected infection. Sepsis defined by Sequential Organ Failure Assessment (SOFA) score increase≥ 2 . Models evaluated: SIRS≥2, q SOFA≥2, MEWS≥5. Statistical analysis included area under the receiver operating characteristic curve (AUROC) comparison and calibration.
Study design: Retrospective cohort
Setting: Single tertiary ICU
Participants: Adult patients with suspected infection
Variables: SIRS, qSOFA, Modified Early Warning Score (MEWS), mortality
Bias: Selection and retrospective bias acknowledged
Study size: 268 patients
Statistical methods: AUROC, sensitivity, specificity
RESULTS:
The Modified Early Warning Score (MEWS) showed the best discrimination (AUROC 0.81), followed by qSOFA (0.72) and SIRS (0.68). Patients detected early had improved survival and shorter ICU stays.
Table 1. Baseline Characteristics
|
Variable |
Sepsis (n=156) |
No Sepsis (n=112) |
|
Age (years) |
62 ± 15 |
59 ± 14 |
|
Male (%) |
61% |
58% |
|
APACHE II (Acute Physiology and Chronic Health Evaluation II) |
21 ± 5 |
18 ± 4 |
|
Serum Lactate (mmol/L) |
4. 2 ± 1.9 |
2. 1 ± 1.1 |
Table 2. Diagnostic Performance of Prediction Models
|
Model |
AUROC |
Sensitivity |
Specificity |
|
Systemic Inflammatory Response Syndrome (SIRS) |
0.68 |
74.3% |
55.2% |
|
quick Sequential Organ Failure Assessment (qSOFA) |
0.72 |
63.5% |
78.4% |
|
Modified Early Warning Score (MEWS) |
0.81 |
82.1% |
73.4% |
DISCUSSION:
Composite early warning models such as the Modified Early Warning Score may outperform traditional criteria. Early identification likely contributes to timely therapy and improved outcomes.
CONCLUSION:
The Modified Early Warning Score–based prediction system may enhance early sepsis recognition and improve intensive care unit performance metrics.
REFERENCES:
1. Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016.
2. Seymour CW, Liu VX, Iwashyna TJ, et al. Assessment of Clinical Criteria for Sepsis. JAMA. 2016.
3. Churpek MM, Snyder A, Han X, et al. Early warning scores for clinical deterioration. Am J Respir Crit Care Med. 2017.
4. Henry KE, Hager DN, Pronovost PJ, Saria S. A targeted real-time early warning score for septic shock. Sci Transl Med. 2015.
5. Freund Y, Lemachatti N, Krastinova E, et al. Prognostic accuracy of qSOFA score. JAMA. 2017.
6. Subbe CP, Kruger M, Rutherford P, Gemmel L. Validation of a Modified Early Warning Score in medical admissions. QJM. 2001.
7. Jones AE, Trzeciak S, Kline JA. The Sequential Organ Failure Assessment score for predicting outcome. Crit Care Med. 2009.
8. Kaukonen KM, Bailey M, Pilcher D, et al. Systemic inflammatory response syndrome criteria in sepsis. N Engl J Med. 2015.
9. Vincent JL, Moreno R, Takala J, et al. The SOFA score to describe organ dysfunction. Intensive Care Med. 1996.
10. Levy MM, Evans LE, Rhodes A. The Surviving Sepsis Campaign bundle. Intensive Care Med. 2018.