Dark Mode
Synthetic Sepsis Prediction Dataset
Patient Health Records & Digital Health
Tags and Keywords
Trusted By
"No reviews yet"
£79.99
About
The Synthetic Sepsis Prediction Dataset is a large-scale, anonymised synthetic dataset designed for research and education in critical care and infectious disease analytics. It includes patient demographic, clinical, pharmacological, and outcome-related features, simulating real-world scenarios to support the prediction, diagnosis, and treatment analysis of sepsis.
Dataset Features
- cod: Unique patient identifier (integer).
- Fecha_ing: Date of hospital admission.
- Sexo: Biological sex (Male/Female/Other).
- Edad: Patient age at admission (years).
- Hospital: Name of the hospital.
- Proced & Reg_salud: Healthcare procedural and regional codes.
- Peso / Talla / IMC: Weight (kg), height (cm), and BMI.
- Initial & 12h/24h Vitals: Systolic, diastolic, mean arterial pressures; heart and respiratory rates; saturation; capillary glucose.
- HTA, DM_2, Hipotiroidismo, Obesidad, Tabaco, ERC, Enf_coronaria, Dislipidemia, ACV, Fib_aur, Autoinmune: Binary variables indicating presence of chronic conditions.
- Antihypertensives: IECA, ARA_2, beta-blockers, calcium antagonists, diuretics, and others.
- Antidiabetics: Metformin, iSGLT2, GLP1a, DPPIV inhibitors, insulin (basal & preprandial), and others.
- Norepinephrine/Vasopressin: Use and dosage at 0h and 24h.
- Lactate, WBC, Neutrophils, Lymphocytes, NL Ratio, PCR, Procalcitonin, HCO3, pH, SOFA Score. All at baseline and after 24 hours.
- Glucose Monitoring: Glucose at 3 time points and coefficient of variation.
- Sepsis: Binary variable indicating presence of sepsis.
- Cultivos / ATB_1/2/3: Culture tests and antibiotic administration.
- IOT, Dialysis, IRA: Mechanical ventilation, renal failure, and related interventions.
- Dias_iot: Duration of intubation.
- Muerte: Mortality outcome.
Distribution
Usage
This dataset can be used for:
- Predictive Modeling: Forecast onset of sepsis, mortality risk, or treatment responsiveness.
- Clinical Decision Support: Test triage and alert systems for sepsis detection.
- Healthcare Analytics: Examine treatment trends and comorbidity impact in critical care.
- Education: Train students in clinical data science, epidemiology, and bioinformatics.
Coverage
All data is synthetically generated and anonymized to resemble real ICU patients while ensuring full privacy compliance. It includes realistic variability in patient responses and care outcomes.
License
CC0 (Public Domain)
Who Can Use It
- Healthcare AI Researchers and Clinicians: For prototyping early detection and treatment recommendation systems.
- Data Scientists: To build time-aware or multi-modal models in critical care settings.
- Medical Educators and Students: As a complex dataset for hands-on training in clinical data analysis and prediction.
QUALITY