Electronic Health Record Data

Where it comes from

Highlighted claims

• The study used MIMIC-IV version 2.2, a de-identified EHR database developed by MIT with Beth Israel Deaconess Medical Center. — Synthetic data-augmented machine learning for 30-day readmission prediction in patients with chronic conditions: a retrospective real-world study
• The final analytical sample included 33,882 adult ICU patients across COPD, HF, and T2DM cohorts. — Synthetic data-augmented machine learning for 30-day readmission prediction in patients with chronic conditions: a retrospective real-world study
• MIMIC-IV contains clinical notes, laboratory data, pharmacologic data, and demographic data from more than 60,000 ICU admissions between 2008 and 2019. — Synthetic data-augmented machine learning for 30-day readmission prediction in patients with chronic conditions: a retrospective real-world study
• Structured EHR variables included demographics, comorbidities, laboratory values, vital signs, severity scores, medication classes, and prior healthcare utilisation. — Synthetic data-augmented machine learning for 30-day readmission prediction in patients with chronic conditions: a retrospective real-world study
• Unstructured clinical notes supplied behavioral and care-continuity indicators such as medication non-adherence, substance use, social support, discharge education, and outpatient follow-up planning. — Synthetic data-augmented machine learning for 30-day readmission prediction in patients with chronic conditions: a retrospective real-world study