You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

YAML Metadata Warning: empty or missing yaml metadata in repo card (https://huggingface.co/docs/hub/datasets-cards)

MIMIC-IV Clinical Notes - Patients with 3+ Notes (From Original Preprocessed Dataset)

Dataset Description

This dataset contains discharge notes for all patients from the MIMIC-IV Note dataset who have 3 or more clinical notes. This is a subset of the GTE2 (≥2 notes) dataset, further filtered to include only patients with at least 3 clinical notes. The data has been merged with admissions information to provide comprehensive patient context.

Dataset Statistics

  • Total Clinical Notes: 192,478
  • Unique Patients: 33,458
  • Unique Hospital Admissions (hadm_id): 192,478
  • Total Columns: 15

Notes per Patient

Metric Value
Mean 5.75
Median 4.0
Max 89
Min 3
Std Dev 4.87

Source

  • Original data: MIMIC-IV Note Dataset (discharge.csv)
  • Merged with: MIMIC-IV Admissions data
  • Filter criteria: Patients with >= 3 discharge notes
  • Parent dataset: mimic-capstone/mimic_clinical_notes_from_og_preprocessed (GTE2 subset)

Columns

The dataset includes the following 15 columns:

  1. note_id
  2. subject_id
  3. hadm_id
  4. note_type
  5. note_seq
  6. charttime
  7. storetime
  8. text
  9. admittime
  10. dischtime
  11. deathtime
  12. admission_type
  13. marital_status
  14. race
  15. input

Key Features

  • Clinical Notes: Full text of discharge summaries
  • Patient Demographics: Subject ID, marital status, race
  • Admission Details: Admission time, discharge time, death time, admission type
  • Note Metadata: Note ID, note sequence, chart time, store time

Data Quality

  • All patients verified to have >= 3 clinical notes
  • Inner join performed on hadm_id with admissions data
  • Derived from the GTE2 preprocessed dataset

Use Cases

  • Clinical NLP research
  • Multi-document summarization
  • Longitudinal patient analysis
  • Healthcare AI applications

Citation

If you use this dataset, please cite the original MIMIC-IV database:

Johnson, A., Bulgarelli, L., Pollard, T., Horng, S., Celi, L. A., & Mark, R. (2023). 
MIMIC-IV (version 2.2). PhysioNet. https://doi.org/10.13026/6mm1-ek67

License

This dataset is derived from MIMIC-IV and is subject to the same usage restrictions. You must be a credentialed user with the appropriate data use agreement.

Downloads last month
17