---
pretty_name: Synthetic ROCS Dataset (100K Patients)
language: en
license: other
tags:
- synthetic-data
- healthcare
- cerebral-palsy
- rehabilitation
- real-world-evidence
- dbbun
size_categories:
- 10K<n<100K
task_categories:
- feature-extraction
- text-generation
---
# Synthetic ROCS Dataset (100K Patients)

## Overview
This dataset is a **fully synthetic reproduction** of the patient cohort described in [A Big Data Approach to Evaluate Receipt of Optimal Care in Childhood Cerebral Palsy](https://pubmed.ncbi.nlm.nih.gov/36755522/).

It simulates the **Receipt of Optimal Care Score (ROCS)** framework across 100K synthetic patients, modeling event sequences, adherence probabilities, and component-level quality-of-care scores.  

All data were algorithmically generated — **no real patient data are included**.

---

## Motivation
The purpose of this dataset is to provide a **privacy-free reference cohort** for:
- Testing analytic pipelines on pediatric rehabilitation data,
- Practicing longitudinal cohort analysis (PM&R, PT, OT, Spasticity management),
- Teaching or demonstrating real-world evidence (RWE) methodologies,
- Benchmarking ROCS-style scoring frameworks.

---

### Files
| File | Description |
|------|--------------|
| **patients.csv** | Patient-level attributes (sex, race, ethnicity, GMFCS, phenotype, spasticity flags, and ROCS summary scores). |
| **events.csv** | Longitudinal event records for PMR, PT, OT, and Spasticity components; each row = one care event. |
| **scores.csv** | Component-level and adjusted ROCS scores per patient. |

---

### Reference
Mitelpunkt A, Stodola MA, Vargus-Adams J, Kurowski BG, Greve K, Bhatnagar S, Aronow B, Zahner J, Bailes AF. A big data approach to evaluate receipt of optimal care in childhood cerebral palsy. Disabil Rehabil. 2024 Feb;46(4):723-730. doi: 10.1080/09638288.2023.2175919. Epub 2023 Feb 8. PMID: 36755522; PMCID: PMC10406971.

---