# Data dictionary

## `data/rwe_patient_features.csv`

One row per synthetic patient. This dataset supports the trial-eligibility funnel and documents the features used to define the trial-like population.

| Variable | Description |
|---|---|
| `patient_id` | Sequential public identifier for the synthetic patient. |
| `birth_date` | Synthetic date of birth. |
| `gender` | Gender field carried from the synthetic source data. |
| `age_at_analysis_date` | Age at the fixed analysis date (2026-06-09). |
| `n_conditions` | Number of distinct condition descriptions in the source history. |
| `n_medications` | Number of distinct medication descriptions in the source history. |
| `n_encounters` | Total recorded encounters. |
| `n_emergency` | Number of emergency encounters. |
| `n_inpatient` | Number of inpatient encounters. |
| `followup_years` | Time between first and last recorded encounter in years. |
| `hypertension` | Indicator derived from condition descriptions. |
| `hyperlipidemia` | Indicator derived from condition descriptions. |
| `coronary_heart_disease` | Indicator derived from condition descriptions. |
| `stroke_history` | Indicator for a recorded stroke/cerebrovascular condition. |
| `myocardial_infarction_history` | Indicator for a recorded myocardial infarction/heart attack. |
| `heart_failure` | Indicator for recorded heart failure. |
| `chronic_kidney_disease` | Indicator for recorded chronic kidney disease. |
| `cancer_history` | Indicator for recorded cancer/malignant neoplasm. |
| `crit_age` | Meets the age criterion (45-75 years). |
| `crit_alive` | Meets the alive-at-analysis criterion. |
| `crit_followup` | Has at least two years of recorded follow-up. |
| `crit_cv_risk` | Meets the cardiovascular-risk definition. |
| `crit_comorbidity` | Meets the maximum comorbidity-count criterion. |
| `crit_polypharmacy` | Meets the maximum medication-count criterion. |
| `crit_no_stroke` | Has no recorded stroke history. |
| `crit_no_mi` | Has no recorded myocardial infarction history. |
| `crit_no_hf` | Has no recorded heart-failure history. |
| `crit_no_ckd` | Has no recorded chronic kidney disease. |
| `crit_no_cancer` | Has no recorded cancer history. |
| `crit_utilization` | Meets the emergency/inpatient utilization criterion. |
| `trial_eligible` | Meets all trial-like eligibility criteria simultaneously. |

## `data/rwe_htn_outcome.csv`

One row per synthetic patient in the hypertension outcome cohort. This dataset supports the five-year diabetes-risk, regression and transportability analyses.

| Variable | Description |
|---|---|
| `patient_id` | Sequential public identifier matching the patient-level dataset. |
| `index_date` | First recorded hypertension date used as the cohort index date. |
| `trial_eligible` | Trial-like eligibility status derived from the patient-level rules. |
| `pred_at_index` | Prediabetes recorded on or before the hypertension index date. |
| `age_index` | Age at the hypertension index date. |
| `gender` | Gender field carried from the synthetic source data. |
| `n_conditions` | Number of distinct recorded conditions. |
| `n_medications` | Number of distinct recorded medications. |
| `n_emergency` | Number of emergency encounters. |
| `n_inpatient` | Number of inpatient encounters. |
| `diabetes_5y` | Indicator that diabetes was recorded after index and within five years. |