Build a Multi-Center Trial Recruitment Diversity and Inclusion Modeler — implementation

## Build
Analyze synthetic cohort records across simulated medical centers to project demographic representation in clinical trial recruitment. Formulate recruitment adjustments to address severe demographic underrepresentation.

## User
Trial Equity and Feasibility Officer

## Workflow
1. Ingest synthetic multi-center participant records with geographic and demographic covariates.
2. Compute site-specific eligibility yields against specified trial inclusion criteria.
3. Calculate disparity indices across age, sex, and ethnicity brackets against regional baselines.
4. Deliver a feasibility adjustment report with site-level cohort rebalancing suggestions.

## Interface
Multi-site demographic breakdown dashboard with disparity index gauges and scenario simulation toggles.

## Data
Construct mock demographic records in normalized CSV or JSON format representing five simulated research hospital sites.
Starter examples below are synthetic. They are not downloaded records.

## Source
[HL7 FHIR overview](https://hl7.org/fhir/overview.html)
Employ FHIR ResearchStudy and Group demographic structure patterns to define cohort stratification criteria.

## Tools
anthropic-claude, pandas, scipy.
anthropic-claude processes complex demographic disparity summaries to synthesize coherent site rebalancing strategies.
Hosted Claude API for structured analytical summarization and protocol sensitivity review.

## Deliverables
Submit runnable source, pinned dependencies, setup commands, a working interface, example outputs, and a recorded end-to-end demo. Include tests and an action log.

## Acceptance
- Ingests synthetic multi-site cohort files and validates demographic schema fields.
- Computes demographic representation ratios relative to target population distributions.
- Flags sites where inclusion yield drops below predetermined equity thresholds.
- Outputs structured JSON containing disparity metrics and site-specific recommendations.

## Starter Examples
```json
[
  {
    "input": {
      "study_id": "DIV-2026-T1",
      "target_distribution": {
        "group_a": 0.5,
        "group_b": 0.5
      },
      "site_records": [
        {
          "site": "Site_North",
          "group": "group_a",
          "eligible": true
        },
        {
          "site": "Site_North",
          "group": "group_a",
          "eligible": true
        },
        {
          "site": "Site_North",
          "group": "group_b",
          "eligible": false
        },
        {
          "site": "Site_South",
          "group": "group_b",
          "eligible": true
        }
      ]
    },
    "expected_output": {
      "study_id": "DIV-2026-T1",
      "total_eligible": 3,
      "observed_distribution": {
        "group_a": 0.67,
        "group_b": 0.33
      },
      "disparity_detected": true,
      "priority_site_for_outreach": "Site_North"
    }
  },
  {
    "input": {
      "study_id": "DIV-2026-T2",
      "target_distribution": {
        "group_a": 0.5,
        "group_b": 0.5
      },
      "site_records": []
    },
    "expected_output": {
      "study_id": "DIV-2026-T2",
      "total_eligible": 0,
      "observed_distribution": {},
      "disparity_detected": false,
      "priority_site_for_outreach": "none"
    }
  }
]
```

## Safety
Research protocol planning system using synthetic demographics; makes no individual clinical care decisions.
Use mock external writes. Require human review before any real deployment.