Hello! I’m Sherminta Lawrence, a Data & Interface Analyst with over 10 years of experience integrating and analyzing healthcare data. My expertise lies in HL7 interface migration, HL7/FHIR validation, data analysis and Agile SDLC methodologies. This repository showcases some of my skills in data analytics and interoperability solutions.

This repository is designed for:

• Healthcare data professionals looking for examples of HL7/FHIR integration.
• Data analysts interested in healthcare-focused JavaScript, SQL, and Python projects.
• Developers & engineers working on interoperability solutions.
• Anyone exploring healthcare data modeling, Software Development Life Cycle (SDLC), and visualization techniques.

Real-world patient healthcare data isn’t available, however, there are datasets avaialble as well as ways to create our own data using Python.

How the data used in this repository was created and sourced: 📈 Practice Data

• Programming: Python, SQL, JavaScript
• Databases: PostgreSQL
• Healthcare Data Standards: HL7 (ADT, ORM, ORU), FHIR (Patient Resource), JSON Validation
• Data Processing & ETL: Pandas, SQLAlchemy, SSIS
• Data Visualization: Matplotlib, Seaborn

Challenge: Migrate clients from a legacy VB-based interface to a Mirth-powered HL7 interface engine while ensuring seamless data exchange with EMRs like Epic Beaker, Meditech, Vista, and Cloverleaf.

SDLC Approach:

• Followed Agile methodology with sprint-based releases.
• Participated in backlog grooming, sprint planning, and iterative testing cycles.
• Conducted interface validation using HL7 message types (ADT, ORM, ORU) to ensure compliance.
• Deployed using parallel testing strategies to minimize downtime.

Solution:

• Configured and optimized HL7 transactions (ADT, ORM, ORU) to align with EMR requirements.
• Developed FHIR validation scripts to support JSON-based patient resource exchange.
• Created SQL queries for data validation and integrity checks during migration.
• Designed monitoring workflows to track message success rates post-migration, utilizing Mirth’s channel statistics and built-in dashboards.

Outcome:

✅ Reduced HL7 transaction errors, improving data reliability.
✅ Enhanced interoperability with major EMRs like Epic Beaker, Meditech, and Cloverleaf.

graph TD;

    A[Assess Legacy HL7 Interfaces] --> B[Define Requirements]
    B --> C[Present to Stakeholders]
    C --> D[Develop Timeline & Planning]
    D --> E[Design & Configure Interfaces]
    E --> F[Test Messages & Resources]
    F --> G[Deploy & Validate]
    G --> H[Monitor & Support]

    %% Feedback loops for Agile process
    H -->|If issues found| F
    G -->|Iterative improvements| F
    D -->|Continuous Sprints| F

    %% Optional: Major redesign loop
    H -->|Major redesign needed| A

    %% Styling
    classDef default fill:#f9f9f9,stroke:#333,stroke-width:2px;
    classDef highlight fill:#e1f3d8,stroke:#82b366,stroke-width:2px;
    class F,G highlight

This diagram illustrates:

1. Legacy Interface → Mirth Engine Migration Process
2. Sprint-based development cycle (Agile Kanban representation)
3. Testing & Validation Phases for HL7 message compliance

Here’s a generalized example of how JavaScript is used in Mirth to parse JSON patient data and create an HL7 message (with error handling).
➡️ JavaScript for Parsing JSON in Mirth

• Data Cleaning & Transformation – Used SQL and Python to handle missing values, outliers, and format inconsistencies, improving data quality for analysis.
• Predictive Modeling – Built a logistic regression model to predict hospital readmissions, helping optimize patient care.
• FHIR Resource Validation – Developed Python scripts to validate HL7 FHIR patient resources against schema standards.

• Objective: Identify eligible patients for a clinical trial based on lab results and demographic data.
• Skills Used: SQL joins, CASE statements, filtering, and patient cohort identification.
• Outcome: Helped refine patient selection criteria, improving recruitment efficiency.

➡️ Project 1_Notebook

• Objective: Validate JSON-based FHIR patient resources against HL7 standards.
• Skills Used: Python, JSON parsing, schema validation, and FHIR standards.
• Outcome: Ensured compliance with FHIR interoperability guidelines, reducing integration errors.

➡️ Project 2_Notebook

• Skills Used: Python, JSON parsing, schema validation, and FHIR standards.

• Expand FHIR validation scripts for more resources (Observations, Encounters, Medications).
• Add more interactive visualizations
• Machine Learning
• Develop API integrations for real-time data exchange.
• Instructions for how to use this repository.

Objective: Predict readmission rates within a 30 day period by generating features from our three data sets:

• Patient_Vitals: Capturing patient health status
• Encounters: Understanding patient visits and length of stay
• Patient_RX: Analyzing medication history

For questions or collaboration opportunities, reach out via LinkedIn or email me at shermintalawrence@gmail.com.

This README is continuously updated as I refine and expand my projects. 🚀