Skip to main content Link Menu Expand (external link) Document Search Copy Copied
PrecisionMedicineUnit Docs

Last update: 20240919

Example project

Objective {#objective .unnumbered}

Develop a predictive model to forecast the need for ICU admission based on daily clinical measurements, thereby enhancing patient care and resource allocation.

Funding and Project Initiation {#funding-and-project-initiation .unnumbered}

  • Funding Source: Grant application starts with project variables.

    • Example Variables: project_id, the scientific basis of the product e.g. code_repository, public_documentation, and examples of outcomes that we expect, e.g. action_taken

  • Project Registration: Register the project within the precision medicine unit framework to ensure consistent data handling and integration with existing structures.

    • Example Variable: project_id

Patient Enrolment and Data Collection {#patient-enrolment-and-data-collection .unnumbered}

  • Patient Enrolment: Follow a standardised protocol for patient recruitment to ensure uniform data collection across the unit.

    • Example Variables: project_id, patient_id, enrolment_date, consent_status, public_documentation

  • Data Collection: Implement automated systems to use daily clinical measurements from a centralised database.

    • Example Variables: measurement_date, clinical_metrics

Analysis and Software Development {#analysis-and-software-development .unnumbered}

  • Software Development: Use the published predictive algorithms in a version-controlled environment where code is systematically documented and linked to project variables.

    • Example Variables: algorithm_version, code_repository, public_documentation

  • Data Processing: Use automated scripts to process the incoming data daily, structuring it according to predefined templates that facilitate easy integration and analysis.

    • Example Variables: processed_data_output, analysis_date, code_repository

Result Generation and Clinical Integration {#result-generation-and-clinical-integration .unnumbered}

  • Result Generation: Generate predictive results daily, storing them in an accessible format within the centralised system.

    • Example Variables: patient_id, result_id, prediction_score, prediction_date, public_documentation

  • Data Return to Clinic: Integrate the predictive results back into the clinical workflow, enabling real-time decision-making.

    • Example Variables: patient_id, result_id, clinical_integration_date, action_taken, public_documentation

Outcome and Impact {#outcome-and-impact .unnumbered}

  • Actionable Clinical Outcomes: Provide clinicians with daily reports of patients potentially requiring ICU care, supporting timely and effective clinical interventions.

    • Example Variables: report_id, report_date, public_documentation

  • Feedback Loop: Regularly update the predictive model based on clinician feedback and outcome data to enhance accuracy and relevance.

    • Example Variables: feedback_id, modification_date

Project Tracking and Management {#project-tracking-and-management .unnumbered}

  • Variable Tracking: Maintain a centralised log of all project variables, ensuring that updates are propagated automatically to all linked documents and systems.

  • Documentation: Employ Markdown or LaTeX for all project documentation, linked directly to the variable tracking system for consistency and real-time updates.

    • Example Variables: document_id, last_updated

Benefits of the Unified Approach {#benefits-of-the-unified-approach .unnumbered}

  • Efficiency: Streamlines project management by reducing redundancy and automating updates, ensuring that project components are efficiently managed.

  • Transparency: Provides clear visibility into project operations through centralised tracking of all critical variables and documentation.

  • Scalability: Facilitates easy scaling of the project framework to include new patient cohorts or measurement types without extensive modifications.

|L0.2|L0.2|L0.2|L0.4|


Variable Name & Example Content & Phase(s) Used & Description/Usage\

Table  – continued from previous page
Variable Name & Example Content & Phase(s) Used & Description/Usage\

project_id & PMU2024-001 & Funding and Project Initiation, Tracking and Management & Unique identifier for the project within the Precision Med framework.
patient_id & PID123456 & Patient Enrollment, Data Collection, Result Generation, Data Processing & Unique identifier for each enrolled patient.
enrolment_date & 2024-01-01 & Patient Enrollment, Data Processing, Result Generation & Date when the patient was enrolled in the study.
consent_status & Consented & Patient Enrollment, Data Processing & Consent status of the patient for participation in the study.
measurement_date & 2024-01-02 & Data Collection, Data Processing, Result Generation & Date on which clinical measurements were taken.
clinical_metrics & Blood pressure, Heart rate & Data Collection, Data Processing, Result Generation & Types of clinical measurements collected daily.
algorithm_version & v1.0 & Software Development, Data Processing, Result Generation & Version of the predictive algorithm used.
code_repository & https://github.com & Funding and Project initiation, Software Development, Data Processing, Result Generation & URL of the version-controlled repository storing the project’s code.
data_output & Data_20240102.csv & Data Processing, Result Generation & Filename or identifier for the output from data processing.
analysis_date & 2024-01-03 & Data Processing, Result Generation & Date on which the data was analyzed.
result_id & RES12345678 & Result Generation & Unique identifier for a set of predictive results.
prediction_score & 0.85 & Result Generation & Score indicating the likelihood of ICU need.
prediction_date & 2024-01-04 & Result Generation & Date when the prediction was made.
clinical_integration_date & 2024-01-05 & Clinical Integration & Date when predictive results were integrated into the clinical workflow.
action_taken & Reviewed by clinician & Clinical Integration & Description of the clinical action taken based on predictions.
report_id & REP20240105 & Outcome and Impact, Data Processing & Unique identifier for the generated daily report.
report_date & 2024-01-05 & Outcome and Impact, Data Processing & Date when the report was generated.
feedback_id & FB20240106 & Feedback Loop & Unique identifier for feedback entry from clinicians.
modification_date & 2024-01-07 & Feedback Loop, Data Processing & Date when modifications were made to the predictive model.
document_id & DOC1234 & Project Tracking and Management, Data Processing & Identifier for a specific project document.
last_updated & 2024-01-08 & Project Tracking and Management, Data Processing & Last date the document was updated.\