top of page

Tracking the Vital Signs of Clinical Data Integrity



I had defined the Reliability Pulse of Monitoring Clinical Data's Rhythm of Quality in both my previous EMR (Patient Electronic Health Record Technology) company and in the Clinical Trial Research Organization and thought to share some of the Key Performance Indicators (KPI) we've used. I tried to incorporate elements related to clinical trials, data quality dimensions, compliance with GCP (Good Clinical Practice) principles, and the pharmaceutical/CRO environment.


Here are some examples of data quality KPIs (Key Performance Indicators) in the healthcare industry

Accuracy KPIs:

  • Percentage of patient diagnoses/medications matching clinical documentation

  • Percentage of claim entries aligning with medical coding standards

  • Number of dosage value errors caught per treatment cycle

Completeness KPIs:

  • Percentage of patient records with all required fields populated

  • Percentage of missed appointments with no-show reasons captured

  • Percentage of clinical trial data sets with zero missing data points

Consistency KPIs:

  • Number of variations in format/layout across electronic medical record (EMR) interfaces

  • Percentage consistency in code set/terminology usage across departments

  • Deviation from standard data dictionary definitions

Integrity KPIs:

  • Number of duplicate patient medical record instances

  • Percentage of referential integrity violations between systems

  • Number of data validation rule failures prior to ETL loads

Timeliness KPIs:

  • Percentage of patient vitals readings transmitted in real-time

  • Aged data: Number of months since last patient history review

  • Lag between lab results data availability vs defined service level agreements (SLAs)

Validity KPIs:

  • Number of values violating allowable value constraints

  • Percentage of data inputs failing validation rules/sanity checks

  • Count of records failing HIPAA data privacy rules


Here are some examples of data quality KPIs that are general to the clinical trial data:

Accuracy KPIs:

  • Percentage of data points that match source data (e.g. medical records, lab reports)

  • Number of transcription errors or data entry mistakes identified

  • Percentage of data queries resolved satisfactorily

Completeness KPIs:

  • Percentage of case report forms with all mandatory fields filled

  • Number of missing key data points (e.g. adverse events, concomitant medications)

  • Percentage of protocol deviations with reasons documented

Consistency KPIs:

  • Number of coding inconsistencies across sites/associates

  • Deviation from standard data formats, terminologies and conventions

  • Percentage of data differences between EDC and external data sources

Integrity KPIs:

  • Number of duplicate subject records identified

  • Rate of database errors, corruptions or constraint violations

  • Percentage of logical inconsistencies between related data points

Timeliness KPIs:

  • Duration between last patient visit and data entry

  • Percentage of serious adverse events reported within required timelines

  • Lag between site data transfer and availability in clinical database

Validity KPIs:

  • Number of values out of expected range (e.g. vitals, dosages)

  • Percentage of data failing edit checks or validation rules

  • Count of protocol deviations or violations against study criteria


Here are some examples of data quality KPIs relevant for a clinical research organization (CRO):

Accuracy KPIs:

  • Percentage of data transcribed accurately from source documents to EDC system

  • Number of data queries or issues identified during monitoring visits

  • Ratio of confirmatory SDV (source data verification) failures

Completeness KPIs:

  • Percentage of case report forms submitted with all expected data present

  • Number of missing critical safety data points (e.g. SAEs, con-meds)

  • Rate of protocol deviations with missing descriptors or reasons

Consistency KPIs:

  • Deviation from CDISC data standards across therapeutic areas

  • Number of terminology inconsistencies across sites/languages

  • Percentage of data transfer inconsistencies between EDC and clinical database

Integrity KPIs:

  • Number of duplicate subject records across systems/extracts

  • Rate of database errors, audit trail issues or edit check failures

  • Percentage of data failing cross-form, cross-visit integrity checks

Timeliness KPIs:

  • Average lag between patient visit and site data entry

  • Percentage of SAEs reported to sponsor within required timelines

  • Duration between site data transfer and availability in data repository

Validity KPIs:

  • Number of values out of expected range based on study plan/protocol

  • Rate of data failing up-front edit checks in EDC

  • Count of protocol deviations without justifications provided


By tracking and reviewing these KPIs, the CRO can proactively identify issues, monitor site performance, and ensure high quality clean data is being delivered to the sponsor pharmaceutical company. This allows them to uphold GCP standards, ensure subject safety, and maintain solid inspection readiness. By monitoring these clinical trial data quality KPIs, sponsors and CROs can ensure the reliability, traceability and quality of the underlying data that drives insights and submissions. High quality data is critical for patient safety, adherence to GCP principles, and regulatory compliance.


Sash Barige

Oct-29-2023

Comentarios


bottom of page