top of page

Fraudulent Data

Are you being compliant with the regulatory expectations regarding the monitoring of your trials and specifically fraud/misconduct?

Fraudulent data checks are now required by the FDA as part of routine data checks. However, there can be many different methods used on many different types of data to identify fraud.

 

Shafi Consultancy have spent many years working with, and reviewing such methods and have developed statistical techniques used to identify fraudulent data on different types of data. We know which techniques can be successfully used on the different types of data. In our process the results can also be presented to our clients in many different formats, from charts and plots to analysis tables.

​

As fraudulent data check is something that must be done for each trial, our experienced team is ready to assist you in performing the checks, standardise the process across trials and to minimise the effort and costs involved in achieving this.

See our papers on Fraudulent Data:
Fraudulent Data Checks and How to Develop them
Fraudulent data detection in clinical trial using dynamic clustering

fraudelentdata_tp.png

Why Look at

Fraudulent Data?

  • Because people’s lives and health will be at risk
     

  • To protect the rights and well-being of patients enrolled in a trial by verifying the authenticity of clinical trial data

​

  • To identify and address problems early to limit the serious implications by analyzing data quality
     

  • Maintain the research integrity in the public eye

Data Susceptible

to Fraud

  • Eligibility criteria e.g. age, medical history of the patients
     

  • Patient diary data

​

  • Repeated measurements e.g. blood pressure and laboratory data
     

  • Adverse events reporting
     

  • Assessment of medical compliance
     

  • Dates of assessment
     

  • Efficacy results
     

  • Data collected on Weekends and National Holidays or outside of normal hours

Detection of Fraud at Clinical Sites

  • Check for Inliers and Outliers
     

  • Check for Incorrect dates: Mosaic Plots

​

  • Check for Under-reporting of Adverse Events: Scatter Plots, 2D and 3D.
     

  • Rounding of integers: Line and Scatter Plots
     

  • Last Digit Preference: Volcano Plots
     

  • Check for duplicate patients by comparing key data, including age, height, weight, vital signs and visit schedules. Box plots
    can be used for these checks

We use a pool of template programs to identify fraud

Fraudulent_data2.png

Benefits of using us for fraud detection

Standardised analysis ensuring quality and consistency

Can be reproduced quickly and easily on an ongoing basis

Additional study-specific checks are always carried out

Manual data interrogation always performed on every run

Dedicated team specialized in fraudulent data checks

bottom of page