Anomaly detection and secure data collection for healthcare research
A research organization running survey studies with non-specialist interviewers needed to trust its data. We added a statistical anomaly detection component to its survey system, built on open-source big data tools, and implemented REDCap for secure, compliant and transparent data collection. Two systems, one outcome: research that stands up.
- Statistical
- anomaly detection over survey responses, flagging interviewer error and biasEngagement
- REDCap
- implemented for secure, compliant, transparent data collectionEngagement
- Integrated
- into the client's existing survey infrastructure rather than replacing itEngagement
- Client
- A healthcare research organization
- Sector
- Healthcare research and survey studies
- Year
- 2024
- Stack
- Java integration with the existing survey system; open-source big data tooling for statistical analysis and reporting; REDCap for data capture
- Offer today
- AI Integration for Java Systems (classical statistics and anomaly detection) and integration work
Survey data collected by non-specialists, and research that depended on it
The client needed to ensure the quality of data collected during research studies, especially those involving surveys conducted by non-specialist interviewers. Human error and potential interviewer bias could compromise the integrity of the data and, with it, the credibility of the research. A quantitative analysis and statistical anomaly detection methodology was needed to identify and mitigate data quality issues, and the collection itself had to become more secure and transparent.
A detection component on the existing system, and REDCap for collection
We developed an additional component for the client's existing survey system that aggregates and analyses collected data to detect anomalies. It was integrated into the existing infrastructure and uses open-source big data tools to perform the statistical analysis and generate detailed reports, which are presented to the research teams so they can quickly identify and correct anomalies.
In a companion engagement we implemented REDCap, the research electronic data capture platform, for secure, compliant and transparent data collection, improving research data integrity at the point of capture.
What changed for the research teams
- Higher
- data quality standards across the organization's studiesClient
- Faster
- identification and correction of anomalies from the generated reportsClient
- Credible
- more reliable data, producing more credible and impactful researchClient
The offers this engagement proves
Book a scoping call
Thirty minutes with an architect, not a salesperson. You leave with a written view of scope, price model and whether we are the right team for it.