Good Research Practices at Duke: Improve data accuracy with multiple data checks by different people

Author: 
Emilia Chiscop-Head, PhD, Scientific Integrity Associate

Today we launch a new “Integrity Insights” series. The goal of this series is to promote good research practices that Duke researchers and administrators implement in their units to encourage integrity and reduce errors or bias.

This series is inspired by the feedback from our research community about the need to support and promote positive conversations. Every day at Duke, the overwhelming majority of the research community carry out research with passion and integrity - to learn, innovate and discover in service to society. We want to share the invaluable practices that research groups are using to foster integrity.

If you have such a story to share, please email us at asist@duke.edu or fill out the survey “Recognizing and sharing good practices in research”.

 

At one of our recent Research Integrity Roundtables, participants were reflecting upon common patterns or “ringing bells” that could signal a case of questionable research or research misconduct, such as: excessive trust combined with the absence of oversight and weak supervision of the research work.

We asked participants what research practices their research teams use to minimize the potential for human bias or error.

The answers were truly inspirational.

One of them came from Dr. Alfredo Daniel Guerron. An Assistant Professor of Surgery at Duke and Director of the Minimally Invasive & Bariatric Surgery Fellowship, Dr. Guerron created a collaborative system that fosters peer mentoring and shared responsibility on all research projects. As a practice, his fellows meet regularly to provide peer feedback on their work, review results and decide next steps. “This way I ensure that data is checked twice or three times by individuals with different eyes, different approaches and, most importantly, different interests. This minimizes the chances for data manipulation”, explains Dr. Guerron.

A challenge of collaborative work is finding the best platforms for data sharing. Dr. Guerron’s team is using REDCap for clinical data and Duke Box for documents and papers. However, resolving the platform challenges of working collaboratively are worth it, because the benefits of team-based research approaches are great. The collaborative approach creates team cohesion, allows for peer mentoring and cultivates solid positive work relationships. “It reduces the competition among fellows. Everyone knows what project each person is leading and which one is juicier. They can collaborate on all of them, be co-authors on papers they will publish, and develop good work relationships”, says Dr. Guerron.

Dr. Guerron’s current research focuses on lung inflammatory markers after bariatric surgery. Our overall goal in this proposal is to not only elucidate the immediate benefits of bariatric surgery in asthma but also achieve important insights that can better aid understanding of the mechanism by which these procedures improve the disease in obese patients. Our central hypothesis is that bariatric surgery induces early weight-loss independent changes in systemic markers of inflammation that positively predict post-surgery asthma outcomes and distinguish pathologic mechanisms in early-onset and late-onset obese asthma”, explains Dr. Guerron.

Based on a database of asthma patients his team analyzes patient response to medications pre and post-surgery at multiple time intervals.  “We notice that medication count drops as early as 3 weeks. The hypothesis is that the medication drop is not related to weight loss. It must be something else that impacts the inflammatory environment. Based on a paper published in Surgical Endoscopy we thought we should look at biological markers to make it more robust. The first goal is to have the profile of inflammatory markers and the second goal is to identify which inflammatory markers are working. “ 

Dr. Guerron in the news: "Latino Leader of the Week".