With the growing public attention on food allergy, there has been a surge of interest in clinical trials. Internet-based enrollment tools offer the benefits of efficiency, organization and fairness to the recruitment process. However, there are concerns about selection bias that may be introduced by the use of these tools.
In 2015, the UNC Food Allergy Initiative (UNC FAI) developed an IRB-approved web-based REDCap database for patients and families interested in research participation. We compared the demographics of children identified as food allergic in the 2012 North Carolina Statewide Child Health Assessment and Monitoring Program (CHAMP) with children registered in the UNC FAI database.
The data showed no differences in gender between the UNC FAI database and the CHAMP survey (male 63% vs 57%, female 37% vs 42% respectively). In the UNC FAI database, African Americans were underrepresented (3.7% vs 25%) and whites were over-represented (85% vs 61%) but the percentage of “other” races was similar (12% vs 12.6%). For age, a higher percentage of children under five (36% vs 21%) and a lower percentage of 14 to 17 year olds (7.3% vs 36.8) were found in the UNC FAI database compared to the CHAMP survey.
A web-based REDCap database can be an effective tool for study recruitment. However, the risk for selection bias exists with African Americans and older children underrepresented in the UNC FAI database as compared to the general population of NC. Use of these tools should be combined with efforts to attract these populations.