2.6. Data Collection Process

The following characteristics were selected from the articles: authors and year of publication, country of the study, study aim, study outline, starch sample/blood sample analysis method, type of flour/starch base, production (if available in local markets or developed by the study’s authors), enrichment, in vivo/in vitro, number of samples/subjects tested, use of control samples and method to determine GI. Calibration exercises were performed before starting the review to ensure consistency among reviewers. Reviewers solved disagreements by discussion, and the third reviewer (BR) adjudicated unresolved disagreements.

All literature searches and results were documented in the search plan (which included the study question, month and year of the literature review, inclusion and exclusion criteria, and search terms). As mentioned, the report was based on the PRISMA flowchart. After completion of the data extraction and quality appraisal process, data were synthesized by three researchers (GP, SC, BR) using a standardized table containing information about reference; country; aim; bread starch sample/blood sample analysis method, type of flour/starch base, enrichment (yes/no and which ingredient used), in vivo/in vitro, number of samples/subjects tested (if triplicates/duplicates, or if healthy subjects or not), use of control samples (yes/no) and method to determine GI, since synthesizing evidence summaries involves combining relevant and valid information [18]. The Wordcloud® tool was used to highlight the most mentioned ingredients and the studies used as gluten-replacements or starch sources.