Strategy for data synthesis
When possible, data extracted for sensitivity and specificity will be
analysed using random effects meta-analysis using hierarchical summary
receiver operating characteristic (ROC) curve analysis in case of three
or more studies for a given combination of index test and food.
The GRADE approach will be used to assess
heterogeneity13,14, and will not solely rely on
statistical methods such as the I2 statistic.
Depending on the included studies, the following reasons for
heterogeneity will be discussed:
- Multiple definitions of target condition
- Multiple thresholds of test positivity for index and/or comparator
- Age of the participants
- Origin of study population as indication for patient spectrum/pre-test
probability
- All domains of risk of bias of the included studies
The GRADE approach will be used to evaluate the certainty of the body of
evidence.