Table 1 Direction of bias on study outcome for each key question
Bias direction | Within domain/key question | Methodological alternatives considered a source of high risk of bias |
|---|---|---|
In favour of an effect of exposure (+) | Performance bias: - Was the level and method of blinding appropriate? | - no blinding of research personal during sessions - insufficient removal of any clues that could reveal exposure status and no tests done to control blinding |
In favour of a null result (−) | Selection bias: | |
- Were individuals excluded whose symptoms may be explained by somatic diseases or mental disorders? | - not sufficiently considered/not reported | |
- Was the contrast in the severity of symptoms between situations with/without exposure verified? | - not reported | |
- Were EMF exposures (type of exposure source, frequency range and exposure level) applied that individuals associate with their symptoms? | - not reported | |
- Were exposure durations and assessment times applied that matched the time scales for the symptoms to appear? | - not reported | |
- Were the symptoms registered in the trials matched with those experienced in everyday exposure situations? | - not reported | |
Exposure bias: | ||
- Was the background exposure level controlled and minimized? | - not reported | |
- Was the exposure level controlled? | - not reported | |
Uncertain direction on study outcome (±) | Selection bias: | |
- Were the intervals between exposure sessions sufficiently long to allow for recovery and to avoid carry-over effects? | - not reported | |
Performance bias: | ||
- Were biases related to sequence and period of the exposure conditions minimized (for studies with cross-over design)? | - same sequence and period of the exposure conditions for all participants or for all participants of a group - not reported | |
Confounding bias: | ||
- Were biases related to confounders and cofactors minimized (for studies comparing parallel groups of IEI-EMF participants with different exposure conditions)? | - not randomized | |
Attrition bias: | ||
- Were biases minimized that are related to attrition and to incomplete data included in the analysis? | - high attrition/exclusion rate or incomplete data in analysis | |
Selective reporting bias: | ||
- Was bias related to selective outcome reporting minimized? | - selective outcome reporting | |