Table 3.

Analysed risk of bias domains (based on¹⁸ and²³) described in this publication.

Bias domain	Type of bias	Operationalisationa
Baseline characteristics	Selection bias	Where the CF and control subjects comparable at baseline, at least for age and sex?
Housing	Performance bias	Was housing of animals mixed over the cages or adequately randomised/counterbalanced?
Blinding	Performance bias	Were all involved in contact with the subjects adequately blinded?
Outcome assessment	Detection bias	Was the order of measuring CF and control subjects adequately randomised/counterbalanced?
Blinding	Detection bias	Were those performing the nPD and those analysing the traces adequately blinded?
Missing outcome data	Attrition bias	Were missing outcome data transparently reported and equally distributed over CF and control groups?
Selective outcome reporting (“cherry picking”)	Reporting bias	Was a protocol posted before study start, and were all planned outcomes reported?
Other	Other bias	Were there no other reasons for concerns of biassing the results?

^aYes reflects a low risk of bias, no reflects a high risk of bias, unclear reflects insufficient information to be sure.