Bignold 2011.

Methods	Randomised controlled cross‐over trial
Participants	POSA Participants randomised: 16 Gender: male:13; female: 2 Age (years): 58.2 ± 13.9 Inclusion criteria People with POSA Overall AHI ≥ 15 Supine AHI twice the non‐supine AHI ≥ 20 min or more of sleep in supine and non‐supine postures Non‐supine AHI < 15 Exclusion criteria Exclusion criteria for existing treatments for OSA (e.g. ventilator requirement) Mobility‐limiting problems inhibiting lateral sleep Cardiac pacemaker use
Interventions	Intervention: positional monitoring and supine alarm device in active mode Control: device in inactive mode
Outcomes	AHI, snoring loudness, percentage of time in supine posture
Notes	Setting: home‐based regimen Duration of treatment: 7 days; 1 week active vibration and 1 week inactive vibration in random order, separated by an intervening washout week No period effect as demonstrated using linear mixed‐model analysis with an auto regressive covariance structure (using supine time data among participants ‐ i.e. the control phase data) Funding: Flinders Medical Centre foundation grant; not industry sponsored Comment: study results are applicable only to POSA.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	The procedure for randomisation is not mentioned. High risk may be mitigated by the cross‐over design of the study.
Allocation concealment (selection bias)	Unclear risk	Allocation concealment not explicitly stated. The contribution of allocation concealment to the risk of bias is low by virtue of the design of the study
Blinding of participants and personnel (performance bias) Objective outcomes ( AHI)	High risk	Quote: "It is difficult to blind patients to active versus inactive treatment" (Protocol first paragraph page 378) Comment: likely unblinded. Participant blinding not possible
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Blinding not mentioned
Incomplete outcome data (attrition bias) All outcomes	Low risk	No attrition
Selective reporting (reporting bias)	Unclear risk	Unclear, as we could not access the protocol of the study
Other bias	Low risk	None identified