eTable 4. Overall results for the comparison of intraoral appliances vs placebo on the AHI.

	Data	Trials	Effect	95% CI	p	95% PrI	τ2	I2 (%)
Meta-analysis	MD	12	–11.69	[–15.38; –8.01]	<0.001	–22.55; –0.85	20.15	93.6
Factor	Data	Trials	Effect	95% CI	p		τ2	I2 (%)
Follow-up (months)	Coefficient	11	0.25	[–0.24; 0.75]	0.273		21.23	94.7
	Constant		–14.40	[–21.20; –7.61]	0.001
% of maximum protrusion	Coefficient	8	0.03	[–0.42; 0.48]	0.87		34.21	96.3
	Constant		–14.18	[–48.70; 20.34]	0.354
Appliance type (thermoplastic vs impression-based)	Coefficient	13	6.25	[–3.73; 16.24]	0.196		17.12	89.4
	Constant		–12.17	[–15.68; 8.66]	<0.001
Appliance type (1- or 2-piece)	Coefficient	9	–0.09	[–11.25; 11.08]	0.986		15.79	87.9
	Constant		–14.24	[–24.81; –4.30]	0.012
Baseline-BMI (kg/m2)	Coefficient	12	1.94	[–0.84; 4.72]	0.152		17.42	88.7
	Constant		–68.59	[–150.39; 13.21]	0.091
Baseline AHI (events/hour)	Coefficient	11	–0.53	[–0.72; –0.34]	<0.001		2.84	46.3
	Constant		1.17	[–3.64; 5.99]	0.595
Baseline age (years)	Coefficient	12	0.44	[–1.05; 1.93]	0.524		21.46	91.3
	Constant		–33.59	[–107.42; 40.25]	0.335
Ratio of male patients	Coefficient	12	–28.65	[–78.34; 21.05]	0.228		18.8	93.7
	Constant		10.39	[–28.05; 48.82]	0.561
Total sample	Coefficient	12	0.14	[0.04; 0.24]	0.01		9.78	73.7
	Constant		–19.13	[–25.29; –12.98]	<0.001
Study design (crossover or parallel)	Coefficient	12	–0.52	[–8.85; 7.81]	0.892		22.5	94.2
	Constant		–11.38	[–18.25; 4.51]	0.004
Data type 1	Coefficient	12	–1.20	[–10.98; 8.58]	0.79		22.39	94.2
	Constant		–11.49	[–15.81; –7.17]	<0.001
Data type 2	Coefficient	12	8.07	[–1.72; 17.86]	0.096		15.9	89.2
	Constant		–12.77	[–16.47; –9.07]	<0.001
Data type 1	Coefficient	12	0.22	[–9.36; 9.79]	0.274		18.15	90.2
Data type 2	Coefficient		8.12	[–2.63; 18.88]
	Constant		–12.82	[–17.25; –8.39	<0.001
	Data	Trials	Effect	95% CI	p
Reporting bias (Egger’s test)	Coefficient	12	–1.24	[–5.35; 2.87]	0.516

CI. confidence interval; PrI. predictive interval; MD. mean difference; BMI. body mass index; Data type 1. origin of data used in the analysis 1 (increment calculated from parallel or cross-over trials); Data type 2. origin of data used in the analysis 2 (increment calculated from final values of cross-over trials)