TABLE VIII.D. 1.

Evidence for positional therapy for positional OSA

Study		Year	LOE	Study design		Study groups	Clinical endpoint		Conclusion
Eijsvogel et al.1396		2015	2b	RCT		55 patients with positional OSA (defined as supine AHI ≥ 2 times nonsupine AHI and non-supine AHI < 10) with chest worn device (n = 29, SPT, NightBalance) vs. tennis ball technique (n = 26)	AHI, compliance (use ≥4 h/day and ≥5 days/week)		Equal reduction AHI (SPT: 11.4 vs. 3.9, TBT: 13.1 vs. 5.8), better compliance in SPT arm (SPT: 75.9% vs. TBT: 42.3%, p = 0.01).
Laub et al.1397		2017	1b	RCT		101 patients with positional OSA (defined as supine AHI ≥2 times nonsupine AHI, supine AHI > 10 and non-supine AHI < 10) with chest worn device (n = 52, SPT, NightBalance) vs. no treatment (n = 49)	AHI, % supine sleep, sleepiness, compliance (SPT use >4 h/night all week)		At 2 months, AHI improved in the SPT group (from 18.1 to 10.4 on per-protocol analysis, p = 0.008) and remained unchanged in the control group (p = 0.2). Percentage of time supine improved in both groups and ESS did not change in either group. Compliance with SPT was 75.5%.
Benoist et al.1392		2017	2b	RCT		99 patients with positional OSA (defined as supine AHI ≥ 2 times nonsupine AHI) with chest worn device (n = 48, SPT, NightBalance) vs. oral appliance therapy (n = 51)	AHI, compliance		SPT and oral appliance equally effective. At 3 months, AHI decreased in both groups (SPT: 13.9–8.7, oral appliance: 13.2–8.1, p < 0.001 for both). Only SPT decreased percentage of time in supine sleep (42.4–14.2, p < 0.001).
de Ruiter et al.1395		2017	1a	RCT		58 patients with positional OSA (defined as supine AHI ≥ 2 times nonsupine AHI) with chest worn device (n = 29, SPT, NightBalance) vs. oral appliance therapy (n = 29)	AHI, compliance		SPT and oral appliance equally effective. At 12 months, AHI decreased in both groups (SPT: 13.2–7.1, oral appliance: 13.4–5.0, p < 0.001 for both). Only SPT decreased percentage of time in supine sleep (41.6–12.7, p < 0.001).
Berry et al.1393		2019	1b	Non-inferiority RCT, cross over		117 patients with positional OSA (defined as AHI ≥15 or AHI 10–15 with ESS > 10 and supine AHI ≥ 2 times nonsupine AHI) with CPAP vs. chest worn device (SPT, NightBalance) followed by crossover	Adherence, AHI, PROMs		SPT non-inferior to CPAP with improved AHI in both groups (SPT: 21.5–7.3, CPAP: 21.5–3.7). SPT had greater nightly adherence than CPAP (345.3 vs. 286.98 min, p < 0.0001).
Mok et al.1403	2020		2b	Non-inferiority RCT, crossover	40 patients with ESS 10–16, positional OSA (defined as supine AHI ≥ 2 times nonsupine AHI, supine AHI > 10 and non-supine AHI < 10) with CPAP vs. neck worn device (Night Shift)		AHI, ESS	With regard to decrease in AHI, PT inferior to CPAP (PT: 23.4–13.0, CPAP: 23.4–4.0, p < 0.001). With regard improvement in ESS, PT noninferior to CPAP.
Srijithesh et al.1405	2019			SR, Cochrane	323 patients with OSA treated with positional therapy vs. CPAP (n = 72) or positional therapy vs. inactive control (n = 251)		AHI, ESS	Positional therapy was less effective than CPAP for reducing AHI (6.4 fewer events per hour with CPAP, 95% CI 3.00–9.79; low-certainty evidence) but compliance may be higher. Positional therapy was more effective than inactive control in reducing AHI and ESS.
van Maanen et al.1401	2013		2b	Single arm	36 patients with positional OSA (defined as supine AHI > 2 times nonsupine AHI) with chest worn device (Sleep Position Trainer [SPT]NightBalance)		AHI, sleep efficiency, ESS, FOSQ, compliance	At 1 month, the median percentage of supine sleeping time decreased from 49.9% to 0.0% (p < 0.001) and median AHI decreased from 16.4 to 5.2 (p < 0.001).
van Maanen et al.1402	2012		2b	Single arm	36 patients with positional OSA (defined as supine AHI > 2 times nonsupine AHI) with simple neck worn device		AHI, sleep quality	Mean AHI decreased from 27.7 to 12.8 (p < 0.01) and mean percentage of supine sleeping time decreased from 40% to 19% with use of the device.
van Maanen et al.1400	2014		2b	Single arm	A total of patients with positional OSA (defined as supine AHI > 2 times nonsupine AHI) and data from chest worn device (SPT, NightBalance)		Sleep efficiency, ESS, FOSQ, compliance	Long-term reduction of percentage of supine sleep from 21% to 3% (p < 0.001) at 6 months and improvement of ESS and QOL.
Levendowsky et al.1398	2014		1b	Single arm	30 patients with positional OSA (defined as supine AHI ≥ 1.5 times nonsupine AHI) with neck worn device (NightShift)		AHI, SO2, depression score, sleep architecture	Improvement of mean AHI from 24.7 to 7.5 (p < 0.00001), decrease in percentage of time O2 sat <90% (from 4.2 to 1.2, p < 0.01), and improved sleep architecture.
Beyers et al.1394	2019		1b	Single arm	58 patients with positional OSA (defined as supine AHI ≥ 2 times nonsupine AHI) with chest worn device (SPT, NightBalance)		Long-term AHI, % supine sleep, sleep quality, compliance	At 1 year, 85% of patients were still using the device and 75% reported subjective improvement in sleep quality. Mean AHI improved from 16.4 to 6.2 (p < 0.001), percentage of time in supine sleep was reduced from 37.1% to 1.4% (p < 0.001).
Dieltjens et al.1404	2015		1b	RCT	20 patients with residual positional OSA after mandibular advancement device use (defined as supine AHI ≥ 2 times nonsupine AHI, supine AHI>10 and non-supine AHI<10) with chest worn device (SPT, NightBalance) vs. combination of chest worn device and mandibular advancement device		AHI, compliance	SPT and MAD equally effective at reducing AHI and ODI, combination more effective with decrease in AHI from 20.9 to 5.5, p < 0.008).