Table 6.
Study PMID |
Baseline AHI (SD) [eligibility] |
Baseline ESS (SD) | Duration (design) | Interventions | Number analyzed | Baseline (SD) | Change (final) |
Net difference or difference |
95% CIa | P | Dropout (%) | Study quality |
---|---|---|---|---|---|---|---|---|---|---|---|---|
d'Ortho 2000 [32] 11035671 |
58 (6) [≥10] |
12.7 (5.3) | 2 months (XO) |
APAP | 25 | 66.5 (13.6) |
85.2 | -1.4 | nd | nd | 0 | B |
CPAP | 25 | 66.5 (13.6) |
86.6 | |||||||||
Galetke 2008 [12] 17148931 |
33 (19) [> 10] |
10.3 (5.7) | 2 months (XO) |
APAP | 20 | 77.8 (8.4) | 8.7 | -1.8 | -3.8 to 0.2 | nd | nd | C |
CPAP | 20 | 77.8 (8.4) | 10.5 | |||||||||
Hussain 2004 [20] 15072173 |
47 (36) [> 15] |
11.1 (6.4) | 1 month (XO) |
APAP | 10 | 67.8 (12.5) | 14 | -3.9 | -7.3 to -0.5 | nd | 0 | C |
CPAP | 10 | 67.8 (12.5) | 17.9 | |||||||||
Konermann 1998 [31] 9515848 |
38 (14) [> 20} |
nd | 3 to 6 months (PL) |
APAP | 25 | 76.5 (12.4) | 90.3 | 1.1 | nd | NS | 4 | B |
CPAP | 23 | 74.5 (10.7) |
87.2 | |||||||||
Meurice 2007 [13] 17638595 |
55 (10) [nd] |
11.8 (4.9) | 6 months (PL) |
APAP (AutoSet) | 15 | 82.1 (12.8) | 0.2 | -2.9 | -7.1 to 7.5 | nd | 15 | B |
CPAP | 14 | 82.3 (9.9) |
0.5 | |||||||||
Nolan 2007 [14] 17326544 |
15 (8) [≥5) |
12.3 (4) | 2 months (XO) |
APAP | 29 | 79 (11.5) | 8.5 | 4.8 | -7.4 to 17.0b | 0.44 | 15 | B |
CPAP | 29 | 79 (11.5) | 3.7 | |||||||||
Patruno 2007 [16] 17494789 | 46 (14) [> 20] |
15 (2.7) | 3 months (PL) |
APAP | 15 | 71.7 (10.6) | 16.4 | -4.4 | -11.8 to 3.0 | nd | 23 | C |
CPAP | 16 | 70.0 (11.7) |
20.8 | |||||||||
Randerath 2001 [26] 11254519 |
35 (26) [≥10] |
11.1 (5.1) | 1.5 months (XO) |
APAP | 52 | 81 (8.0) | 7.0 | -1.0 | -2.1 to 0.1 | nd | 12 | B |
CPAP | 52 | 81 (8.0) | 8.0 | |||||||||
Resta 2004 [24] 15679008 |
47 (11) [> 30] |
13.9 (3.2) | 1 month (PL) |
APAP | 10 | 72.4 (10.5) | 15.9 | 0.9 | -7.4 to 9.2 | nd | 0 | C |
CPAP | 10 | 74.1 (10.8) | 15.0 |
AHI: apnea-hypopnea index (events/hour); APAP: auto-titrating CPAP; CPAP: continuous positive airway pressure; ESS: Epworth Sleepiness Scale; nd: no data; NS: non-significant; PL: parallel design; XO: cross-over design. aEstimated from reported data. bEstimated from reported P value.