| Design (n) | Casanova et al7 | Clini et al8 | McEvoy et al11 | Diaz et al15 | Windisch et al18 |
| RCT (n = 24 vs 20) | RCT (n = 47 vs 39) | RCT (n = 72 vs 72) | RCT (n = 18 vs 18) | Uncontrolled (n = 34) | |
| Age (years) | 68 vs 64 | 66 vs 64 | 69 vs 67 | 67 vs 67 | 63 |
| BMI (kg/m2) | 25 vs 25 | 25 vs 26 | 25.4 vs 25.5 | 25 vs 24.9 | 28.3 |
| FEV1 (l) | 0.87 vs 0.82 | – | 0.55 vs 0.63 | 0.81 vs 0.72 | 1.03 |
| FEV1 (% pred) | 31 vs 29 | 31 vs 27 | 23.1 vs 25 | 38 vs 32 | – |
| Pao2 (mm Hg) | 57.5 vs 55.7 | 49.5 vs 50 | 52.5 vs 54.8 | 50 vs 45 | 51.7 |
| Paco2 (mm Hg) | 53 vs 50 | 55.5 vs 54 | 54.4 vs 52.6 | 56 vs 57 | 53.3 |
| LTOT (months) | 21 vs 30 pre-trial | 29 vs 30 pre-trial | At least 3 | – | “Majority receiving” |
| SGRQ total score | – | 62 vs 66 | 64 vs 69 | – | |
| Hospital days for acclimatisation to NIV | At least 2 | 10 | 3–4 | – | 13 (6.9) |
| Compliance | 5.9 h/day11% <3 h/day | 9.2 h/night in compliant group (>5 h/night) | 4.5 h/day60% >5 h/night | 3 h/day for 5 days/week for 3 weeks | – |
| IPAP/EPAP (cm H2O) | 12/4 | 14/2 | 12.9/5.1 | 18/2 | 27.7 |
| Mode | Bilevel, spontaneous | Bilevel, timed,back-up rate 8 | Bilevel, spontaneous | Bilevel, spontaneous | Pressure-controlled |
| Therapeutic target for NIV | Reduced respiratory muscle activity | Reduced CO2 by day, Sao2 >90% overnight | Reduced sleep-disordered breathing | Maximum tolerated IPAP during day | Normalisation of Paco2 |
| Demonstration of physiological effect during NIV | – | – | + | +++ | +++ |
| 12-month survival | 78% vs 78% | – | 80% vs 72% | – | – |
| 24-month survival | – | 83% vs 82% | 68% vs 53% | – | 86% |
BMI, body mass index; EPAP, expiratory positive airway pressure; FEV1, forced expiratory volume in 1 s; IPAP, inspiratory positive airway pressure; LTOT, long-term oxygen therapy; NIV, non-invasive ventilation; Paco2, Pao2, arterial carbon dioxide and oxygen tensions; RCT, randomised controlled trial; SGRQ, St George Respiratory Questionnaire.