Outcomes for non-randomised control trials
| Author, year | Total sample size | Mortality | Additional findings | Adverse events | Reported limitations | Reported conclusions |
| Balfour-Lynn, 201416 | 106 | 2 (2%) | N/A | 0 (0%) | Possibility of missing data | NIPPV can be a simple and cost-effective way to treat patients with acute respiratory failure |
| Bjorklund, 201917 | 83 | 8 (10%) | Patients with severe illness based on Tal score: 0 hours: 64/83 (77%) 2 hours: 12/83 (15%) | Severe: 0 Mild: 5 (6%) Mild AE included nasal tissue irritation and abdominal distension | Evaluations for complications based only on clinical exam, not powered to evaluate effectiveness, differences in pretrial and trial patients | SEAL-bCPAP is safe for treatment of respiratory distress in non-neonatal children in LMIC with a trend towards decreased mortality |
| Bonora, 201118 | 154 | Avoided intubation: 3.8% Required intubation: 38.8% | No need for intubation: 80/154 (52%) Duration of NIV (days, median (IQR)): Avoided intubation: 4 (2.25–6) Required intubation: 2 (1–4) Duration of hospital stay (days, median (IQR)): Avoided intubation: 6 (5–9) Required intubation: 13 (9–24) | Skin breakdown noted but number of adverse events not reported | Retrospective study design with no control group, no rigid protocol to determine when therapies should be escalated or discontinued | NIV avoided mechanical ventilation in a high proportion of children |
| Brown, 201327 | 1 | 0 (0%) | Duration of bCPAP: 4 days Duration of hospital stay: 6 days | 0 (0%) | Limitations not reported | A low-cost bCPAP could reduce child mortality in Africa |
| Figueroa, 201719 | 120 | Not reported | Success of bCPAP: 72% Duration of bCPAP (hours, mean (CI)): 75 (65–85) Duration of ICU stay (days, mean (CI)): 10 (9–11) | 4 (3%) Complications included abdominal bloating and pneumothorax | Limitations not reported | A reduction in respiratory rate, heart rate and TAL scores at 2 hours after starting intervention were predictors of success |
| Ghiggi, 200020 | 42 | 2 (5%) | Duration of nasopharyngeal CPAP (days, mean (SD)): 4.12 (3.71) Need for mechanical ventilation: 13/42 (31%) | 8 (19%) Complications included tube obstructions and apnoea due to excessive sedation | Small sample size | Nasopharyngeal CPAP was useful to avoid mechanical ventilation |
| Kinikar, 201121 | 36 | 0 (0%) | Duration of ICU stay (days, median (range)): 2 (2–5) Duration of hospital stay (days, median (range)): 7 (6–11) Decrease in mean RR after 6 hours: H1N1 positive: 20 H1N1 negative: 17 | 0 (0%) | Limitations not reported | Indigenous NB-CPAP improves hypoxemia and signs and symptoms in hemodynamically stable children with acute respiratory failure due to influenza-like injury |
| Lum, 201122 | 129 | 19 (15%) | Duration of NIV (days, median (IQR)): 4 (2–8) Duration of PICU stay (days, median (IQR)): 4.5 (2–9) Avoided mechanical ventilation for ≥5 days: 98 (76%) | 29 (22%) Complications included pneumonia while on NIV, pressure from mask and problems with mask fitting | Not an RCT, no routine use of blood gas sampling, shortage of NIV machines | NIV represents a viable strategy that provided effective respiratory support and prevented intubation in majority of patients |
| Machen, 201523 | 79 | 23 (29%) | Duration of CPAP (days, mean): 3.12 Duration of hospitalisation (days, mean): 8.41 Had lower RISC score after 24 hours: 63 (80%) | Not reported | Clinical diagnoses could have led to misclassification | bCPAP was most beneficial to patients with bronchiolitis |
| McCollum, 201128 | 1 | 0 (0%) | Duration of bCPAP (days): 7 | 0 (0%) | Limitations not reported | bCPAP was successful in treating an infant with PJP pneumonia secondary to HIV infection |
| Myers, 201924 | 117 | 38 (33%) | Required intubation: 15/115 (13%) Duration of treatment (hours, median (IQR)): 24 (24–60) | 13 (11%) Complications included blocked nostrils or nasal prongs, interruption of oxygen supply, nasal septum lesions and aspiration | Observational study design, small sample size, limited human resources and some missing data points | It is feasible to use bCPAP in the hospital management of critically ill children in resource-limited settings |
| Pulsan, 201925 | 64 | 35 (55%) | RDS (mean (IQR)): Pre-CPAP: 11 (10–12) 1 hour: 9 (8–11) 84 hours: 6.5 (6–8) | Not reported | Observational study design, bCPAP only used when oxygen failed | bCPAP improves oxygenation and reduces respiratory distress in some children but children with comorbidities continue to do poorly |
| Walk, 201626 | 77 | 36 (47%) | Duration of treatment (days, median (IQR)): 3 (3–5) | 13 (17%) | Non-randomised and uncontrolled, small sample size, understaffing, missing vital sign data | bCPAP can be feasibly implemented into a tertiary African hospital with high-risk patients |
AEs, adverse events; bCPAP, bubble continuous positive airway pressure; CPAP, continuous positive airway pressure; RR, respiratory rate in breaths per minute.