Table 2. Effect of the full oxygen system intervention on mortality—Primary outcome analysis comparing the full oxygen system intervention to the prior standard of care plus pulse oximetry.
| Population and study period | Deaths | Mixed-Model Adjusted Odds Ratio (95% CI)α | ICC | |||
|---|---|---|---|---|---|---|
| n/N (%)* | Basic Model | Severity-Adjusted | (95% CI) | |||
| p-Value | p-Value | |||||
| Childβ | ||||||
| - Pulse oximetry | 227/6440 (3.5) | - | - | 0.03 | ||
| - Full O2 system | 422/9203 (4.6) | 1.06 (0.77–1.46) | 0.721 | 1.12 (0.79–1.59) | 0.527 | (0.01–0.07) |
| Child ALRIβ | ||||||
| - Pulse oximetry | 44/759 (5.8) | - | - | 0.16 | ||
| - Full O2 system | 97/1182 (8.2) | 1.09 (0.50–2.41) | 0.824 | 1.42 (0.60–3.36) | 0.427 | (0.06–0.36) |
| Neonate | ||||||
| - Pulse oximetry | 204/2627 (7.8) | - | - | 0.11 | ||
| - Full O2 system | 486/4360 (11.1) | 1.45 (1.04–2.00) | 0.026 | 1.76 (1.22–2.54) | 0.002 | (0.04–0.28) |
| Preterm/LBW neonate | ||||||
| - Pulse oximetry | 82/611 (13.4) | - | - | 0.07 | ||
| - Full O2 system | 203/1042 (19.5) | 1.30 (0.76–2.23) | 0.366 | 1.74 (0.97–3.11) | 0.062 | (0.02–0.24) |
Preterm defined as <37 weeks’ gestational age.
*Denominators vary according to the population included.
αMixed-model odds ratios account for the clustering of patients within hospitals and adjust for time trends. Under the stepped-wedge design, the adjusted odds ratios are calculated with the use of all data points in the intervention period versus the comparison period and therefore represent the average odds of exposure to the intervention.
βChild under 15 years of age, excluding neonates (<28 days of age).
Abbreviations: ALRI, acute lower respiratory infection, defined as cough or difficulty breathing and any of the following: fast breathing, lower chest wall indrawing; CI, confidence interval; ICC, intracluster correlation coefficient, indicating the degree of outcome similarity within clusters (hospitals); LBW, low birth weight, <2,500 g