Table 2. Antibiotic prescription and strategy failure.
| Analysis and outcome | Number/total (percentage) | Unadjusted | Adjusted | |||
|---|---|---|---|---|---|---|
| Pre-intervention | Intervention | OR* (95% CI) | p-Value† | OR‡ (95% CI) | p-Value† | |
| Primary analyses | ||||||
| Intention-to-treat population | ||||||
| Antibiotic prescription | 179/597 (30%) | 101/402 (25%) | 1.06 (0.61–1.85) | 0.84 | 1.07 (0.57–2.01) | 0.75 |
| Strategy failure | 131/572 (23%) | 61/381 (16%) | 0.56 (0.34–0.93) | 0.02 | 0.53 (0.32–0.88) | 0.01 |
| Per-protocol population | ||||||
| Antibiotic prescription | 179/597 (30%) | 83/359 (23%) | 0.89 (0.5–1.61) | 0.71 | 0.96 (0.49–1.88) | 0.92 |
| Strategy failure | 131/572 (23%) | 57/340 (17%) | 0.60 (0.36–1.00) | 0.05 | 0.56 (0.34–0.93) | 0.03 |
| Sensitivity analyses | ||||||
| Truncated baseline and post-rollout periods§ | ||||||
| Antibiotic prescription | 66/276 (24%) | 64/279 (23%) | 0.81 (0.45–1.46) | 0.48 | 0.71 (0.38–1.32) | 0.27 |
| Strategy failure | 58/261 (22%) | 46/269 (17%) | 0.57 (0.35–0.94) | 0.03 | 0.54 (0.33–0.90) | 0.02 |
| Strategy failure, including missing values | ||||||
| Assumption missing = failure | 156/597 (26%) | 82/402 (20%) | 0.56 (0.36–0.88) | 0.01 | 0.55 (0.35–0.87) | 0.01 |
| Assumption missing = no failure | 131/597 (22%) | 61/402 (15%) | 0.59 (0.36–0.96) | 0.03 | 0.56 (0.34–0.91) | 0.02 |
| Secondary analyses | ||||||
| Compliance (Feverkidstool calculated and patient treated according to advice) | NA | 359/402 (89%) | ||||
| Complications¶ | 1/572 (0.1%) | 1/381 (0.2%) | ||||
| Strategy failure: reasons | ||||||
| Secondary antibiotic prescription | 45/572 (8%) | 29/381 (8%) | ||||
| Changed antibiotics during follow-up | 14/572 (2%) | 5/381 (1%) | ||||
| Secondary hospitalisation** | 16/572 (3%) | 13/381 (3%) | ||||
| Oxygen need at day 7 | 9/572 (2%) | 1/381 (0.2%) | ||||
| Fever at day 7 | 47/572 (8%) | 13/381 (3%) | ||||
Bolding indicates statistical significance.
*Main model: clustered by hospital, adjusted for time period. Time-adjusted intracluster correlation coefficient for antibiotic prescription = 0.04, for strategy failure = 0.
†p-Values based on multivariable logistic regression.
‡Adjusted model: main model further adjusted for age, sex, season, ill appearance, and duration of fever.
§Using data from 4 weeks before until 4 weeks after the rollout period, resulting in 9 time periods of equal length, truncating the prolonged baseline and post-rollout periods.
¶Complications were 1 admission to intensive care unit in the pre-intervention phase and 1 pleural empyema in the intervention phase (both unrelated to study intervention).
**Including 1 admission to the intensive care unit in the pre-intervention group.
NA, not applicable.