Segmented regression analysis of drug-cost data
| Pre-intervention mean (unadjusted) | Post-intervention mean (unadjusted) | ß1 – baseline trend (95% CI) | ß2 – step level change | ß3 – change in trend | |
|---|---|---|---|---|---|
| Total spend on anitbiotics per week (£) | 10,719 | 11,260 | –43.58 (–43.58 to –19.60) p<0.001 | +2,807.50 (+1,367.83 to +4,247.07) p<0.001 | – |
| Antibiotics spend per patient per week (£) | 26.44 | 25.40 | – | -3.75 (-6.52 to -0.98) p=0.008 | +0.10 (+0.02 to +0.18) p=0.015 |
| 100 mg oral doxycycline capsules spend per patient per week (£) | 0.020 | 0.023 | – | +0.003 (+0.001 to +0.005) p<0.001 | – |
| Total inhaler spend per week (£) | 309.29 | 227.74 | – | – | – |
| Inhaler spend per patient per week (£) | 5.29 | 4.17 | – | – | –0.03 (–0.06 to –0.01) p=0.11 |
β1: weekly change in spending before intervention.
β2: step change in weekly spend after intervention.
β3: absolute change in trend in weekly spend after intervention, compared with baseline trend.
β1 + β3 = β4 post-intervention slope.
Inhaler models include a dummy variable to adjust for a change in local prescribing policy. Only statistically significant variables are included in the parsimonious models. Gaps in the table reflect that variables were not significant and were not included in the final model.
CI = confidence interval