Skip to main content
. 2022 Jan 2;22:9. doi: 10.1186/s12913-021-07314-2

Table 8.

Estimates from segmented linear regression for adverse events in Site 1

Outcome Estimate, ß
(95% CI)
P Value Interpretation
Adverse events
 Intercept 3.08 (2.93, 3.24) < 0.00001
 Pre-intervention trend 0.02 (0.00, 0.03) 0.04 Adverse events were very gradually but significantly increasing during this period. Given the low overall rates the clinical impact of this increase is difficult to determine.
 Change in slope (implementation period vs. pre-intervention period) 0.03 (−0.03, 0.09) 0.29 There was a trend towards an increasing rate of adverse events (against the expected trend) but this was not significant. The wide confidence intervals mean the trend could have been in either direction should a greater sample size have been available.
 Immediate change in level (implementation period vs. pre-intervention period) 0.15 (− 0.34, 0.64) 0.55
 Change in slope (post-intervention period vs. implementation period) -0.09 (−0.15, − 0.05) < 0.001 Adverse event rates decreased by nearly 10% in this period, compared to the implementation period, which was statistically significant.
 Immediate change in level (post-intervention period vs. implementation period) −0.43 (−1.03, 0.17) 0.16