TABLE 3.

Results of interrupted time-series analysis in subgroups of participating hospitals.

Subgroups	Number of hospitals	Slope before intervention (95% CI)	Slope during intervention (95% CI)	P-value of test of difference in slopes	Change in level (95% CI)	P-value
Participating through complete intervention period	45	0.133 (0.03, 0.236)	−0.018 (−0.12, 0.085)	0.042	−1.133 (−3.405, 1.138)	0.328
Participating through complete intervention period and early implementation of quality managementa	8	−0.089 (−0.345, 0.167)	−0.035 (−0.29, 0.22)	0.771	2.841 (−2.784, 8.466)	0.323
Not participating through complete intervention period	29	−0.076 (−0.193, 0.041)	0.165 (−0.085, 0.415)	0.084	−1.67 (−5.525, 2.184)	0.396
Number of beds ≤700	40	0.017 (−0.117, 0.152)	0.02 (−0.153, 0.194)	0.98	−0.997 (−4.468, 2.474)	0.573
Number of beds >700	34	−0.015 (−0.073, 0.042)	0.047 (−0.033, 0.127)	0.21	−0.285 (−1.828, 1.257)	0.717

Results of piecewise hierarchical models on the difference in the risk-standardized mortality rate (RSMR) in patients with sepsis between GQNS hospitals and the national German diagnosis-related-groups statistic considering different subgroups. Slopes give the linear trajectory of RSMR-difference in % per month across time before and after start of the intervention, change in level gives the change at the time of the beginning of the intervention. Time of beginning of the intervention is defined for each individual hospital as the time of supply of the first quality report. ^aEarly implementation of quality management was defined based on the survey of local quality improvement leaders of participating hospitals in autumn of 2016, if the implementation of a quality improvement team as well as analyses of quality reports was reported. Survey data were available for 28 of 45 hospitals (62%).