Cardiogenic shock in Taiwan from 2003 to 2017 (CSiT-15 study)

Chien, Shih-Chieh; Hsu, Chien-Yi; Liu, Hung-Yi; Lin, Chao-Feng; Hung, Chung-Lieh; Huang, Chun-Yao; Chien, Li-Nien

doi:10.1186/s13054-021-03820-1

Critical Care

Table 2 The marginal effect of AMI accreditation on in-hospital mortality of AMI-CS patients, by level of hospital

From: Cardiogenic shock in Taiwan from 2003 to 2017 (CSiT-15 study)

	Overall			Medical center			Regional hospital			District hospital
	dy/dx	(95% CI)	P	dy/dx	(95% CI)	P	dy/dx	(95% CI)	P	dy/dx	(95% CI)	P
Policy: 2009
t	− 0.016	(− 0.022, − 0.010)	< .0001*	− 0.007	(− 0.012, − 0.002)	0.004*	− 0.021	(− 0.031, − 0.012)	< .0001*	− 0.006	(− 0.025, 0.013)	0.514
P_t	0.005	(− 0.019, 0.028)	0.688	− 0.003	(− 0.021, 0.015)	0.708	0.012	(− 0.021, 0.045)	0.477	0.024	(− 0.054, 0.101)	0.549
[t − T_I ]P_t	0.005	(− 0.002, 0.012)	0.171	0.002	(− 0.003, 0.008)	0.344	0.009	(− 0.001, 0.019)	0.078	− 0.007	(− 0.028, 0.014)	0.519
Policy lag: 2010
t	− 0.015	(− 0.020, − 0.010)	< .0001*	− 0.008	(− 0.012, − 0.004)	< .0001*	− 0.017	(− 0.025, − 0.010)	< .0001*	− 0.008	(− 0.023, 0.008)	0.323
P_t	0.006	(− 0.017, 0.029)	0.612	0.004	(− 0.013, 0.022)	0.624	0.002	(− 0.030, 0.035)	0.897	0.040	(− 0.039, 0.118)	0.321
[t − T_I]P_t	0.004	(− 0.002, 0.009)	0.224	0.003	(− 0.002, 0.007)	0.234	0.006	(− 0.003, 0.015)	0.164	− 0.008	(− 0.027, 0.011)	0.385
Policy lag: 2011
t	− 0.016	(− 0.020, − 0.012)	< .0001*	− 0.008	(− 0.012, − 0.005)	< .0001*	− 0.016	(− 0.022, − 0.010)	< .0001*	− 0.010	(− 0.023, 0.003)	0.120
D_t	0.023	(0.000, 0.046)	0.051	0.021	(0.004, 0.038)	0.016	− 0.007	(− 0.040, 0.026)	0.676	0.110	(0.031–0.189)	0.006
[t − T_I]P_t	0.002	(− 0.004, − 0.008)	0.489	0.001	(− 0.003, 0.005)	0.624	0.006	(− 0.002, 0.014)	0.165	− 0.020	(− 0.039, − 0.001)	0.041*
Policy lag: 2012
t	− 0.013	(− 0.016, 0.010)	< .0001*	− 0.007	(− 0.010, 0.004)	< .0001*	− 0.014	(− 0.019, − 0.010)	< .0001*	− 0.005	(− 0.015, 0.006)	0.376
D_t	0.005	(− 0.018, 0.028)	0.678	0.009	(− 0.010, 0.028)	0.364	− 0.015	(− 0.047, 0.018)	0.382	0.088	(0.006, 0.171)	0.035*
[t − T_I]P_t	0.003	(− 0.003, 0.009)	0.292	0.001	(− 0.003, 0.006)	0.559	0.008	(− 0.000, 0.016)	0.056	− 0.029	(− 0.051, 0.008)	0.006*
Policy lag: 2013
t	− 0.012	(− 0.015, − 0.009)	< .0001*	− 0.007	(− 0.009, 0.004)	< .0001*	− 0.014	(− 0.018, − 0.010)	< .0001*	− 0.010	(− 0.023, 0.003)	0.120
D_t	− 0.009	(− 0.034, 0.016)	0.473	0.004	(− 0.017, 0.024)	0.723	− 0.022	(− 0.058, 0.014)	0.224	0.110	(0.031–0.189)	0.006
[t − T_I]P_t	0.005	(− 0.002, 0.012)	0.174	0.001	(− 0.005, 0.007)	0.677	0.010	(− 0.001, 0.020)	0.054	− 0.026	(− 0.051, 0.001)	0.040*

Time variable referred as t was beginning at 1 in the year of 2002 and incrementing by 1 up to the year of 2017; policy variable referred as P_t (an intervention time indicator) was coded 0 pre-intervention and 1 post-intervention; the marginal effect of the policy referred as [t − T_I]P_t (a slope change variable) was equal to zero at the time of the intervention (T_I) and incrementing by 1 up to the year of 2017. dx/dy referred as marginal effect was estimated by using multi-level mixed effect logistic regression models which had a random intercept for center and adjusted for individual-level demographics and disease history listed in Table 1
*P < 0.05

Back to article page

ISSN: 1364-8535

Contact us

Submission enquiries: journalsubmissions@springernature.com