The impacts of Chinese drug volume-based procurement policy on the use of policy-related antibiotic drugs in Shenzhen, 2018–2019: an interrupted time-series analysis

Table 2 Results of the segmented linear regression models for the volume of winning and non-winning products

	Coefficient	Standard Error	t	p-value	95 % CI
	Coefficient	Standard Error	t	p-value	Lower	Upper
Model 1, Winning products
Secular trend, β₁	3.94	3.28	1.20	0.245	-2.92	10.80
Change in level, β₂	185.07	50.14	3.69	0.002	80.13	290.02
Change in trend, β₃	-0.82	8.01	-0.10	0.920	-17.58	15.95
Cold, β₄	73.25	31.36	2.34	0.031	7.61	138.89
Constant, β₀	87.74	30.10	2.92	0.009	24.75	150.73
Model 2, Non-winning products
Secular trend, β₁	0.33	0.98	0.33	0.744	-1.73	2.38
Change in level, β₂	-26.35	15.03	-1.75	0.096	-57.81	5.11
Change in trend, β₃	-4.27	2.40	-1.78	0.091	-9.30	0.75
Cold, β₄	23.97	9.41	2.55	0.020	4.27	43.66
Constant, β₀	59.08	9.02	6.55	0.000	40.21	77.95

Model 1, F = 25.18, p-value < 0.001, R² = 0.841, Adjusted R² = 0.808; Model 2, F = 13.02, p-value < 0.001, R² = 0.733, Adjusted R² = 0.676

ISSN: 1472-6963