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 3 Results of the segmented linear regression models for the volume of Cefuroxime and its alternative drugs

	Coefficient	Standard Error	t	p-value	95 % CI
	Coefficient	Standard Error	t	p-value	Lower	Upper
Model 1, Cefuroxime
Secular trend, β₁	4.12	3.09	1.33	0.199	-2.35	10.59
Change in level, β₂	161.16	48.61	3.32	0.004	59.43	262.90
Change in trend, β₃	-5.41	7.52	-0.72	0.480	-21.16	10.33
Cold, β₄	89.67	32.04	2.80	0.011	22.62	156.72
Constant, β₀	149.22	28.04	5.32	0.000	90.54	207.89
Model 2, Alternative drugs
Secular trend, β₁	20.52	5.42	3.79	0.001	9.18	31.86
Change in level, β₂	273.65	87.66	3.12	0.006	90.17	457.12
Change in trend, β₃	-47.57	12.91	-3.69	0.002	-74.59	-20.55
Cold, β₄	313.94	65.63	4.78	0.000	176.56	451.31
Constant, β₀	634.46	47.66	13.31	0.000	534.70	734.22
Model 3, Total
Secular trend, β₁	24.70	7.25	3.41	0.003	9.52	39.88
Change in level, β₂	436.31	117.29	3.72	0.001	190.81	681.81
Change in trend, β₃	-54.09	17.30	-3.13	0.006	-90.31	-17.88
Cold, β₄	385.09	87.21	4.42	0.000	202.55	567.63
Constant, β₀	786.20	63.89	12.30	0.000	652.47	919.94

Model 1, F = 19.63, p-value < 0.001, R² = 0.805, Adjusted R² = 0.764; Model 2, F = 32.63, p-value < 0.001, R² = 0.873, Adjusted R² = 0.846; Model 3, F = 36.12, p-value < 0.001, R² = 0.884, Adjusted R² = 0.859

ISSN: 1472-6963