Optimizing the medical equipment investment in primary care centres in rural China: evidence from a panel threshold model

Table 3 Estimates of the effects of the medical equipment input on the service utilization in different THC subgroups

	THCs in Neighboring towns of county seat (N = 175)			THCs in remote towns (N = 144)
	Subgroup	β	95%CI	subgroup	β	95%CI
Impact of input
Number of equipment	Regime1	155.23	(-66.41, 376.88)	Regime1	1052.54**	(830.30, 1,274.79)
Number of equipment	(urban ≤ 69.89%)	155.23	(-66.41, 376.88)	(urban ≤ 5.99%)	1052.54**	(830.30, 1,274.79)
	Regime2	774.81**	(495.63, 1,053.98)	Regime2	237.00**	(74.42, 399.59)
	(urban > 69.89%)	774.81**	(495.63, 1,053.98)	(urban > 5.99%)	237.00**	(74.42, 399.59)
Impact of covariates
Population size		-0.23	(-0.49, 0.03)		0.02	(-0.25, 0.29)
Aging degree		-320.05	(-772.28, 132.18)		-94.97	(-386.62, 196.68)
Number of doctors		112.39	(-124.15, 348.94)		114.09	(-89.59, 317.78)
Number of technicians		1316.73	(-52.29, 2,685.75)		520.52	(-380.40, 1,421.44)
Intercept		40776.59**	(28,990.76, 52,562.42)		19188.49**	(12,262.49, 26,114.49)
Number of observation¹		875			720
F-value²		8.96		16.09
Prob > F		< 0.0001		< 0.0001

Note: *p < 0.05 **p < 0.01. (1) Number of observations: the number of observations in the panel data, where the data for the variables were updated on a yearly basis; (2) F test indicates the overall significance of estimated regression model, where the null hypothesis is all of the regression coefficients are equal to zero.

ISSN: 1472-6963