Impacts of the diagnosis-intervention packet reform on costs and healthcare resource utilization: evidence from Guangzhou, China

Table 4 The dynamic impacts of DIP on costs, LOS, and IHM using the SDID method without covariates (2018–2020)

	Per-episode inpatient costs (CNY)	Average length of stay (days)	In-hospital mortality (%)
ATT (2018–2020)	1574.735** [148.330, 3001.140]	0.006 [−1.362, 1.374]	−0.031 [−0.185, 0.122]
ATT₁ (2018)	603.296 [−241.000, 1450.000]	0.691 [−0.163, 1.545]	−0.005 [−0.090, 0.080]
ATT₂ (2019)	1662.495** [198.425, 3130.000]	−0.705 [−2.138, 0.728]	−0.083 [−0.263, 0.097]
ATT₃ (2020)	2889.953*** [719.63588, 5060.000]	0.033 [−1.962, 2.028]	−0.005 [−0.235, 0.226]

Abbreviations: DIP diagnosis-intervention packet, LOS length of stay, IHM in-hospital mortality, SDID synthetic difference-in-differences, ATT average treatment effect
^***, **, and * denote the significance at the 1%, 5%, and 10% level. 95% confidence intervals in brackets. The SDID estimates use panel data from the other 20 cities in Guangdong Province to synthesize Guangzhou. 2018 was the first year of DIP implementation, 2019 was the second year of DIP implementation, 2020 was the third year of DIP implementation

ISSN: 2191-1991