Table 8.

The effect of the PC organizational change, difference-in-difference estimations selecting different minimum cell sizes excluding the years 2010 and 2011

	(1)	(2)
	Min cell size = 20	Min cell size = 40
Days in hospital	-0.043***	-0.085***
	[0.011]	[0.013]
No.Obs.	9,900	8,695
R-squared	0.962	0.972
Re-hospitalization rate (same MDC)	-0.009***	-0.009***
	[0.002]	[0.003]
No.Obs.	9,900	8,695
R-squared	0.962	0.972
Re-hospitalization rate (same Hosp., same MDC)	-0.009***	-0.005***
	[0.002]	[0.002]
No.Obs.	9,900	8,695
R-squared	0.907	0.927
Mortality rate	-0.001	-0.002
	[0.002]	[0.002]
No.Obs.	9,900	8,695
R-squared	0.922	0.942
MDC x Year interaction	Yes	Yes
Hosp. and MDC Interactions with:
Discharges	Yes	Yes
Age	Yes	Yes
Male	Yes	Yes

Note: The table shows the estimate of the coefficient of the PC variable, for different variables, and respectively the log of the number of days of hospitalization (days in hospital), the log of the rate at which a discharged patient happens to be re-hospitalized in the same MDC within 30 days (re-hospitalization rate, same MDC), the log of the rate at which a discharged patient happens to be re-hospitalized in the same hospital and MDC within 30 days (re-hospitalization rate, same Hosp. and same MDC) and the log of the mortality rate within 30 days after discharge (mortality rate). All regressions are estimated always controlling for the year fixed effects, average number of discharged patients, average age and share of male patients for MDC j, hospital h and year t. Additional interaction terms have been added sequentially, as shown at the bottom of the table. Standard errors have been clustered at the hospital level

* 10% significance level, ** 5%, *** 1%