Skip to main content

Table 2 Results of DiD analysis

From: The 2010 expansion of activity-based hospital payment in Israel: an evaluation of effects at the ward level

DiD coefficient (δ) lndis lnALoS
Model 1 Model 2 Model 1 Model 2
Estimate δ (CI) Estimate δ (CI) Estimate δ (CI) Estimate δ (CI)
Wave1*period1 0.068 (− 0.011 0.147) 0.068 (− 0.011 0.147) 0.017 (− 0.014 0.048) 0.017 (− 0.014 0.048)
Wave1*period2 0.026 (− 0.141 0.193) 0.026 (− 0.141 0.193) −0.001 (− 0.087 0.084) − 0.001 (− 0.087 0.084)
Wave2*period2 −0.097 (− 0.291 0.097)     − 0.055 (− 0.141 0.030)    
General surgery*period2     −0.030 (−0.194 0.134)     −0.043 (− 0.126 0.039)
urology*period2     0.030 (−0.103 0.164)     −0.063 (−0.139 0.012)
Ophthalmology*period2     −0.151 (−0.307 0.005)     −0.025 (− 0.105 0.054)
head neck*period2     −0.239* (−0.409 − 0.070)     −0.086* (− 0.168 −0.004)
R Square 0.48    0.59    0.47    0.51   
Number of cases 1828    1828    1828    1828   
number of hospitals 29    29    29    29   
  1. Notes: lndis = natural logarithm of number of discharges, lnALoS = natural logarithm of average length of stay, wave 1 = orthopedics, wave 2 = general surgery, urology, \’/ ophthalmology, head and neck surgery; period 1 = 2011–2013; period 2 = 2014–2015; CI = 95% Confidence Interval in parenthesis; *p < 0.05; **p < 0.01. Regression OLS, clustered by ward. ALoS were weighted by ward size. Model 1 includes reform waves as predictors, model 2 includes each ward separately as predictors