Skip to main content

Table 3 ICCs hospital level and C-statistics of the models per diagnosis group of the index admission, N = 53 hospitals

From: To what degree can variations in readmission rates be explained on the level of the hospital? a multilevel study using a large Dutch database.

Diagnosis group (CCS1 code)

ICC hospital (% and 95 CI)

C-statistic model with random effect hospital (95% CI)

C-statistic model without random effect hospital (95% CI)

Biliary tract disease (149)

0.48 (0.23–0.68)

0.651 (0.644–0.657)*

0.641 (0.635–0.648)

Osteoarthritis (203)

1.81 (1.26–2.41)

0.620 (0.611–0.630)*

0.597 (0.587–0.607)

Complication of device; implant or graft (237)

1.73 (1.22–2.15)

0.641 (0.634–0.649)*

0.625 (0.618–0.632)

Fracture of neck of femur (hip) (226)

2.33 (1.74–2.83)

0.617 (0.606–0.628)*

0.575 (0.564–0.586)

Complications of surgical procedures or medical care (238)

0.70 (0.45–1.15)

0.576 (0.568–0.584)*

0.556 (0.548–0.564)

Appendicitis and other appendiceal conditions (142)

1.45 (0.67–1.79)

0.583 (0.571–0.596)*

0.520 (0.506–0.533)

Calculus of urinary tract (160)

2.31 (1.57–2.64)

0.615 (0.604–0.625)*

0.574 (0.563–0.585)

Abdominal hernia (143)

1.36 (0.63–1.76)

0.648 (0.636–0.661)*

0.628 (0.615–0.641)

Cardiac dysrhythmias (106)

1.11 (0.65–1.18)

0.590 (0.585–0.595)*

0.568 (0.562–0.573)

Hyperplasia of prostate (164)

2.70 (1.37–3.86)

0.636 (0.620–0.651)*

0.592 (0.576–0.608)

  1. 1CCS = Clinical Classifications Software [55]
  2. *significantly (p < 0.05) higher compared to the C-statistics of the models without a random effect on the hospital level