Table 3.
SHR and AHR Rates and ERR Changes by Hospital Characteristics
| Characteristic | Mean Readmission Rate, % | No. of Hospitals Whose ERR Changed From SHR to AHRb |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| Unadjusted | Adjusted | ||||||||
| SHR | AHR | Underestimation, %a | SHR | AHR | Underestimation, %a | ERR >1 to ≤1c |
ERR ≤1 to >1d |
Total | |
| Overall | 2.38 | 3.10 | 23.18 | 2.72 | 3.59 | 24.22 | 10 | 10 | 20 |
| Children’s hospital status | |||||||||
| Children’s | 5.30 | 5.84 | 9.16 | 4.36 | 4.83 | 9.68 | 0 | 0 | 0 |
| Nonchildren’s | 2.17 | 2.90 | 25.23 | 2.60 | 3.50 | 25.67 | 10 | 10 | 20 |
| Teaching status | |||||||||
| Teaching | 3.67 | 4.23 | 13.22 | 3.56 | 4.15 | 14.20 | 2 | 0 | 2 |
| Nonteaching | 2.04 | 2.80 | 27.16 | 2.50 | 3.44 | 27.41 | 8 | 10 | 18 |
| Annual pediatric volume, quartilee | |||||||||
| 1 (1 to <385) | 2.02 | 2.84 | 28.92 | 2.24 | 3.22 | 30.49 | 2 | 2 | 4 |
| 2 (385 to <1179) | 1.93 | 2.57 | 24.71 | 2.52 | 3.37 | 25.06 | 3 | 3 | 6 |
| 3 (1179 to <3947) | 1.86 | 2.64 | 29.50 | 2.39 | 3.37 | 28.87 | 3 | 4 | 7 |
| 4 (3947 to 44 685) | 3.67 | 4.31 | 14.78 | 3.69 | 4.38 | 15.56 | 2 | 1 | 3 |
| Hospital location | |||||||||
| Urban | 2.20 | 2.88 | 23.57 | 2.74 | 3.62 | 24.50 | 5 | 6 | 11 |
| Large rural city or town | 2.20 | 2.76 | 20.27 | 2.79 | 3.53 | 20.96 | 4 | 2 | 6 |
| Small and isolated small rural town |
3.25 | 4.29 | 24.10 | 2.59 | 3.50 | 26.04 | 1 | 2 | 3 |
Abbreviations: AHR, all-hospital readmission; ERR, excess readmission ratio; SHR, same-hospital readmission.
Calculated as the relative percentage that SHRs underestimate AHRs using the formula |(AHR − SHR)/AHR|.
Represent predicted divided by expected readmissions. Hospitals with an ERR greater than 1 incur penalties. Predicted readmissions are calculated by estimating the probability of readmission for each patient at a hospital and adding the probabilities for all of the hospital’s patients. Expected readmissions are calculated by estimating the probability of readmission for each of a hospital’s patients if admitted to an average hospital and then adding the probabilities for all of the hospital’s patients.
Indicates hospitals whose anticipated performance improved (and who would therefore no longer be subject to penalties under a program with the penalty criteria used in Medicare’s Hospital Readmissions Reduction Program) if AHR rates were used in ERR calculations.
Indicates hospitals whose anticipated performance worsened (and who would become subject to penalties) if AHR rates were used in ERR calculations.
Given as number of pediatric admissions.