Table 1. Mixed-Effect Models of Adjusted HCAHPS-SS Using Independent Time Variablesa.
| 2020 | Q1/2020 (n = 5 861 470 patients; n = 1648 hospitals) | Q2/2020 (n = 5 807 876 patients; n = 1737 hospitals) | Q3/2020 (n = 6 118 270 patients; n = 3224 hospitals) | Q4/2020 (n = 6 104 234 patients; n = 3295 hospitals) | ||||
|---|---|---|---|---|---|---|---|---|
| Est (SE), % | P value | Est (SE), % | P value | Est (SE), % | P value | Est (SE), % | P value | |
| Post-2019 indicator | −0.04 (0.05) | .52 | −1.20 (0.06) | <.001 | −2.00 (0.05) | <.001 | −1.91 (0.05) | <.001 |
| Time, linear mo | 0.08 (0.01) | <.001 | 0.08 (0.01) | <.001 | 0.08 (0.01) | <.001 | 0.08 (0.01) | <.001 |
| Q2 indicator | 0.42 (0.03) | <.001 | 0.42 (0.03) | <.001 | 0.42 (0.03) | <.001 | 0.42 (0.03) | <.001 |
| Q3 indicator | 0.35 (0.03) | <.001 | 0.35 (0.03) | <.001 | 0.35 (0.03) | <.001 | 0.35 (0.03) | <.001 |
| Q4 indicator | 0.04 (0.03) | .19 | 0.04 (0.03) | .21 | 0.04 (0.03) | .26 | 0.04 (0.03) | .22 |
| 2021 | Q1/2021 (n = 6 099 727 patients; n = 3219 hospitals) | Q2/2021 (n = 6 109 713 patients; n = 3181 hospitals) | Q3/2021 (n = 6 084 396 patients; n = 3232 hospitals) | Q4/2021 (n = 5 990 289 patients; n = 3137 hospitals) | ||||
| Post-2019 indicator | −1.93 (0.06) | <.001 | −2.34 (0.06) | <.001 | −3.36 (0.06) | <.001 | −3.57 (0.06) | <.001 |
| Time, linear mo | 0.08 (0.01) | <.001 | 0.08 (0.01) | <.001 | 0.08 (0.01) | <.001 | 0.08 (0.01) | <.001 |
| Q2 indicator | 0.42 (0.03) | <.001 | 0.42 (0.03) | <.001 | 0.42 (0.03) | <.001 | 0.42 (0.03) | <.001 |
| Q3 indicator | 0.35 (0.03) | <.001 | 0.35 (0.03) | <.001 | 0.35 (0.03) | <.001 | 0.35 (0.03) | <.001 |
| Q4 indicator | 0.04 (0.03) | .27 | 0.04 (0.03) | .25 | 0.04 (0.03) | .24 | 0.04 (0.03) | .23 |
Abbreviations: est, estimation; HCAHPS-SS, Hospital Consumer Assessment of Healthcare Providers and Systems summary score; Q, quarter.
a
Each model includes Q1/2018 to Q4/2019 discharges, plus 1 post-2019 quarter (each model had 9 quarters). Each model was limited to hospitals with at least 25 completed surveys in their post-2019 quarter and included data from all quarters for which these hospitals had at least 25 completed surveys. Models also included hospital random effects.