Table 3. Controlled Interrupted Time-Series Analysis Estimating Differences in Short-Stay Quality Outcomes for SNFs Undergoing a Change in Ownership vs Control SNFs That Did Not Undergo Ownership Changes .
| Quality outcome | Coefficient (95% CI)a | P value |
|---|---|---|
| Percentage community discharge | ||
| Control group longitudinal trend | −1.61 (−1.75 to −1.47) | <.001 |
| Pre–change in ownership difference in longitudinal trend between treatment and control SNFs | 0.17 (−0.28 to 0.63) | .46 |
| Post–change in ownership difference in outcome between treatment and control SNFs | ||
| Short-term | −0.10 (−1.36 to 1.15) | .87 |
| Long-term difference in longitudinal trends | −0.28 (−2.86 to 2.30) | .83 |
| Nonprofit to for-profit conversion only | ||
| Short-term | −0.07 (−5.19 to 5.06) | .98 |
| Long-term | −3.85 (−14.33 to 6.65) | .47 |
| Percentage emergency department visits | ||
| Control group longitudinal trend | −0.61 (−0.64 to −0.58) | <.001 |
| Pre–change in ownership difference in longitudinal trend between treatment and control SNFs | 0.23 (0.12 to 0.34) | <.001 |
| Post–change in ownership difference in outcome between treatment and control SNFs | ||
| Short-term | 0.32 (0.03 to 0.62) | .03 |
| Long-term difference in longitudinal trends | −0.05 (−0.13 to 0.43) | .83 |
| Nonprofit to for-profit conversion only | ||
| Short-term | 0.26 (−0.81 to 1.34) | .63 |
| Long-term | −0.04 (−1.88 to 1.80) | .96 |
| Percentage hospital readmissions | ||
| Control group longitudinal trend | 0.03 (0.003 to 0.07) | .04 |
| Pre–change in ownership difference in longitudinal trend between treatment and control SNFs | 0.51 (0.37 to 0.64) | <.001 |
| Post–change in ownership difference in outcome between treatment and control SNFs | ||
| Short-term | 0.27 (−0.08 to 0.61) | .09 |
| Long-term difference in longitudinal trends | 0.13 (−0.46 to 0.71) | .67 |
| Nonprofit to for-profit conversion only | ||
| Short-term | −0.01 (−1.25 to 1.23) | .99 |
| Long-term | −0.19 (−2.40 to 2.02) | .87 |
Abbreviation: SNF, skilled nursing facility.
a
Models also adjusted for state fixed effects, ownership, staffing ratings, percentage Medicaid patients, urban vs rural county, chain-affiliated vs independent status, hospital-based vs freestanding location, and occupancy rate.