Introduction
Home health services play a central role in post-acute care transitions for Medicare beneficiaries. Variability in visit allocation across disciplines has increased under the Patient-Driven Groupings Model (PDGM), which decoupled payment from therapy volume. While concerns have been raised that reductions in therapy intensity may affect patient outcomes, empirical evidence examining discipline-specific early visit intensity and subsequent acute utilization remains limited at the national level.
This study evaluates the association between early therapy and nursing visit intensity and 30-day acute utilization following home health admission in a large Medicare fee-for-service cohort.
Key Findings
1. Acute Utilization Is Common
13.4% of admissions resulted in rehospitalization within 30 days. 20.7% resulted in any acute utilization (rehospitalization, ED, or observation). Approximately 1 in 5 home health admissions is followed by acute utilization within 30 days.
2. Early Therapy Intensity Is Associated with Lower Acute Utilization
After adjusting for patient risk, PDGM case mix, and clustering at the agency level: higher early therapy intensity was associated with 13% lower odds of 30-day composite acute utilization. For rehospitalization alone, early therapy intensity was associated with 7% lower odds. This association was slightly stronger among patients with high functional impairment.
Key Finding
Front-loading therapy in week 1 appears to be modestly associated with improved short-term stability.
3. Early Nursing Intensity Appears to Reflect Medical Instability
Higher early nursing intensity was associated with a modest increase in acute utilization risk. This likely reflects underlying medical complexity rather than nursing causing higher utilization. Nursing intensity functions primarily as a marker of patient severity.
Operational Implications for HHAs
1. Early therapy allocation is not neutral. Indiscriminate reduction of early therapy may increase downstream instability risk.
2. Functional stratification matters. Therapy appears most impactful in functionally impaired patients.
3. Week 1 scheduling discipline is strategically important. Early allocation patterns show measurable associations with 30-day outcomes.
4. Nursing allocation should be interpreted as a severity indicator, not a preventive lever alone.
Warning
Important Caveats: This is an observational study reporting associations, not causation. Effect sizes are modest at the individual level. Results are population-level and risk-adjusted.
Methods
Data Source and Study Population
We used Medicare fee-for-service claims data accessed via the CMS Virtual Research Data Center (VRDC). The analytic cohort included all home health agency admissions from January 1, 2022 through December 31, 2025. Each observation represented a distinct home health start of care.
Exposure Variables
Early therapy intensity was defined as the number of physical therapy, occupational therapy, and speech-language pathology visits occurring within 7 days of home health admission, divided by 7. Early nursing intensity was defined analogously. Both variables were standardized (mean 0, SD 1) to facilitate interpretation.
Outcomes
Primary outcomes included 30-day rehospitalization, 30-day emergency department visit, and composite acute utilization (rehospitalization, ED visit, or observation stay). Outcomes were measured within 30 days of home health admission.
Covariates
Models adjusted for age, CMS HCC risk score (v28 Model), dual eligibility (binary), ESRD status, PDGM case-mix weight, and functional impairment tier (derived from PDGM code).
Statistical Analysis
Multivariable logistic regression models were estimated. Standard errors were clustered at the home health agency level using PROC SURVEYLOGISTIC to account for within-agency correlation. Interaction terms between early therapy intensity and functional impairment were tested to evaluate heterogeneity.
Results
Cohort Characteristics
Among 13,065,775 admissions from 2022–2025: rehospitalization rate was 13.4%, ED visit rate was 8.0%, observation stays were 2.1%, and composite acute utilization was 20.7%. Rehospitalization increased across functional impairment tiers (Low: 10.2%, Medium: 12.9%, High: 16.6%).
Mean early therapy intensity was 0.256 visits per day (approximately 1.8 visits during week 1). Mean early nursing intensity was 0.241 visits per day (approximately 1.7 visits during week 1). Approximately 25.6% of admissions had no therapy visits during the first 7 days.
Table 1. Cohort Acute Utilization Summary
| Outcome | Rate | N |
|---|---|---|
| Rehospitalization | 13.4% | 1,750,814 |
| ED Visit | 8.0% | 1,045,262 |
| Observation Stay | 2.1% | 274,381 |
| Composite Acute Utilization | 20.7% | 2,704,616 |
Multivariable Models
After clustering at the agency level and adjusting for severity and case mix:
Table 2. Adjusted Odds Ratios per 1-SD Increase in Early Visit Intensity
| Variable | Composite Acute Utilization OR (95% CI) | Rehospitalization OR (95% CI) |
|---|---|---|
| Early Therapy Intensity | 0.870 (0.864–0.876) | 0.928 (0.926–0.930) |
| Early Nursing Intensity | 1.054 (1.048–1.059) | 1.038 (1.036–1.039) |
Interaction models indicated slightly stronger protective association of early therapy intensity among patients with high functional impairment.
Discussion
In this national Medicare cohort, higher early therapy intensity was consistently associated with modest reductions in short-term acute utilization following home health admission. The association persisted after adjustment for baseline risk, PDGM case-mix weight, and agency-level clustering.
In contrast, early nursing intensity demonstrated a small positive association with acute utilization, likely reflecting residual confounding by underlying medical instability.
These findings suggest that early therapy allocation may contribute to transitional stability, particularly among functionally impaired patients. While effect sizes were modest at the individual level, small relative risk differences may translate into meaningful population-level impacts in value-based payment environments.
Limitations
This study is observational and subject to residual confounding. Functional impairment was derived from PDGM classification rather than raw OASIS scores. Causal inference cannot be established.
Conclusion
Higher early therapy intensity following home health admission is independently associated with lower 30-day acute utilization in Medicare beneficiaries. Findings support the importance of early functional stabilization strategies within home health practice.