High-Intensity Home-Based Rehabilitation in a Medicare Accountable Care Organization

ABSTRACT

Objectives: Patients are often discharged to a skilled nursing facility (SNF) for postacute rehabilitation. Functional outcomes achieved in SNFs are variable, and costs are high. Especially for accountable care organizations (ACOs), home-based postacute rehabilitation offers a high-value option if outcomes are not compromised. The objective was to compare outcomes for episodes in a novel high-intensity home-based rehabilitation (HIHR) model vs an SNF.

Study Design: Retrospective cohort study.

Methods: Medicare patients from a large integrated multihospital health system who had low to moderate medical complexity and mild to moderate mobility deficits at hospital discharge were included. The primary exposure was discharge to HIHR (intervention) or an SNF (control) after hospitalization. The primary outcome was Activity Measure for Post-Acute Care (AM-PAC) mobility score. Secondary outcomes were Medicare costs within 30 and 90 days post hospitalization, 30-day readmission rate, and index hospital length of stay (LOS). Inverse probability of treatment-weighted regression was used for comparison between cohorts.

Results: There were 171 patients discharged to HIHR and 841 to SNFs. The adjusted AM-PAC mobility T-score was 8.2 (95% CI, 6.3-10.1) points higher after HIHR vs SNF. Adjusted Medicare costs were lower for the HIHR cohort (within 90 days, –$17,123; 95% CI, –$20,757 to –$13,490). Hospital LOS and odds for readmission did not differ between cohorts.

Conclusions: The HIHR cohort demonstrated better functional outcomes and lower posthospital costs. HIHR may be a high-value option for patients attributed to a Medicare ACO who have moderate medical complexity and moderate functional deficits at the time of hospital discharge.

Am J Manag Care. 2025;31(1):In Press

Takeaway Points

• For patients in a Medicare accountable care organization who had low to moderate medical complexity and mild to moderate mobility deficits at hospital discharge, postacute care in a high-intensity home-based rehabilitation (HIHR) model may be a feasible alternative to a skilled nursing facility.
• Patients in HIHR had better functional scores than those in skilled nursing facilities at the end of the care episode after adjusting for individual clinical characteristics.
• After adjustment, all-cause Medicare reimbursement costs were also lower for the HIHR cohort (–$17,123) within 90 days of index hospital discharge.
• Scaling HIHR models will remain challenging until postacute care reimbursement policies support it.

Most patients prefer to be discharged from the hospital to their home,^1-4 but many require rehabilitation in a postacute care (PAC) facility.^5-8 Twenty percent of hospitalized Medicare patients are discharged to a skilled nursing facility (SNF).⁹ The extent to which these patients improve their function and return to the community following PAC is highly variable.^10-12 Additionally, costs of SNFs are high, generally 3 to 4 times those of home health care (HHC).¹⁰ If patients can achieve the same or better outcomes at home as in an SNF, we reasoned that home-based rehabilitation should offer better value (outcomes/costs).¹³

Medicare’s current models tie PAC reimbursement to the physical location of care rather than the intensity of care (ie, HHC is reimbursed less than SNF care even for similar scope and service intensity).^14,15 Conversely, in accountable care organizations (ACOs) and similar value-based arrangements, the financial incentive is to reduce the total costs of care while improving or maintaining quality. In such models, if outcomes are not compromised, a less expensive home-based PAC is an attractive option.

Our health system developed and implemented a high-intensity home-based rehabilitation (HIHR) model, “Home Care Plus,” in 2019 for a subgroup of patients attributed to our Medicare ACO.⁴ The aim of this study was to examine mobility status at PAC discharge for patients in HIHR vs those in an SNF. We also compared hospital length of stay (LOS), 30-day readmission, and 30- and 90-day posthospital health care costs between HIHR and SNF cohorts.

METHODS

The HIHR model has been described elsewhere.⁴ Briefly, it differs from traditional HHC in 3 ways. First, its primary emphasis is on rehabilitation via physical, occupational, and speech therapy, with nursing services provided as needed. Second, rehabilitation service intensity is higher, with visits occurring near daily during the first 10 to 14 days instead of intermittently distributed across 30- or 60-day episodes. Lastly, private-duty home health aides visit daily to assist with cooking, cleaning, bathing, and so on. Although this service would typically require out-of-pocket payment by the patient, it is covered in HIHR by the ACO. The provision of near-daily therapy and wraparound support in HIHR is modeled after SNFs, which differ in that they have a rehabilitation gym and nursing staff on site. In HIHR, however, patients navigate the real-world functional challenges of their homes.

HIHR also differs from hospital-at-home models, which provide acute inpatient-level medical care in the home.^16,17 HIHR provides PAC and is reimbursed by Medicare as a standard HHC episode without adjustment. In our study, HIHR services were provided by an HHC agency integrated within our health system.

Study Design and Data Sources

This was a retrospective cohort study. In May 2019, HIHR became a PAC option for patients discharged from 4 hospitals in the Cleveland Clinic health system (based in Cleveland, Ohio). A complete registry of patients discharged to HIHR was maintained prospectively; variables pertaining to their characteristics, care, and outcomes were identified retrospectively. We identified the SNF cohort retrospectively.

Data for each hospital, HIHR, and SNF episode were extracted from medical records and linked across settings. Health care utilization and associated costs in the 90 days following hospitalization were extracted from all-setting Medicare claims for services provided between May 2019 and August 2022.

Study Sample

Eligibility criteria included (1) attribution to our ACO (insured by traditional Medicare with a primary care physician in our system); (2) 30-day readmission risk less than 35% (as a proxy for medical complexity), estimated using an internally developed and validated risk tool¹⁸; and (3) an Activity Measure for Post-Acute Care (AM-PAC) 6-Clicks basic mobility short form raw score between 16 and 22. The AM-PAC 6-Clicks short form is a valid, reliable measure of function for inpatients.^19,20 Scores in this range suggest that tasks such as getting out of bed and walking a few feet require some assistance—making rehabilitation indicated—but deficits are not significant enough to impede being discharged home. As part of routine discharge planning, clinical teams considered appropriateness of HIHR. If appropriate, they communicated this option to patients and their at-home caregivers who would make the decision. We included HIHR episodes between May 1, 2019, and April 30, 2022. For patients with multiple HIHR episodes, we included only the first.

We included patients in the SNF cohort if they were discharged in the same time frame from any of 11 hospitals in our system to any of 52 SNFs within our preferred network. To ensure similarity to HIHR patients, we excluded patients who did not meet HIHR eligibility criteria, and we included only the first episode.

For patients who had both an SNF and an HIHR episode within the study period, we excluded the subsequent episode if it occurred within 90 days of the first. Finally, we excluded patients missing both an AM-PAC mobility score (ie, it was not collected during the care episode) and 90-day cost data (ie, not insured by Medicare, so no claims data were available). For the primary analysis, we excluded patients missing an AM-PAC mobility score (the primary outcome) at PAC discharge. For the cost analysis, we excluded patients missing all cost data.

Variables

We described HIHR services by episode length and the number of visits provided by a nurse, occupational therapist (OT), physical therapist (PT), speech-language pathologist (SLP), and private-duty home health aide. In all analyses, cohort designation (HIHR vs SNF) was the primary exposure variable. The primary outcome was mobility status at the end of the PAC episode, measured using AM-PAC basic mobility short forms. In SNFs, the AM-PAC 6-Clicks short form was scored by a PT (as it was validated),¹⁹ and in HIHR, the AM-PAC Home Care short form was scored by patient self-report.^21,22 The 6-Clicks short form assesses tasks ranging from rolling in bed to navigating 3 to 5 steps. The Home Care short form assesses tasks ranging from sitting up in bed to climbing a full flight of stairs. For both short forms, item-level scores range from 1 (dependent) to 4 (independent). Total raw scores from both short forms were converted to AM-PAC T-scores for comparison.^23,24 The T-score range for the 6-Clicks short form is 16.6 to 57.7, and for the Home Care short form it is 30.4 to 70.7.

Additional outcomes were 30-day hospital readmission (as recorded in claims data or the SNF record) and index hospital LOS. Examination of hospital LOS is warranted because discharge to an SNF can require time-consuming processes (eg, finding an SNF with bed capacity and obtaining insurance preauthorization) that may not present barriers to being discharged to HIHR.

Costs included payments for equipment or services provided within 30 and 90 days of hospital discharge, excluding Part D drug costs. For the HIHR cohort, we added costs incurred by the ACO for the private-duty aide services. For 17 patients in the HIHR cohort who were missing costs for the HHC episode, we used the sample median ($1504). For 12 of these 17 episodes that lasted less than 30 days, we multiplied the sample median by 60% to estimate the cost; this matches reimbursement practice in the Patient-Driven Groupings Model.²⁵ For 101 patients in the SNF cohort missing SNF costs, we estimated the cost by multiplying the mean cost per SNF day ($256) by the LOS.

Statistical Analyses

To estimate the effect of HIHR compared with SNF on functional status at PAC discharge and 30-day readmission, we used inverse probability of treatment-weighted regression modeling with a logit treatment model (HIHR vs SNF).²⁶ We used a continuous outcome model for functional status at PAC discharge and a logit outcome model for readmission. In the models for functional status, both the treatment and outcome models included age, sex, assistance available at home, prehospital functional status, hospital LOS, surgery during hospitalization, last recorded readmission risk, and last recorded AM-PAC mobility score. The treatment and outcome models for 30-day readmission included these same covariates plus the attending physician’s service line. Following model estimation, we summarized raw and weighted standardized differences of the covariates to check for balance between cohorts, where weighted differences closer to zero indicated improved between-group balance.

Because some patients in the HIHR cohort were missing final AM-PAC scores, we conducted a prespecified sensitivity analysis in which we derived missing AM-PAC scores from the Continuity Assessment Record and Evaluation section GG (CARE-GG) mobility score.²⁷ There is no published standard for matching AM-PAC and CARE-GG scores, so we used clinical rationale to match 6 of the 15 items included on CARE-GG to the 7 items on the AM-PAC Home Care short form. Details of the derivation are described in eAppendix 1 (eAppendices available at ajmc.com). As a check, we examined the correlation between recorded and derived scores among patients who had both. After observing a potential ceiling effect in the SNF cohort, we conducted an additional post hoc sensitivity analysis comparing AM-PAC scores between cohorts using median regression.^28,29

Due to the skewed distribution of hospital LOS and costs, we used generalized linear models with a log link and gamma family to compare the HIHR and SNF cohorts on these outcomes. The LOS model included age, sex, Charlson Comorbidity Index score, surgery during hospitalization, attending physician’s service line, and first and last recorded AM-PAC mobility scores. The cost models used the same covariates, excluding first recorded AM-PAC mobility scores and adding hospital LOS and last recorded readmission risk. From model coefficients, we used marginal estimation to predict adjusted mean LOS and costs for each cohort and the adjusted differences between them.

The Cleveland Clinic Institutional Review Board approved this study. The ideas and opinions expressed in this publication are not attributable to CMS or the Medicare Shared Savings Program.

RESULTS

Between May 2019 and April 2022, 181 patients were discharged to HIHR, and 35,990 were discharged to an SNF. After exclusions, 171 patients in the HIHR cohort and 841 in the SNF cohort remained in the study sample. The Figure details the sample size for each of the analyses. Characteristics of patients in the primary analysis are shown in Table 1. Most patients in both cohorts were categorized into general (eg, endocrine, gastrointestinal, respiratory) surgical and medical services. Other patients were distributed among cardiothoracic, vascular, neurological, oncology, and orthopedic services; the sample size is suppressed in the tables per Medicare reporting regulations. Characteristics of all patients in the study sample, sensitivity analysis, and cost analyses appear in eAppendix 2. Compared with no weighting, propensity weighting improved covariate balance between groups.

The median (IQR) HIHR episode length was 32.0 (22.0-48.0) days vs 16.0 (11.0-24.0) days in SNFs. The median (IQR) number of visits from nursing personnel (registered nurse or licensed practical nurse) was 7.0 (4.0-12.0). The median (IQR) numbers of OT, PT, and SLP visits were 5 (2-7), 11 (8-15), and 0 (0-0), respectively. Ninety (52.6%) of the patients received private-duty aide services from HIHR (median [IQR], 11 [5-13] visits). Among the 81 who did not receive aide services, 57 (70.4%) declined them.

As shown in Table 2, the unadjusted final AM-PAC mobility score—both recorded and derived—was higher in the HIHR vs SNF cohort. Among those with recorded scores, the mean (SD) for the SNF cohort was 48.9 (8.1) compared with 55.6 (7.3) for the HIHR cohort. Using derived scores, the mean (SD) for the HIHR cohort was 54.7 (6.8). There was a moderate correlation (r = 0.54; P < .05; n = 45) between recorded and derived scores. Compared with those who were discharged to SNFs, patients who were discharged to HIHR had a shorter hospital LOS and lower 30- and 90-day Medicare costs, and a similar proportion were readmitted.

In the primary analysis (Table 3), the adjusted final AM-PAC mobility score at HIHR discharge compared with SNF discharge was 7.8 (95% CI, 5.9-9.8) points higher. In the sensitivity analysis using derived scores, the difference was 6.1 (95% CI, 4.9-7.4) points. In the sensitivity analysis to account for a potential ceiling effect in the SNF cohort, where 32.6% of patients had the maximum score, the adjusted difference in medians was 6.0 (95% CI, 3.9-8.0) points. The adjusted hospital LOS was shorter for the HIHR cohort. The odds of readmission did not differ significantly between cohorts. Adjusted Medicare costs were significantly lower for the HIHR cohort (including the ACO-covered cost of the aide) than the SNF cohort (within 90 days, –$17,123; 95% CI, –$20,757 to –$13,490).

DISCUSSION

In this retrospective study, participation in a novel HIHR model for PAC was associated with better mobility at the conclusion of the PAC episode compared with PAC for similar patients in SNFs. HIHR was also associated with shorter index hospital LOS. There was no significant difference in readmission rates between groups. Ninety-day posthospital Medicare costs for the HIHR cohort were nearly one-third those of the SNF cohort, even after accounting for private-duty aide services. Those aide services typically require out-of-pocket payments. In HIHR, however, our ACO covered this cost to ensure that patients—and their at-home caregivers—had sufficient support to go home. Because outcomes were positive, HIHR appears to offer higher value than SNFs after hospitalization for these patients.

Our findings resemble those of Augustine et al,³ who found that a majority of patients in a rehabilitation-at-home program achieved meaningful functional benchmarks within 30 days. Their study delivered a shorter service (mean, 14.2 days) with more frequent PT visits (5.2 visits per week) than ours did. Our therapy frequency was more than double the average number of visits received in standard HHC,³⁰ but it was lower than intended because HIHR services were often downgraded to standard HHC soon after hospital discharge. The HHC agency had a financial disincentive to provide more intensive services because HIHR was reimbursed by Medicare as standard HHC.

Nevertheless, functional outcomes in HIHR exceeded those in SNFs. Several studies comparing standard HHC with SNF care have had similar findings.^10-12,31,32 Werner et al³² found no significant differences in function between those in HHC and those in SNFs but lower costs in HHC. Whether functional outcomes are better (which we found with HIHR) or the same (which others have found with HHC generally), if costs are lower, then value is improved.

The HIHR model requires ongoing development. There are 2 challenges facing the present model. First, our study identifies only potential characteristics of patients who can benefit from HIHR. These include low to moderate medical complexity and mild to moderate mobility impairments.^18,33-37 To scale this model to other systems or patient populations, these eligibility criteria will need to be refined.

Second, reimbursement policies must support the model. Our observation that hospital LOS is shorter for patients being discharged to HIHR is compelling for hospitals because the Inpatient Prospective Payment System incentivizes the use of fewer resources and greater bed availability. Meanwhile, home care delivery and reimbursement are complex.³⁸ Alternative payment models (eg, ACOs or bundled payments) can allow for higher-value PAC services—such as HIHR—by aligning incentives between hospital systems and patients.^39-42 For an integrated system like ours that owns its hospitals and HHC agency, there is financial incentive to create new programs capable of delivering high-value, patient-centered services. However, the financial incentives for growing these programs to broader preferred networks are often lacking. Even in an ACO, there is no standard practice for attributing and then directing shared savings—if they are achieved—to partnered PAC entities. Further, performance-based payments are unreliable revenue sources disbursed months after the performance year has ended, if at all. For potentially cost-saving models such as HIHR to be financially sustainable, reimbursement needs to be granular, timely, and aligned across settings.⁴³ From an operational perspective, scaling programs such as HIHR to the majority of patients—beyond those in value-based arrangements—will remain challenging without such specific incentives.

Limitations

Mobility measures were collected as part of standard clinical care. Although pragmatic, this introduces limitations. We used validated standard-of-care AM-PAC measures (clinician-rated 6-Clicks and patient-rated Home Health short form). AM-PAC T-scores are comparable between short forms, but only 1 study has compared clinician-rated scores with patient-rated scores on the 6-Clicks short form; agreement was moderate.⁴⁴ Additional studies are needed to further validate agreement between clinician-reported and patient-reported functional outcomes. Because HIHR episodes were longer than SNF episodes, improved mobility in the HIHR group could be the result of a longer recovery time. Many patients had missing AM-PAC scores at PAC discharge. Although we conducted a sensitivity analysis with derived scores for patients who had a CARE-GG assessment completed, the validity of that derivation has not been established. However, a moderate correlation with recorded scores suggests that derived scores were a reasonable estimation. The ceiling effect in the SNF cohort cannot be ruled out as an explanation for observed effects but is less concerning given the persistence of the effect in our quantile regression analysis. Lastly, because the AM-PAC is scored subjectively, scores could be manipulated, but such a practice would constitute fraud.

Introducing HIHR as a standard discharge option was also pragmatic but problematic in this study. We did not capture whether HIHR was offered to patients, so we do not know whether or how clinical processes and patient preferences influenced HIHR use. We are conducting a separate implementation study to examine this. Also, our study may be subject to confounding by indication (ie, patients who are functionally worse and/or have greater medical complexity are more likely to both be discharged to an SNF and have poorer functional outcomes in the SNF). It was for this reason that we used an inverse probability of treatment-weighted regression approach with adjustment for clinically important variables. Although this analysis was robust and adjustment was thorough, it was not comprehensive. For example, we were unable to control for discharge practice patterns that may differ between hospitals. Finally, our sample was small and from 1 health system, limiting generalizability.

CONCLUSIONS

This study demonstrates that for patients with no greater than moderate medical complexity and functional deficits, an HIHR model may effectively substitute for PAC in an SNF. Rehabilitation service intensity was lower than expected, but higher than usual care. HIHR patients had shorter hospital stays, had better function at PAC discharge, and were not readmitted more often than similar patients who were discharged to SNFs. From Medicare’s perspective, these outcomes were achieved at 33% to 48% of the cost of an SNF, including the cost of private-duty aide services. If future research validates these findings in larger prospective studies, scaling HIHR and ensuring PAC reimbursement provides appropriate incentives may be warranted.

Author Affiliations: Department of Physical Medicine and Rehabilitation (JKJ) and Rehabilitation and Sports Therapy (JKJ, KJG, MS), Neurological Institute, Cleveland Clinic, Cleveland, OH; Center for Value-Based Care Research, Cleveland Clinic (JKJ, MBR, JAH), Cleveland, OH; Department of Medicine, Cleveland Clinic Lerner College of Medicine at Case Western Reserve University (JKJ, MBR, JED, JAH), Cleveland, OH; now with Department of Orthopedic Surgery, Duke University School of Medicine (JKJ), Durham, NC; Department of Internal Medicine and Geriatrics, Community Care, Cleveland Clinic (MBR, JAH), Cleveland, OH; Center for Population Health Research, Cleveland Clinic (JED), Cleveland, OH; Department of Quantitative Health Sciences, Lerner Research Institute, Cleveland Clinic (JED), Cleveland, OH; Center for Connected Care, Cleveland Clinic (WZ, DC, SP, LO, JM), Cleveland, OH.

Source of Funding: Agency for Healthcare Research and Quality award No. K01HS028529.

Author Disclosures: Dr Rothberg reports serving as a consultant for Blue Cross Blue Shield Association. The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (JKJ, MBR, JED, WZ, DC, SP, LO, JM, KJG, JAH); acquisition of data (JKJ, SP, LO, JM, JAH); analysis and interpretation of data (JKJ, MBR, JED, WZ, DC, MS, JAH); drafting of the manuscript (JKJ, MBR, KJG, JAH); critical revision of the manuscript for important intellectual content (JKJ, MBR, JED, WZ, DC, SP, LO, JM, KJG, MS, JAH); statistical analysis (JKJ, JED); provision of patients or study materials (JKJ, WZ, DC, KJG, JAH); obtaining funding (JKJ); administrative, technical, or logistic support (DC, SP, JM); and supervision (MBR, WZ, LO, MS, JAH).

Address Correspondence to: Joshua K. Johnson, DPT, PhD, Department of Orthopedic Surgery, Duke University School of Medicine, 311 Trent Dr, Durham, NC 27710. Email: joshua.johnson@duke.edu.

REFERENCES

1. Ackerly DC, Grabowski DC. Post-acute care reform—beyond the ACA. N Engl J Med. 2014;370(8):689-691. doi:10.1056/NEJMp1315350

2. Robinson SM, Ní Bhuachalla B, Ní Mhaille B, Cotter PE, O’Connor M, O’Keeffe ST. Home, please: a conjoint analysis of patient preferences after a bad hip fracture. Geriatr Gerontol Int. 2015;15(10):1165-1170. doi:10.1111/ggi.12415

3. Augustine MR, Davenport C, Ornstein KA, et al. Implementation of post-acute rehabilitation at home: a skilled nursing facility-substitutive model. J Am Geriatr Soc. 2020;68(7):1584-1593. doi:10.1111/jgs.16474

4. Johnson JK, Hohman JA, Vakharia N, et al. High-intensity postacute care at home. NEJM Catalyst. 2021;2(6). doi:10.1056/CAT.21.0125

5. Burke RE, Juarez-Colunga E, Levy C, Prochazka AV, Coleman EA, Ginde AA. Rise of post-acute care facilities as a discharge destination of US hospitalizations. JAMA Intern Med. 2015;175(2):295-296. doi:10.1001/jamainternmed.2014.6383

6. Burke RE, Juarez-Colunga E, Levy C, Prochazka AV, Coleman EA, Ginde AA. Patient and hospitalization characteristics associated with increased post-acute care facility discharges from US hospitals. Med Care. 2015;53(6):492-500. doi:10.1097/MLR.0000000000000359

7. Burke RE, Lawrence E, Ladebue A, et al. How hospital clinicians select patients for skilled nursing facilities. J Am Geriatr Soc. 2017;65(11):2466-2472. doi:10.1111/jgs.14954

8. Falvey JR, Burke RE, Ridgeway KJ, Malone DJ, Forster JE, Stevens-Lapsley JE. Involvement of acute care physical therapists in care transitions for older adults following acute hospitalization. J Geriatr Phys Ther. 2019;42(3):E73-E80. doi:10.1519/JPT.0000000000000187

9. Tian W. An All-Payer View of Hospital Discharge to Postacute Care, 2013. Healthcare Cost and Utilization Project statistical brief 205. Agency for Healthcare Research and Quality; 2016. Accessed June 7, 2019.
https://www.ncbi.nlm.nih.gov/pubmed/27441335

10. Buntin MB, Colla CH, Deb P, Sood N, Escarce JJ. Medicare spending and outcomes after postacute care for stroke and hip fracture. Med Care. 2010;48(9):776-784. doi:10.1097/MLR.0b013e3181e359df

11. Mallinson T, Deutsch A, Bateman J, et al. Comparison of discharge functional status after rehabilitation in skilled nursing, home health, and medical rehabilitation settings for patients after hip fracture repair. Arch Phys Med Rehabil. 2014;95(2):209-217. doi:10.1016/J.APMR.2013.05.031

12. Hong I, Goodwin JS, Reistetter TA, et al. Comparison of functional status improvements among patients with stroke receiving postacute care in inpatient rehabilitation vs skilled nursing facilities. JAMA Netw Open. 2019;2(12):e1916646. doi:10.1001/jamanetworkopen.2019.16646

13. Porter ME. What is value in health care? N Engl J Med. 2010;363(26):2477-2481. doi:10.1056/NEJMp1011024

14. DeCherrie LV, Wardlow L, Ornstein KA, et al. Hospital at home services: an inventory of fee-for-service payments to inform Medicare reimbursement. J Am Geriatr Soc. 2021;69(7):1982-1992. doi:10.1111/jgs.17140

15. Jones CD, Bowles KH. Emerging challenges and opportunities for home health care in the time of COVID-19. J Am Med Dir Assoc. 2020;21(11):1517-1518. doi:10.1016/j.jamda.2020.09.018

16. Leff B, Burton L, Mader SL, et al. Hospital at home: feasibility and outcomes of a program to provide hospital-level care at home for acutely ill older patients. Ann Intern Med. 2005;143(11):798-808. doi:10.7326/0003-4819-143-11-200512060-00008

17. Levine DM, Ouchi K, Blanchfield B, et al. Hospital-level care at home for acutely ill adults. Ann Intern Med. 2020;172(2):77-85. doi:10.7326/M19-0600

18. Misra-Hebert AD, Felix C, Milinovich A, et al. Implementation experience with a 30-day hospital readmission risk score in a large, integrated health system: a retrospective study. J Gen Intern Med. 2022;37(12):3054-3061. doi:10.1007/s11606-021-07277-4

19. Jette DU, Stilphen M, Ranganathan VK, Passek SD, Frost FS, Jette AM. Validity of the AM-PAC “6-Clicks” inpatient daily activity and basic mobility short forms. Phys Ther. 2014;94(3):379-391. doi:10.2522/ptj.20130199

20. Jette DU, Stilphen M, Ranganathan VK, Passek S, Frost FS, Jette AM. Interrater reliability of AM-PAC “6-
Clicks” basic mobility and daily activity short forms. Phys Ther. 2015;95(5):758-766. doi:10.2522/ptj.20140174

21. Andres PL, Haley SM, Ni PS. Is patient-reported function reliable for monitoring postacute outcomes? Am J Phys Med Rehabil. 2003;82(8):614-621. doi:10.1097/01.PHM.0000073818.34847.F0

22. Haley SM, Andres PL, Coster WJ, Kosinski M, Ni P, Jette AM. Short-form activity measure for post-acute care. Arch Phys Med Rehabil. 2004;85(4):649-660. doi:10.1016/j.apmr.2003.08.098

23. Jette AM, Tao W, Norweg A, Haley S. Interpreting rehabilitation outcome measurements. J Rehabil Med. 2007;39(8):585-590. doi:10.2340/16501977-0119

24. Jette AM, Haley SM, Coster WJ, Ni P. Activity Measure for Post-Acute Care Short Forms 4.0 Instruction Manual. Pearson; 2020.

25. Home Health Patient-Driven Groupings Model. CMS. Accessed May 31, 2023. https://www.cms.gov/medicare/medicare-fee-for-service-payment/homehealthpps/hh-pdgm

26. Austin PC, Stuart EA. Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. Stat Med. 2015;34(28):3661-3679. doi:10.1002/sim.6607

27. Deutsch A, Palmer L, McMullen T, et al. A postacute care function process quality measure for the collection of standardized self-care and mobility data: development, implementation, and quality measure scores. Arch Phys Med Rehabil. 2022;103(6):1061-1069. doi:10.1016/J.APMR.2022.01.148

28. Seymour J, McNamee P, Scott A, Tinelli M. Shedding new light onto the ceiling and floor? a quantile regression approach to compare EQ-5D and SF-6D responses. Health Econ. 2010;19(6):683-696. doi:10.1002/hec.1505

29. Crompvoets EAV, Keuning J, Emons WHM. Bias and precision of continuous norms obtained using quantile regression. Assessment. 2021;28(6):1735-1750. doi:10.1177/1073191120910201

30. Falvey JR, Murphy TE, Gill TM, Stevens-Lapsley JE, Ferrante LE. Home health rehabilitation utilization among Medicare beneficiaries following critical illness. J Am Geriatr Soc. 2020;68(7):1512-1519. doi:10.1111/jgs.16412

31. Alcusky M, Ulbricht CM, Lapane KL. Postacute care setting, facility characteristics, and poststroke outcomes: a systematic review. Arch Phys Med Rehabil. 2018;99(6):1124-1140.e9. doi:10.1016/j.apmr.2017.09.005

32. Werner RM, Coe NB, Qi M, Konetzka RT. Patient outcomes after hospital discharge to home with home health care vs to a skilled nursing facility. JAMA Intern Med. 2019;179(5):617-623. doi:10.1001/jamainternmed.2018.7998

33. Jette DU, Stilphen M, Ranganathan VK, Passek SD, Frost FS, Jette AM. AM-PAC “6-Clicks” functional assessment scores predict acute care hospital discharge destination. Phys Ther. 2014;94(9):1252-1261. doi:10.2522/ptj.20130359

34. Pfoh ER, Hamilton A, Hu B, Stilphen M, Rothberg MB. The six-clicks mobility measure: a useful tool for predicting discharge disposition. Arch Phys Med Rehabil. 2020;101(7):1199-1203. doi:10.1016/j.apmr.2020.02.016

35. Warren M, Knecht J, Verheijde J, Tompkins J. Association of AM-PAC “6-Clicks” basic mobility and daily activity scores with discharge destination. Phys Ther. 2021;101(4):pzab043. doi:10.1093/ptj/pzab043

36. Harry M, Woehrle T, Renier C, Furcht M, Enockson M. Predictive utility of the Activity Measure for Post-Acute Care “6-Clicks” short forms on discharge disposition and effect on readmissions: a retrospective observational cohort study. BMJ Open. 2021;11(1):e044278. doi:10.1136/bmjopen-2020-044278

37. Herbold J, Rajaraman D, Taylor S, Agayby K, Babyar S. Activity Measure for Post-Acute Care “6-Clicks” basic mobility scores predict discharge destination following acute care hospitalization in select patient groups: a retrospective, observational study. Arch Rehabil Res Clin Transl. 2022;4(3):100204. doi:10.1016/J.ARRCT.2022.100204

38. Sterling MR, Grabowski DC, Shen MJ. Obtaining and paying for home care—navigating patients through the complex terrain of home care in the US. JAMA Intern Med. 2023;183(8):755-756. doi:10.1001/jamainternmed.2023.2072

39. Agarwal D, Werner RM. Effect of hospital and post-acute care provider participation in accountable care organizations on patient outcomes and Medicare spending. Health Serv Res. 2018;53(6):5035-5056. doi:10.1111/1475-6773.13023

40. Colla CH, Lewis VA, Stachowski C, Usadi B, Gottlieb DJ, Bynum JPW. Changes in use of postacute care associated with accountable care organizations in hip fracture, stroke, and pneumonia hospitalized cohorts. Med Care. 2019;57(6):444-452. doi:10.1097/MLR.0000000000001121

41. Lage DE, Rusinak D, Carr D, Grabowski DC, Ackerly DC. Creating a network of high-quality skilled nursing facilities: preliminary data on the postacute care quality improvement experiences of an accountable care organization. J Am Geriatr Soc. 2015;63(4):804-808. doi:10.1111/jgs.13351

42. Barnett ML, Mehrotra A, Grabowski DC. Postacute care—the piggy bank for savings in alternative payment models? N Engl J Med. 2019;381(4):302-303. doi:10.1056/NEJMp1901896

43. Porter ME, Lee TH. The strategy that will fix health care. Harvard Business Review. October 2013. Accessed January 29, 2020. https://hbr.org/2013/10/the-strategy-that-will-fix-health-care

44. Johnson JK, Lapin B, Bethoux F, Skolaris A, Katzan I, Stilphen M. Patient versus clinician proxy reliability of the AM-PAC “6-Clicks” basic mobility and daily activity short forms. Phys Ther. 2022;102(6):pzac035. doi:10.1093/ptj/pzac035