Impact of Functional Recovery on Patients Having Heart Surgery

ABSTRACT

Objective: To describe the results of a program developed to manage institutional postacute care (IPAC) (postacute skilled nursing, inpatient rehabilitation facility, and long-term acute care) in a CMS Bundled Payments for Care Improvement (BPCI) project for coronary artery bypass graft (CABG) surgery.

Study Design: We compared pre- and postutilization patterns during a 3-year period by evaluating risk-adjusted national, state, and other BPCI participant comparisons using a difference-in-differences (DID) analysis in a large urban community tertiary center with a CABG surgery program. Included in the analysis were all Medicare patients receiving CABG surgery at the institution (n = 504), across the nation (n = 213,423), and at other BPCI institutions (n = 4939).

Methods: The intervention included (1) use of a standardized tool for evaluation and prognostication of patient placement, (2) programmatic changes to manage patient functional recovery, and (3) patient and family engagement in postacute placement and functional recovery plan.

Results: Physical therapist/occupational therapist time with patients who had undergone CABG surgery increased by more than 179% between the pre- and postintervention periods. This was associated with a 41.2% and 51.6% decline in IPAC use at the institution on an observed basis and adjusted basis, respectively. DID comparison demonstrated a 14.40% (95% CI, –19.30% to –9.60%) greater reduction at the target hospital than at other participating BPCI hospitals.

Conclusions: A strong association exists between a focused patient functional recovery program and IPAC use reduction after CABG surgery. Using a structured approach to clinical analytics and hypothesis testing of redesign efforts when managing postacute care populations removes waste from care delivery.

Am J Manag Care. 2024;30(10):504-509. https://doi.org/10.37765/ajmc.2024.89619

Takeaway Points

Managing episode-of-care risk contracts requires care redesign to improve patient care and reduce postacute spending. We describe the impact of a directed approach—measuring and improving patient functional status—on institutional postacute care utilization and readmissions in patients undergoing open heart surgery. By using risk-adjusted outcomes measures and specific care redesign developed by physical/occupational therapists, we demonstrate:

• methods of defining opportunity gaps in postacute care utilization;
• a model of managing patient recovery using existing resources at all hospitals; and
• the impact on patient recovery that the deployed functional recovery program had on patients using difference-in-differences analysis.

The Institute of Medicine has estimated that 30% of health care spending in the US does not contribute to better patient outcomes and could be considered waste.¹ Postacute care (PAC) that is provided after hospital discharge demonstrates high variation and has been identified as an area of potential overuse in the Medicare population.^2,3 Institutional PAC (IPAC), which includes skilled nursing facility (SNF), inpatient rehabilitation facility (IRF), and long-term acute care use, is a large driver of IPAC spending. A recently published analysis demonstrated that although discharge to SNF or IRF increased from an adjusted 21.0% in 2000 to 26.3% in 2015 among Medicare beneficiaries, there is little evidence linking this increase to better patient outcomes.⁴

To create incentives to manage waste, the Center for Medicaid and Medicare Innovation (CMMI) has implemented population-based programs focused on clinical and financial outcomes as opposed to volume of services.⁵ Initiated in 2013, Bundled Payments for Care Improvement (BPCI) is a CMMI payment redesign initiative providing participants an opportunity to take financial risk for care delivered during an index hospitalization and the ensuing 90 days within an episode-of-care model by focusing on better organization of care and the removal of waste.^6,7 Because the Medicare diagnosis related group (DRG) payment for the episode initiating hospitalization is set, any savings resulting in payment, or losses resulting in recoupment, occur during the PAC period. To be successful, participating organizations must lower PAC spending, which is largely driven by IPAC and rehospitalization following discharge from the index hospitalization.⁸

A review of the literature demonstrated some clinical programs that focused on PAC clinical management but a paucity of evidence-based approaches to improve the quality of functional recovery or decrease the costs of PAC.^9,10 In this article, we describe a program our hospital developed that systematically improved the functional recovery of patients after heart surgery and provided close clinical postdischarge follow-up. We describe the process of care redesign selection and implementation and discuss some of the barriers and solutions to the transformation of care delivery during this project. We evaluate the impact of the program by comparing our institution with other participants in the BPCI program using a difference-in-differences (DID) analysis of IPAC utilization and unadjusted trends in rehospitalization.

METHODS

Study Setting

We examined changes in IPAC and rehospitalization outcomes for patients undergoing coronary artery bypass graft (CABG) surgery following implementation of functional and clinical recovery interventions at OhioHealth Riverside Methodist Hospital (RMH). RMH is a large tertiary community hospital in central Ohio that became a BPCI participating organization on July 1, 2015. This research was deemed exempt from review by the OhioHealth Institutional Review Board.

To evaluate and manage PAC opportunities, we documented current performance, designed future state of functional recovery, set goals, and developed an implementation plan to close gaps in outcomes.¹¹ As part of value stream mapping for functional recovery of patients, a specific work team led by physical therapists and occupational therapists (PTs/OTs) was chartered to develop and test a systematic approach to recover patients functionally.

The described interventions were the result of an interactive evaluation of impact using outcomes measures to continuously improve the process of patient functional recovery and rehospitalization mitigation.

Care Redesign: Functional Recovery

The functional recovery care redesign team developed standard workflows to approach patients in a consistent manner to assure continuity of recovery across patients with different needs and the ability to test the hypothesis that specific process change had an effect on IPAC utilization rates.

A systematic approach to assess patient functionality during the index hospitalization was developed to achieve 3 goals: (1) assess patient functionality within the framework of the patient’s anticipated postacute setting, (2) develop a plan or prescription of functional recovery during the time the patient was hospitalized, and (3) engage the patient and family on the course of prescribed functional recovery during and after the index hospitalization.

The Activity Measure for Post-Acute Care (AM-PAC) inpatient short form was chosen as the standard tool for patient assessment.¹² The AM-PAC assessment consists of 2 sets of short forms called “6 clicks” that assess basic mobility, such as walking and moving from one position to another, and activities of daily living, such as dressing and toileting. Each item is scored from 1 (unable to perform) to 4 (independent) based on the amount of difficulty a patient has or how much help is needed from another person to complete the task. The sum of the AM-PAC scores ranges from 0 to 24, with lower scores equating to lower levels of function.

To better prepare patients and caregivers for going home, the team developed an educational program for caregiver engagement and training. This included hands-on experience, where caregivers facilitate patients’ mobility and self-care tasks with the patient. Caregivers assist patients with the 12 activities identified in the 2 AM-PAC 6-clicks measures. All PT/OT clinicians on cardiac surgery floors were trained, and the standard work was embedded into clinical competencies. PTs focused on improving mobility and strength, training caregivers in assisting with comfort, and managing patient physical recovery. OTs focused on improving performance with activities of daily living. Utilizing continuous quality improvement techniques guided by outcomes, we used the AM-PAC as a prognostic indicator for PAC placement, generally discharging those patients with an AM-PAC score of less than 15 to IPAC and those with a score of 20 or greater to home. Patients with AM-PAC scores of 15 to 19 were deemed at risk of being discharged to IPAC and were the focus of more intense functional recovery efforts.

Care Redesign: Clinical Recovery

The clinical recovery team focused on several initiatives including standardization of inpatient care through protocols focused on cardiac, pulmonary, and renal management based on guidelines. Patients received a follow-up visit within 7 to 10 days post discharge with the cardiothoracic surgery advanced practice nurse (CTS APN) or surgeon. CTS APN coverage was expanded to 24/7 to increase patient access to surgical providers. The CTS APN received real-time text alerts when a patient presented to the emergency department within 90 days of hospital discharge for CABG and collaborated with the emergency department providers to develop a treatment plan and to coordinate the next level of care, which could include discharge home, outpatient treatment at heart failure or atrial fibrillation clinics, outpatient thoracentesis, or hospital admission when appropriate.

Analysis

The CABG episode population was identified using DRG codes defined by CMS.⁷ The target institution that deployed the intervention (RMH) was compared with participating BPCI hospitals performing similar surgeries during the preimplementation and postimplementation of the intervention period. All participating BPCI groups were mutually exclusive. To address issues of counterfactual assumptions, we included a comparison of all hospitals nationally providing CABG services to Medicare patients during the same time frames.

To identify organizations participating in BPCI, we used CMS Medicare Provider Analysis and Review (MedPAR) files from 2013 to 2016. Comparison BPCI contract time frames were identified through MedPAR file contract start dates, which allowed us to define preintervention and postintervention performances.⁷

Time frames for evaluation were separated into 2 periods for all groups: (1) preintervention period (RMH BPCI, October 1, 2013, to June 30, 2015; national non-BPCI hospitals providing CABG, October 1, 2013, to June 30, 2015; comparison BPCI hospitals, October 1, 2013, to 6 months prior to contract start date); and (2) postintervention period (RMH BPCI, July 1, 2015, to September 30, 2016; national non-BPCI hospitals providing CABG, July 1, 2015, to September 30, 2016; comparison BPCI hospitals, contract start date to September 30, 2016).

IPAC Outcome Measure

We used the MedPAR files to compare performance prior to intervention implementation and evaluate change over time at the target institution. Performance over time was then compared with that of non–target hospital BPCI participants and non-BPCI hospitals nationally for evaluation of trends external to the BPCI program. To validate our comparison groups, we compared characteristics and unadjusted outcomes across the 3 groups using standard tests of association. This comparison allowed us to identify potential confounders and limitations due to significant differences in the populations.

Additionally, our DID approach relied on the assumption that comparison participants’ IPAC performance was the same in the preintervention period as post intervention; therefore, we further tested for differences in IPAC performance prior to implementation across groups. We also addressed potential time-varying confounders that would limit our comparison of groups with different preimplementation and postimplementation periods.

We then fit a multivariable model using CMS data for all CABG cases with IPAC utilization as the dependent variable and included independent variables describing patient demographics, comorbidities, and other procedures occurring during the index hospitalization that might affect patient functionality. Comorbid and procedural variables were created using Agency for Healthcare Research and Quality Clinical Classifications Software.¹³ The model discriminated well, with a C statistic of 0.76. (Full model parameters and ORs are available from the authors.) These scores were aggregated to create an expected rate of IPAC utilization across the 3 groups, which allowed comparison of observed/expected rates.

Significance in trends in performance over the preimplementation and postimplementation periods was tested using χ² analysis for observed rates of IPAC utilization. To compare risk-adjusted differences over time periods, we used generalized linear models and entered the predicted probability and the time to determine significance. We used these scores in the DID analysis comparing the target institution’s performance change from preimplementation to postimplementation periods against other BPCI participants. All analysis was done in SAS 9.3 (SAS Institute Inc).

Rehospitalization Measure

As a measure of effectiveness and safety at the target institution, we continuously evaluated rehospitalization density, defined as the number of rehospitalizations per 100 discharges, during implementation of both the functional recovery and clinical interventions. We evaluated risk adjustment models for readmission rates over 90 days and found that the discrimination was weak (receiver operating characteristic, 0.61) and therefore reported densities on an unadjusted trended basis without comparison.

Key Process Indicators

Key process indicators were defined by the PT/OT team and used to dynamically track effectiveness of deployment of process change and to evaluate the association between the care redesign and the outcome of IPAC utilization. Indicators were developed from internal electronic health record data and included the mean amount of time that a PT or OT spent with a CABG patient.

RESULTS

Using the MedPAR files, we identified a total of 213,423 CABG cases nationally during the time frame. Cases by cohort and demographic information for each comparative cohort are available in Table 1. Statistical comparisons between RMH and the other BPCI hospitals are included. There were significant differences between RMH and other BPCI hospitals in the percentage of patients older than 80 years and African American patients.

Table 2 displays observed, predicted, and observed-to-predicted ratio (O/P) rates across the analytic cohorts. During this time frame, national rates of IPAC utilization increased slightly on an observed basis but were flat after adjustment, supporting our counterfactual assumption. Rates of IPAC utilization at comparison BPCI hospitals decreased on both an observed and adjusted basis, which was not significant. IPAC utilization rates at RMH decreased on both an observed and adjusted basis, which achieved statistical significance. Evaluating preimplementation period results at RMH demonstrated a large opportunity gap to reduce overutilization of IPAC compared with the nation. The adjusted rate of IPAC utilization at RMH decreased from 63% higher than national rates during the preimplementation period to 21% below national rates during the intervention period. This is represented graphically in Figure 1, which displays the O/P for each of the 3 cohorts across the time frames. DID analysis demonstrated a significant reduction in IPAC utilization in non-RMH BPCI hospitals of 3.06% (P = .02) when comparing the preimplementation and postimplementation periods and a 17.50% reduction (P < .001) for RMH. Table 3 compares the adjusted differences between the 2 groups, demonstrating a 14.40% (95% CI, –19.30% to –9.60%) greater reduction of IPAC utilization at RMH vs comparison BPCI hospitals during the time frames.

Key Process Indicator Changes

Figure 2 displays the changes in the mean number of minutes PTs/OTs spent with patients during the specific periods. The mean time spent with patients increased from 62.1 minutes during hospitalizations in the preintervention period to 173 minutes during the postintervention period. This 179% increase in time spent with patients was significant, with a P value less than .0001.

Rehospitalizations

The density of rehospitalization fell between the pre- and postintervention periods from 27.8 per 100 discharges to 22.0 per 100 discharges, a decline of 21% on an unadjusted basis at the target institution.

DISCUSSION

We deployed several strategies to achieve the outcomes demonstrated in this publication. First, we understood the opportunity gap in recovering patients from surgery and developed a program to modify this risk. There is good evidence that a patient’s mobility and functionality are reduced by hospitalization, but there are few protocols to improve functionality during hospitalization.^14-16 We used this knowledge to engage PTs/OTs in developing a program using patient assessment presurgery and to prescribe specific actions for functional recovery. This engaged clinicians, patients, and families on specific actions and goals for recovering the functional decompensation that occurred due to the surgical hospitalization. Second, we engaged leadership in the stated goals of IPAC reduction and created system accountabilities that were reviewed on a monthly basis by leadership. Finally, we used risk-adjusted outcomes of IPAC for PT/OT, surgeon, and leadership feedback on a weekly basis that served to remove doubt about performance attributed to patient differences and drive continuous quality improvement.

Emerging evidence suggests that reduced functionality is associated with increased readmission risk after discharge from inpatient rehabilitation hospitalization.¹⁷ We did improve inpatient protocols around comorbidity management, provide postsurgical patients with greater access to surgeons post hospitalization, and deploy a dedicated clinician to manage patients out of the emergency department at the target hospital, but we believe that the functional recovery program also contributed to the observed rehospitalization reduction. Because the target hospital drew patients broadly across Ohio, we were able to examine readmission reductions among local and distant patients on an unadjusted basis. Although the rehospitalization reduction was more pronounced at the target hospital, we noted reductions in populations that were geographically distant from the target hospital where the CTS APN intervention was not available. With our increasing knowledge that functional recovery of patients is an important aspect of providing value, further research needs to be done on the associations between focused physical recovery and IPAC and readmission risk across a broader spectrum of surgical hospitalizations. We found substantial barriers to the BPCI work at RMH requiring alignment of clinicians’ beliefs, including current performance was all that was achievable, resistance to moving discharge planning from case management to PTs/OTs, and concerns we would be putting patients at risk by changing care. In the end, strong leadership deploying change management through “nudge unit” techniques and defendable analytics kept the focus on improving patient outcomes.¹⁸ Value-based payments can provide the incentive to redesign care that aligns with patient-centered care, but we found that the change represented by the redesign needs to be carefully managed using continuous feedback and leadership engagement as demonstrated in our initiative.

In terms of other BPCI hospital performance, we hypothesize a more generic approach to care redesign than we deployed. A number of recent randomized controlled trials and observational studies that specify generic process changes have demonstrated small effects on improving clinical quality or reducing costs.^19-22 The basis of these findings may be in the process changes deployed in these studies, which were not focused on identifying specific patient risks or providing techniques that could modify the risk. We believe our approach of focusing on specific risks of waste (poor functional recovery and inattention to patients returning to the emergency department) along with the opportunity gap analysis demonstrating excess IPAC utilization created the large improvements at RMH compared with other BPCI hospitals.

This study used national comparative risk-adjusted data that included all providers of CABG episodes and a select group of hospitals engaged in care redesign through participation in BPCI both for exploratory analysis and program evaluation. A DID study design was used to show a significant change in IPAC utilization at the target institution. Methods used were consistent with evolving expectations of quality improvement studies and illuminate the level of analytics needed to manage care redesign in a way that produces demonstrable change in patient outcomes during an episode of care.²³ The target institution, although larger than most providing these services, used clinical resources available at most institutions, supporting both generalizability and exportability of the care redesign described in this manuscript. An internal audit committee tracking the CMS payments and the costs of the program demonstrated a large return on investment from this BPCI contract in shared savings, easily offsetting the increased costs in PT/OT and CTS-APN time spent on the project.

Limitations

Our study has several limitations. With an observational study design, we can only draw conclusions between the association of the PT/OT program and IPAC utilization reduction. The consistency of outcomes after implementation and the comparative trends from across the nation and other BPCI participants strengthen the associations but do not prove impact. Also, the time-varying period differences when comparing non-RMH BPCI hospitals with RMH in the DID analysis may introduce confounders. Furthermore, state policies should be examined as potential confounders. We do not report the costs or the savings of the program; we report only discharge to IPAC vs any IPAC use over the entire episode. We also did not implement a patient survey to better understand patient perceptions of the program, although anecdotal feedback from patients and families was positive with no complaints.

CONCLUSIONS

Our findings demonstrate a strong association between an organized functional recovery program and IPAC placement after CABG surgery. The results we demonstrate at RMH easily exceeded those demonstrated by other BPCI hospitals that had the same incentive during the same time frame. By using existing resources and disciplined care redesign, we believe this work is transferable to other hospitals. 

Author Affiliations: OhioHealth (RJS, LM, GB, KV, AC, JS, LW, KL, TC-G), Columbus, OH; Applied Health Services (RJS, ES), Columbus, OH.

Source of Funding: This project was funded by OhioHealth.

The statements contained in this document are solely those of the authors and do not necessarily reflect the views or policies of CMS. The authors assume responsibility for the accuracy and completeness of the information contained in this document.

Author Disclosures: Ms Will is an employee of OhioHealth. 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 (RJS, LM, JS, LW, KL, ES, TC-G); acquisition of data (RJS, GB, KV, AC, KL, TC-G); analysis and interpretation of data (RJS, LM, GB, KV, AC, JS, KL, ES, TC-G); drafting of the manuscript (RJS, LM, ES, TC-G); critical revision of the manuscript for important intellectual content (RJS, LM, GB, JS, LW, KL, ES, TC-G); statistical analysis (RJS, KV, AC, ES, TC-G); provision of patients or study materials (LW, TC-G); administrative, technical, or logistic support (LM, GB, TC-G); and supervision (TC-G).

Address Correspondence to: Lauren McKown, JD, MPH, OhioHealth, 3430 OhioHealth Pkwy, Columbus, OH 43202. Email: lauren.mckown@ohiohealth.com.

REFERENCES

1. Institute of Medicine. Best Care at Lower Cost: The Path to Continuously Learning Health Care in America. The National Academies Press; 2013.

2. Garret B. Post-acute care and Medicare solvency. Urban Institute. October 2023. Accessed September 17, 2024. https://www.urban.org/sites/default/files/2023-10/Post-Acute%20Care%20and%20Medical%20Solvency.pdf

3. Li Q, Rahman M, Gozalo P, Keohane LM, Gold MR, Trivedi AN. Regional variations: the use of hospitals, home health, and skilled nursing in traditional Medicare and Medicare Advantage. Health Aff (Millwood). 2018;37(8):1274-1281. doi:10.1377/hlthaff.2018.0147

4. Werner RM, Konetzka RT. Trends in post-acute care use among Medicare beneficiaries: 2000 to 2015. JAMA. 2018;319(15):1616-1617. doi:10.1001/jama.2018.2408

5. Innovation models. CMS. Updated September 10, 2024. Accessed September 6, 2023. https://www.cms.gov/priorities/innovation/models#views=models

6. BPCI Advanced. CMS. Accessed September 6, 2023. https://www.cms.gov/priorities/innovation/innovation-models/bpci-advanced

7. Bundled Payments for Care Improvement (BPCI) initiative: general information. CMS. Accessed September 18, 2023. https://www.cms.gov/priorities/innovation/innovation-models/Bundled-Payments

8. Huckfeldt PJ, Mehrotra A, Hussey PS. The relative importance of post-acute care and readmissions for post-discharge spending. Health Serv Res. 2016;51(5):1919-1938. 2016;51(5):1919-1938. doi:10.1111/1475-6773.12448

9. Alqahtani AA. Atrial fibrillation post cardiac surgery trends toward management. Heart Views. 2010;11(2):57-63. doi:10.4103/1995-705X.73212

10. Rich MW, Beckham V, Wittenberg C, Leven CL, Freedland KE, Carney RM. A multidisciplinary intervention to prevent the readmission of elderly patients with congestive heart failure. N Engl J Med. 1995;333(18):1190-1195. doi:10.1056/NEJM199511023331806

11. Jones D, Womack J. Seeing the Whole Value Stream. 2nd ed. Lean Enterprise Institute Inc; 2011.

12. Geelen SJG, Valkenet K, Veenhof C. Construct validity and inter-rater reliability of the Dutch activity measure for post-acute care “6-clicks” basic mobility form to assess the mobility of hospitalized patients. Disabil Rehabil. 2019;41(21):2563-2569. doi:10.1080/09638288.2018.1471525

13. Clinical Classifications Software Refined (CCSR). Healthcare Cost and Utilization Project. April 2024. Updated April 29, 2024. Accessed September 18, 2023. https://hcup-us.ahrq.gov/toolssoftware/ccsr/ccs_refined.jsp

14. Gill TM, Beavers DP, Guralnik JM, et al; LIFE Study Investigators. The effect of intervening hospitalizations on the benefit of structured physical activity in promoting independent mobility among community-living older persons: secondary analysis of a randomized controlled trial. BMC Med. 2017;15(1):65. doi:10.1186/s12916-017-0824-6

15. Pisani MA, Albuquerque A, Marcantonio ER, et al. Association between hospital readmission and acute and sustained delays in functional recovery during 18 months after elective surgery: the Successful Aging after Elective Surgery Study. J Am Geriatr Soc. 2017;65(1):51-58. doi:10.1111/jgs.14549

16. Anderson L, Thompson DR, Oldridge N, et al. Exercise-based cardiac rehabilitation for coronary heart disease. Cochrane Database Syst Rev. 2016;2016(1):CD001800. doi:10.1002/14651858.CD001800.pub3

17. Fisher SR, Graham JE, Krishnan S, Ottenbacher KJ. Predictors of 30-day readmission following inpatient rehabilitation for patients at high risk for hospital readmission. Phys Ther. 2016;96(1):62-70. doi:10.2522/ptj.20150034

18. Patel MS, Volpp KG, Asch DA. Nudge units to improve the delivery of health care. N Engl J Med. 2018;378(3):214-216. doi:10.1056/NEJMp1712984

19. Hussey PS, Wertheimer S, Mehrotra A. The association between health care quality and cost: a systematic review. Ann Intern Med. 2013;158(1):27-34. doi:10.7326/0003-4819-158-1-201301010-00006

20. Peikes D, Chen A, Schore J, Brown R. Effects of care coordination on hospitalization, quality of care, and health care expenditures among Medicare beneficiaries: 15 randomized trials. JAMA. 2009;301(6):603-618. doi:10.1001/jama.2009.126

21. McWilliams JM. Cost containment and the tale of care coordination. N Engl J Med. 2016;375(23):2218-2220. doi:10.1056/NEJMp1610821

22. McWilliams JM, Schwartz AL. Focusing on high-cost patients — the key to addressing high costs? N Engl J Med. 2017;376(9):807-809. doi:10.1056/NEJMp1612779

23. Grady D, Redberg RF, O’Malley PG. Quality improvement for quality improvement studies. JAMA Intern Med. 2018;178(2):187. doi:10.1001/jamainternmed.2017.6875