0
We're unable to sign you in at this time. Please try again in a few minutes.
Retry
We were able to sign you in, but your subscription(s) could not be found. Please try again in a few minutes.
Retry
There may be a problem with your account. Please contact the AMA Service Center to resolve this issue.
Contact the AMA Service Center:
Telephone: 1 (800) 262-2350 or 1 (312) 670-7827  *   Email: subscriptions@jamanetwork.com
Error Message ......
Original Investigation |

Thirty-Day Hospital Readmission Following Discharge From Postacute Rehabilitation in Fee-for-Service Medicare Patients FREE

Kenneth J. Ottenbacher, PhD, OTR1; Amol Karmarkar, PhD, MPH1; James E. Graham, PhD, DC1; Yong-Fang Kuo, PhD2; Anne Deutsch, RN, PhD, CRRN3; Timothy A. Reistetter, PhD, OTR4; Soham Al Snih, MD, PhD1; Carl V. Granger, MD5,6
[+] Author Affiliations
1Division of Rehabilitation Sciences, University of Texas Medical Branch (UTMB), Galveston
2Department of Preventive Medicine and Community Health, UTMB
3RTI International, Washington, DC, and Rehabilitation Institute of Chicago, Chicago, Illinois
4Department of Occupational Therapy, UTMB
5Uniform Data System for Medical Rehabilitation, Buffalo, New York
6Department of Medicine, University at Buffalo, Buffalo, New York
JAMA. 2014;311(6):604-614. doi:10.1001/jama.2014.8.
Text Size: A A A
Published online

Importance  The Centers for Medicare & Medicaid Services recently identified 30-day readmission after discharge from inpatient rehabilitation facilities as a national quality indicator. Research is needed to determine the rates and factors related to readmission in this patient population.

Objective  To determine 30-day readmission rates and factors related to readmission for patients receiving postacute inpatient rehabilitation.

Design, Setting, and Patients  Retrospective cohort study of records for 736 536 Medicare fee-for-service beneficiaries (mean age, 78.0 [SD, 7.3] years) discharged from 1365 inpatient rehabilitation facilities to the community in 2006 through 2011. Sixty-three percent of patients were women, and 85.1% were non-Hispanic white.

Main Outcomes and Measures  Thirty-day readmission rates for the 6 largest diagnostic impairment categories receiving inpatient rehabilitation. These included stroke, lower extremity fracture, lower extremity joint replacement, debility, neurologic disorders, and brain dysfunction.

Results  Mean rehabilitation length of stay was 12.4 (SD, 5.3) days. The overall 30-day readmission rate was 11.8% (95% CI, 11.7%-11.8%). Rates ranged from 5.8% (95% CI, 5.8%-5.9%) for patients with lower extremity joint replacement to 18.8% (95% CI, 18.8%-18.9%). for patients with debility. Rates were highest in men (13.0% [ 95% CI, 12.8%-13.1%], vs 11.0% [95% CI, 11.0%-11.1%] in women), non-Hispanic blacks (13.8% [95% CI, 13.5%-14.1%], vs 11.5% [95% CI, 11.5%-11.6%] in whites, 12.5% [95% CI, 12.1%-12.8%] in Hispanics, and 11.9% [95% CI, 11.4%-12.4%] in other races/ethnicities), beneficiaries with dual eligibility (15.1% [95% CI, 14.9%-15.4%], vs 11.1% [95% CI, 11.0%-11.2%] for no dual eligibility), and in patients with tier 1 comorbidities (25.6% [95% CI, 24.9%-26.3%], vs 18.9% [95% CI, 18.5%-19.3%] for tier 2, 15.1% [95% CI, 14.9%-15.3%] for tier 3, and 9.9% [95% CI, 9.9%-10.0%] for no tier comorbidities). Higher motor and cognitive functional status were associated with lower hospital readmission rates across the 6 impairment categories. Adjusted readmission rates by state ranged from 9.2% to 13.6%. Approximately 50% of patients rehospitalized within the 30-day period were readmitted within 11 days of discharge. Medicare Severity Diagnosis-Related Group codes for heart failure, urinary tract infection, pneumonia, septicemia, nutritional and metabolic disorders, esophagitis, gastroenteritis, and digestive disorders were common reasons for readmission.

Conclusions and Relevance  Among postacute rehabilitation facilities providing services to Medicare fee-for-service beneficiaries, 30-day readmission rates ranged from 5.8% to 18.8% for selected impairment groups. Further research is needed to understand the causes of readmission.

Figures in this Article

The Patient Protection and Affordable Care Act1 created the hospital readmission reduction program to reduce readmissions and improve patient transitions from acute care. Research examining 30-day readmission has focused on patients discharged from acute care hospitals.2,3 Patients discharged to postacute care institutional settings have been excluded from previous research on hospital readmission.36 Little research on hospital readmission has been reported for patients receiving postacute services.

The Medicare Payment Advisory Commission (MedPAC)7 recently began tracking hospital readmission for postacute care settings. In 2010, 12% of patients discharged from inpatient rehabilitation facilities to the community were readmitted to acute care hospitals within 30 days.7 The MedPAC report did not stratify these cases by impairment categories.7

It is important to study readmission after discharge from postacute rehabilitation for 3 reasons. First, it is known that certain patients at high risk for readmission are commonly referred to postacute rehabilitation (eg, stroke and hip fracture).4,6 Second, the Centers for Medicare & Medicaid Services (CMS) recently identified 30-day readmission as a national quality indicator for inpatient rehabilitation facilities.8,9 Reporting will be required by the CMS and is consistent with the Affordable Care Act.1,8 Third, the CMS has proposed bundled payment models (acute and postacute care) to align performance incentives and contain costs.10,11 Understanding the ramifications of bundling requires accurate information regarding readmission rates for patients receiving postacute services.

We examined records for patients from the 6 largest impairment categories receiving postacute inpatient rehabilitation. These include patients with stroke, lower extremity fracture, lower extremity joint replacement, debility, neurologic disorders, and brain dysfunction.4 Debility as a rehabilitation impairment category is defined as generalized deconditioning not attributable to neurologic, orthopedic, or cardiopulmonary diagnoses. We were interested in answering the following questions: What is the 30-day readmission rate following discharge from inpatient rehabilitation? Are there differences in readmission rates across impairment categories? Are readmissions associated with patient sociodemographics, clinical characteristics, functional status, or facility factors?

Source of Data

Data analyzed were from 4 CMS files: (1) Medicare Provider Analysis and Review; (2) Inpatient Rehabilitation Facility–Patient Assessment Instrument (IRF-PAI); (3) Beneficiary Summary file; and (4) Inpatient Rehabilitation Facility Rate settings.

Obtaining CMS Files

Stay-level Medicare data in the Research Identifiable Format were acquired for 2006 through 2011. The study was approved by the University of Texas Medical Branch institutional review board. Use of Medicare data files was reviewed by the CMS and met all federal privacy and confidentiality requirements. A Data Use Agreement was completed following CMS guidelines.

Facilities

Rehabilitation facilities are divided into hospital-based units or freestanding centers.12 Sixty-two percent of inpatient rehabilitation stays in 2010 were in hospital-based units.7

Study Population

Our population was Medicare fee-for-service patients discharged directly from short-term acute care hospitals to inpatient rehabilitation facilities. The patients’ rehabilitation admission diagnoses placed them in 1 of the 6 rehabilitation impairment categories described previously. Ninety-five percent of all patients admitted to inpatient rehabilitation facilities during the study period were from acute care hospitals.7 We selected the 6 impairment categories based on the 2013 MedPAC report4 indicating that approximately 75% of Medicare fee-for-service patients receiving inpatient rehabilitation were classified into 1 of these impairment groups.

Variables
Hospital Readmission

We examined readmission to an acute care hospital that occurred within 30 days following discharge from the rehabilitation facility to the community. In the Medicare rehabilitation inpatient data, community includes home, board-and-care, transitional living, and assisted living residence.13 Descriptions of the rehabilitation impairment categories, numbers of patients in each category, and the International Classification of Diseases, Ninth Revision, Clinical Modification diagnostic codes associated with the categories are included in eAppendix 1 in Supplement.

Sociodemographics

Sociodemographic characteristics included age at rehabilitation admission (continuous and categories: ≤74, 75-84, ≥85 years), sex, race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, and other), married (yes/no), prehospital living status (living alone vs with family/relatives, friends, attendant, or others), disability (receiving disability benefits; yes/no), and Medicaid dual eligibility (yes/no). These variables were extracted from the rehabilitation assessment data (IRF-PAI) and Beneficiary Summary files.

Functional Status

Functional status items are included in the Medicare rehabilitation assessment data (IRF-PAI). The IRF-PAI is administered by physical or occupational therapists (nurses may administer the self-care, transfer, or sphincter control items). The 18 items cover 6 domains: self-care, sphincter control, transfers, locomotion, communication, and social cognition.14 Items are assigned to 1 of 7 levels of function, ranging from complete dependence (level 1) to independence (level 7). Functional status ratings can be divided into motor and cognitive subscales. The functional status items are administered to the patient at admission and within 36 hours of discharge. Motor function ratings range from 13 to 91. Cognitive function ratings range from 5 to 35. The reliability and validity of the ratings have been widely studied and found to be psychometrically adequate.15,16 The motor and cognitive ratings at admission are used to determine patient CMG assignment.12,17

Case-Mix Group

Case-mix groups (CMGs) are used to determine Medicare fee-for-service payment for individual patient stays.12 Rehabilitation patients are classified into a CMG at the time of admission based on impairment category, motor and cognitive functional status, and age. Each impairment category has a set number of CMGs (eg, stroke has 10). Each CMG and comorbidity tier is associated with a projected length of stay and base-level reimbursement provided to the facility.18 Medicare reimbursement rates are also adjusted by several factors (eg, rural vs urban location).

Comorbidity Tier

The CMS developed 3 tiers for inpatient rehabilitation reimbursement: tier 1 (high reimbursement) to tier 3 (low reimbursement).19 In 2012, there were 8 comorbid conditions in tier 1, 11 in tier 2, and 924 in tier 3. An example of a tier 1 comorbidity is renal dialysis.4 Some forms of diabetes are classified as tier 3 comorbidities. Each CMG is paired with the patient’s comorbidity tier status (tier 1, 2, 3, or no tier-level comorbidity) and assigned a weight that reflects the resources required to provide treatment to the average patient with that clinical presentation. Patients are assigned to the tier with the highest level of reimbursement.

Clinical Variable and Facility Factors

Length of stay in days (continuous) was the clinical variable assessed. Facility factors included type of facility (hospital-based unit vs free-standing), location (urban or rural), ownership (government, nonprofit, and for profit), and location by state.

Reason for Readmission

Readmissions to acute care hospitals were categorized using Medicare Severity Diagnosis-Related Groups (MS-DRGs). MS-DRGs classify the reason for a hospitalization based on a series of principal and secondary diagnoses as well as procedure codes.20 MS-DRGs incorporate the severity of the patient’s condition using codes that mark the presence of complications and comorbidities.20 We identified the top 25 MS-DRG codes at readmission for each of the impairment categories using the acute care claims associated with the readmission.

Data Analysis

Summary statistics (means and SDs and/or column percentages) of patient characteristics were calculated for each rehabilitation impairment category and for the combined sample. Thirty-day readmission rates and 95% CIs were computed for each category of the patient- and facility-level variables. We used the normal approximation method for binomial CIs.21 The top 25 reasons for readmission, based on MS-DRGs, were identified within each rehabilitation impairment category.

State-specific risk-standardized readmission rates were calculated using hierarchical generalized linear mixed models to account for clustering of patients within states. The models adjusted for 8 patient demographic and clinical variables: age, sex, race/ethnicity, living situation, rehabilitation impairment category, tier comorbidities, and admission motor and cognitive functioning. The final rates were obtained by taking the ratio of predicted to expected readmissions for each state and multiplying by the global unadjusted rate.22 Data were analyzed using IBM SPSS Statistics 21 (IBM) and SAS 9.3 (SAS Institute Inc).

Patients receiving rehabilitation from January 2006 through November 2011 in 1 of the 6 impairments were the eligible sample (N = 1 705 109). We excluded patients with an atypical rehabilitation stay (length of stay >30 days) (n = 26 750), patients admitted to inpatient rehabilitation facilities from settings other than acute care hospitals (n = 68 319), patients living in noncommunity settings prior to their admission (n = 22 559), and those who died during the rehabilitation stay (n = 3005). Also excluded were patients with rehabilitation stay other than “initial rehabilitation” (n = 46 058) and those with rehabilitation program interruptions (n = 14 471). Our sample included beneficiaries in Medicare’s traditional fee-for-service plan 65 and older, including those eligible for disability and those who had dual (Medicare and Medicaid) eligibility, resulting in exclusion of additional cases (n = 448 164). We selected only those patients discharged to community settings after a rehabilitation stay and excluded persons transferred on the day of discharge to acute care, long-term care hospitals, or nursing homes (n = 293 211). Last, we considered only the first rehabilitation stay for all patients, those who survived 30 days after a rehabilitation stay, and those who stayed in community settings for 30 days or who were readmitted to acute care hospitals within 30 days of discharge from inpatient rehabilitation facilities, excluding additional cases (n = 46 036). The final cohort included 736 536 patients from 1365 rehabilitation facilities, with International Classification of Diseases, Ninth Revision, Clinical Modification codes that placed them in 1 of the 6 rehabilitation impairment categories.

The mean patient age was 78.0 (SD, 7.3) years. The majority of patients were women (62.5%), and 85.1% were non-Hispanic white. Forty-eight percent were married, and 65.7% were living with someone prior to their acute care hospitalization. Thirty percent had at least 1 CMS rehabilitation tier comorbidity. The mean rehabilitation length of stay was 12.4 (SD, 5.3) days. Sixty-four percent received rehabilitation services in hospital-based rehabilitation units. A majority of patients (57%) received care in nonprofit facilities in urban (89%) settings.

Lower extremity joint replacement was the largest rehabilitation impairment category of patients discharged to the community, representing 25% of our cohort, followed by patients with lower extremity fracture (23%), stroke (21%), debility (12%), neurologic conditions (10%), and brain dysfunction (9%) (Table 1).

Table Graphic Jump LocationTable 1.  Sample Characteristics for Each Rehabilitation Impairment Categorya

The gradual increase in percentage of patients readmitted over time in the total sample (Table 2) is a consequence of changes in the relative contributions of certain rehabilitation impairment groups to the combined sample. Joint replacement, which has the lowest readmission rates, showed the largest decrease in sample size over the 6-year study period, from approximately 33% of the total sample in 2006 to 20% in 2011. Conversely, debility, which has the highest readmission rate, demonstrated the largest increase in sample size, from 8% of the total sample in 2006 to 15% in 2011. Examination of the readmission rates by individual impairment categories indicates they remained relatively stable during the 6-year study period. Based on these stable rates within impairment categories, the analyses reported below are aggregated across the 6-year period.

Table Graphic Jump LocationTable 2.  Percentages of Patients Rehospitalized Within 30 Days of Discharge by Sample Characteristics for Each Rehabilitation Impairment

Table 2 reports the unadjusted hospital readmission status by percentage for the total sample and for each impairment group. The 30-day hospital readmission rate across the 6 impairment groups was 11.8% (95% CI, 11.7%-11.8%) and ranged from 5.8% (95% CI, 5.8%-5.9%) for persons with lower extremity joint replacement to 18.8% (95% CI, 18.8%-18.9%) for persons with debility. The readmission rate was higher for men (13.0% [95% CI, 12.8%-13.1%], vs 11.0% [95% CI, 11.0%-11.1%] for women), non-Hispanic blacks (13.8% [95% CI, 13.5%-14.1%], vs 11.5% [95% CI, 11.5%-11.6%] for whites, 12.5% [95% CI, 12.1%-12.8%] for Hispanics, 11.9% [95% CI, 11.0%-11.2%] for other races/ethnicities), beneficiaries with dual eligibility (15.1% [95% CI, 14.9%-15.4%], vs 11.1% [95% CI, 11.0%-11.2%] for beneficiaries without dual eligibility), for persons with lengths of stay of 15 days or more (14.7% [95% CI< 14.5%-14.8%]) vs those with lengths of stay between 10 and 14 days (11.7% [95% CI, 11.6%-11.9%]) and less than 10 days (9.2% [95% CI, 9.1%-9.3%]), and patients with tier 1 comorbidities (25.6% [95% CI, 24.9%-26.3%], vs 18.9% [95% CI, 18.5%-19.3%] for tier 2, 15.1% [95% CI, 14.9%-15.3% for tier 3, 9.9% [95% CI, 9.9%-10.0%] for no tier comorbidities). Readmission rates were similar for rural (11.1% [95% CI, 10.9%-11.3%]) vs urban (11.8% [95% CI, 11.8%-11.9%]) facilities; freestanding (11.8% [95% CI, 11.7%-11.9%]) vs hospital-based (11.7% [95% CI, 11.6%-11.8%]) facilities; and nonprofit (11.6% [95% CI, 11.5%-11.7%]) vs for-profit (12.0% [95% CI, 11.9%-12.1%]), and government-owned (11.3% [95% CI, 10.9%-11.6%]) facilities. Readmission rates varied by CMG for each impairment category (Table 3). For example, the readmission rates for the 10 stroke CMGs ranged from 9.0% (95% CI, 8.4%-9.7%) to 16.7% (95% CI, 15.9%-17.4%).

Table Graphic Jump LocationTable 3.  Case-Mix Group Distributions and 30-Day Readmission Rates for Each Rehabilitation Impairment Categorya

Several MS-DRGs were common reasons for readmission across all impairment categories. MS-DRGs 689/690 (Kidney and Urinary Tract Infection), 193/194 (Simple Pneumonia and Pleurisy), 291/292 (Heart Failure and Shock), 391/392 (Esophagitis, Gastroenteritis and Miscellaneous Digestive Disorders), and 640/641 (Nutritional and Miscellaneous Metabolic Disorder) all occurred in the top 20 MS-DRGs for all rehabilitation impairment categories. MS-DRG 871/872 (Septicemia Without MV96+ Hours) occurred in the top 10 MS-DRGs for all impairment categories except lower extremity joint replacement. The distribution of the top 25 MS-DRGs by impairment category is included in eAppendix 2 in Supplement.

The Figure displays state readmission rates across the United States and Puerto Rico adjusted for age, sex, race/ethnicity, living situation, rehabilitation impairment category, tier comorbidities, and admission motor and cognitive functioning. The rates ranged from 9.2% (Idaho and Oregon) to 13.6% (Michigan). The tertiles by state show a pattern of lower readmission rates in states in the mid-north and northwest.

Place holder to copy figure label and caption
Figure.
Rate of Hospital Readmission Within 30 Days of Discharge From Rehabilitation

State-specific risk-standardized readmission rates were calculated using hierarchical generalized linear mixed models to account for clustering of patients within states. The models adjusted for 8 patient demographic and clinical variables: age, sex, race/ethnicity, living situation, rehabilitation impairment category, tier comorbidities, and admission motor and cognitive functioning. Final rates were obtained by taking the ratio of predicted to expected readmissions for each state and multiplying by the global unadjusted rate.

Graphic Jump Location

The 30-day readmission rate among patients discharged to the community for the 6 impairment categories was 11.8%. Readmission varied from 5.8% for patients with lower extremity joint replacement to 18.8% for patients with debility.

In 2011, MedPAC reported an all-cause readmission rate of 15.3% across all conditions for acute care hospitals.4 An all-cause readmission rate of 19.2% was reported for heart failure, respiratory tract infection, urinary tract infection, septicemia, and electrolyte imbalance in Medicare skilled nursing patients in 2010.7 The American Hospital Association reported an all-cause readmission rate of 13.7% for long-term care hospitals based on 2007-2009 Medicare data.23 These readmission rates are not directly comparable because of differences in adjustment models and patient groups included but provide a context for our overall findings

Readmission rates for our sample were higher for men and non-Hispanic blacks, for beneficiaries with dual eligibility, for persons with longer lengths of stay, and for individuals with rehabilitation tier–level comorbidities. Readmission rates were similar for rural vs urban facilities, freestanding vs hospital-based facilities, and ownership status. Among the 6 different impairment categories, patients with debility had the highest readmission rate. The number of cases in this impairment group has been increasing, and the reason is not known.24 Research to better understand and manage the care of persons in this impairment category represents an opportunity to reduce readmissions.

Higher motor and cognitive ratings, indicating better patient function, were consistently related to lower readmission rates across all impairment categories. Motor and cognitive status information is not available in Medicare data files for patients in short-term acute care hospitals. Analyses by MedPAC suggest that functional status measures, such as those used in the Medicare rehabilitation assessment data (IRF-PAI), improve the ability to predict use of resources in postacute settings.25,26 This is an important area for future research related to readmission in acute and postacute care.14,27

We examined potential reasons for readmission for each impairment category, following the approach used by Jencks et al.3 We found that the 25 most frequent MS-DRGs constituted approximately 40% to 50% of the readmissions across the impairment categories. Several MS-DRGs were common reasons for readmission. Of particular interest are MS-DRGs representing potential targets for intervention to reduce readmission. For example, kidney and urinary tract infections (MS-DRG 689/690), pneumonia (MS-DRG 193/194), and nutritional and miscellaneous metabolic disorders (MS-DRG 640/641) were in the top 20 MS-DRGs associated with readmission for all 6 impairment categories. Septicemia (MS-DRG 871/872) was not in the top 20 MS-DRGs for lower extremity joint replacement but was in the top 10 MS-DRGs for the 5 remaining rehabilitation impairment categories.

Approximately half of the rehospitalized patients were readmitted within 11 days after discharge. A number of investigators have argued that the 30-day readmission window is arbitrary.28,29 Joynt and colleagues29 suggested that Medicare weight the Hospital Readmission Reduction Program penalties for acute care according to the timing of readmission because earlier readmissions may reflect poor coordination of care or inadequate recognition of postdischarge needs.29 (p1177) Care transition research suggests that programs involving early follow-up have reduced readmission for some patients.2,30 Weighting the Medicare penalties based on earlier readmissions would provide incentive to develop innovative programs including patient and family education, home visits, partnerships with community primary care clinicians, and the use of technology to monitor adherence and medication use.

We found geographic variation in readmission, with rates ranging from 9.2% to 13.6%. The variation is similar to that reported for patients discharged from acute care hospitals,3,4 with lower rates in the mid-northern and northwestern states and higher rates in southern and some midwestern states. The analysis was conducted at the state level and adjusted for rehabilitation impairment categories and sociodemographic factors. These findings need to be confirmed with more refined geographic analyses.

Consistent with research on acute care rehospitalizations, we found slightly higher readmission rates for men than for women and for non-Hispanic blacks than for other race/ethnicity categories.3,5,31 Sex and racial disparities in health care are complex issues, and much has been written regarding the need to reduce disparities.5,3133 This need extends to the emerging research literature on hospital readmissions in both acute and postacute settings.

Medicare is currently examining bundled payment models designed to improve quality and contain costs.10,11,34 The payment options cover different periods and include multiple health care professionals and settings.26 In the context of bundled payment, what happens to patients during postacute care becomes important in the management of resources, quality, cost, and readmissions.35 Recent research has demonstrated that most of the variation in Medicare spending across geographic areas is attributable to postacute care.36 Readmission will likely add to the cost variation. For example, the median cost for a 30-day fixed-length episode for a patient with major joint replacement of the lower extremity is $18 128 without readmission and $29 803 with readmission.23 In describing the role of readmission in bundled payment models, O’Malley states that “Hospitals are not going to achieve meaningful reductions in readmission unless they are partnered with post-acute care.”37

Despite the establishment of readmission in acute care and its introduction into postacute care as an indicator of quality, questions remain regarding its validity.38 The evidence linking readmission to improvement in the care transition process and quality outcomes is inconsistent.39 Questions regarding the validity of readmission as a quality indicator are likely to increase as the accountability for readmission expands to include postacute care settings.11 Although readmission is an imperfect quality indicator, it has the potential to serve as a platform for efforts to improve patient transitions and care continuity associated with bundling and other initiatives proposed by the Affordable Care Act to reduce cost and improve health outcomes.2

Our study has several limitations. These include the reliability, accuracy, and completeness of data collected for billing and administrative functions.40 The majority of our analyses are descriptive and included the “population” of individuals meeting our inclusion criteria. We did not include adjustments for potential mediating factors except for the geographic variation analysis by state. Nor did we examine interactions across rehabilitation impairment categories or other subgroups.

The lack of variables directly measuring socioeconomic status and education in the Medicare claims files limits our ability to document the role of these factors in readmission. We did not attempt to differentiate between planned and unplanned readmissions. It is also important to understand that our decision to include only patients discharged to the community influenced the case mix of the impairment groups analyzed. If all patients receiving inpatient rehabilitation were included in our analyses regardless of discharge destination, the different case mix would influence the results.

Among postacute rehabilitation facilities providing services to Medicare fee-for-service beneficiaries, 30-day readmission rates ranged from 5.8% to 18.8% for selected impairment groups. Further research is needed to understand the reasons for readmission.

Corresponding Author: Kenneth J. Ottenbacher, PhD, University of Texas Medical Branch, 301 University Blvd, Galveston, TX 77555-1137 (kottenba@utmb.edu).

Author Contributions: Drs Ottenbacher and Karmarkar had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Ottenbacher, Kuo, Deutsch.

Acquisition of data: Ottenbacher.

Analysis and interpretation of data: Karmarkar, Graham, Kuo, Deutsch, Reistetter, Snih, Granger.

Drafting of the manuscript: Ottenbacher.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Karmarkar, Kuo, Snih.

Obtained funding: Ottenbacher, Reistetter.

Administrative, technical, and material support: Deutsch.

Study supervision: Ottenbacher.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Kuo reported receiving a grant from the Agency for Healthcare Research and Quality. Dr Reistetter reported serving as a consultant for the National Institute on Disability and Rehabilitation Research and RTI. No other authors reported disclosures.

Funding/Support: This study was supported by the National Institutes of Health, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Center for Medical Rehabilitation Research (R24-HD065702, R01-HD069443, and K01-HD068513), by a Clinical and Translational Science Award from the National Center for Advancing Translational Sciences (UL1RR029876), by the Agency for Healthcare Research and Quality (R24-HS022134), and by the National Institute on Disability and Rehabilitation Research, Department of Education (H133G080163).

Role of the Sponsors: The funding organizations had no role in the design and conduct of the study; the collection, management, analysis, and interpretation of the data; the preparation, review, or approval of the manuscript; or the decision to submit the manuscript for publication.

Patient Protection and Affordable Care Act, PL 111-148. US Dept of Health and Human Services website. http://www.hhs.gov/healthcare/rights/law/index.html. 2010. Accessed January 13, 2013.
Goodman  D, Fisher  E, Chang  C. The Revolving Door: A Report on U.S. Hospital Readmissions. Princeton, NJ: Robert Wood Johnson Foundation; 2013.
Jencks  SF, Williams  MV, Coleman  EA.  Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med. 2009;360(14):1418-1428.
PubMed   |  Link to Article
Medicare Payment Advisory Comission (MedPAC). Report to the Congress: Medicare Payment Policy. Washington, DC: Medicare Payment Advisory Commission; 2013.
Joynt  KE, Orav  EJ, Jha  AK.  Thirty-day readmission rates for Medicare beneficiaries by race and site of care. JAMA. 2011;305(7):675-681.
PubMed   |  Link to Article
Lichtman  JH, Leifheit-Limson  EC, Jones  SB,  et al.  Predictors of hospital readmission after stroke: a systematic review. Stroke. 2010;41(11):2525-2533.
PubMed   |  Link to Article
MedPAC. Report to the Congress: Medicare Payment Policy. Washington, DC: Medicare Payment Advisory Commission; 2012.
Centers for Medicare & Medicaid Services. Specifications for the All-Cause Unplanned Readmission Measure for 30 Days Post Discharge from Inpatient Rehabilitation Faculities. Washington, DC: Centers for Medicare & Medicaid Services; 2013.
Centers for Medicare & Medicaid Services (CMS), HHS.  Medicare program; inpatient rehabilitation facility prospective payment system for federal fiscal year 2014: final rule. Fed Regist. 2013;78(151):47859-47934.
PubMed
DeJong  G.  Bundling acute and postacute payment: from a culture of compliance to a culture of innovation and best practice. Phys Ther. 2010;90(5):658-662.
PubMed   |  Link to Article
Hussey P, Mulcahy A, Schnyer C, Schneider E. Bundled Payment Effects on Health Care Spending and Quality: Closing the Quality Gap: Revisiting the State of the Science. Rockville, MD: Agency for Healthcare Research and Quality; 2012. Evidence Report/Technology Assessment No. 208.
Chan  L, Koepsell  TD, Deyo  RA,  et al.  The effect of Medicare’s payment system for rehabilitation hospitals on length of stay, charges, and total payments. N Engl J Med. 1997;337(14):978-985.
PubMed   |  Link to Article
Inpatient Rehabilitation Facility Patient Assessment Instrument (IRF-PAI) Training Manual. Buffalo, NY: Uniform Data System for Medical Rehabilitation; 2001.
Granger  CV, Deutsch  A, Russell  C, Black  T, Ottenbacher  KJ.  Modifications of the FIM instrument under the inpatient rehabilitation facility prospective payment system. Am J Phys Med Rehabil. 2007;86(11):883-892.
PubMed   |  Link to Article
Ottenbacher  KJ, Hsu  Y, Granger  CV, Fiedler  RC.  The reliability of the functional independence measure: a quantitative review. Arch Phys Med Rehabil. 1996;77(12):1226-1232.
PubMed   |  Link to Article
Stineman  MG, Shea  JA, Jette  A,  et al.  The Functional Independence Measure: tests of scaling assumptions, structure, and reliability across 20 diverse impairment categories. Arch Phys Med Rehabil. 1996;77(11):1101-1108.
PubMed   |  Link to Article
Carter  GM, Relles  DA, Ridgeway  GK, Rimes  CM.  Measuring function for Medicare inpatient rehabilitation payment. Health Care Financ Rev. 2003;24(3):25-44.
PubMed
Federal Register: Quality Report Program for IRFs. Washington, DC: Department of Health and Human Services; 2011. Report No. 76.
Carter  G, Totten  M. Preliminary Analyses for Refinement of the Tier Comorbidities in the Inpatient Rehabilitation Facility Prospective Payment System.TR-201-CMS ed. Santa Monica, CA: Rand Corp; 2005.
McNutt  R, Johnson  TJ, Odwazny  R,  et al.  Change in MS-DRG assignment and hospital reimbursement as a result of Centers for Medicare & Medicaid changes in payment for hospital-acquired conditions: is it coding or quality? Qual Manag Health Care. 2010;19(1):17-24.
PubMed   |  Link to Article
Blyth  C, Still  H.  Binomial confidence intervals. J Am Stat Assoc. 1983;78:108-116.
Link to Article
Krumholz  HM, Merrill  AR, Schone  EM,  et al.  Patterns of hospital performance in acute myocardial infarction and heart failure 30-day mortality and readmission. Circ Cardiovasc Qual Outcomes. 2009;2(5):407-413.
PubMed   |  Link to Article
American Hospital Association. Issue Brief: Moving Towards Bundled Payment. Washington, DC: American Hospital Association; 2013.
Galloway  RV, Granger  CV, Karmarkar  AM,  et al.  The Uniform Data System for Medical Rehabilitation: report of patients with debility discharged from inpatient rehabilitation programs in 2000-2010. Am J Phys Med Rehabil. 2013;92(1):14-27.
PubMed   |  Link to Article
3M Health Information Systems. The Impact of Disability Measures on Expected Medicare Payments and Expected Provider Charges for Event-Based Episodes That Include Post-acute Care. Salt Lake City, UT: 3M Health Information Systems; 2013.
MedPAC. Report to the Congress: Medicare and the Health Care Delivery System. Washington, DC: Medicare Payment Advisory Commission; 2013.
Deutsch A, Mallinson T, Gage B. Analysis of Crosscutting Medicare Functional Status Quality Metrics Using Continuity and Assessment Record and Evaluation (CARE) Item Set. Washington, DC: RTI International; 2012. Report 0212050.
Chen  LM, Jha  AK, Guterman  S, Ridgway  AB, Orav  EJ, Epstein  AM.  Hospital cost of care, quality of care, and readmission rates: penny wise and pound foolish? Arch Intern Med. 2010;170(4):340-346.
PubMed   |  Link to Article
Joynt  KE, Jha  AK.  A path forward on Medicare readmissions. N Engl J Med. 2013;368(13):1175-1177.
PubMed   |  Link to Article
Hansen  LO, Young  RS, Hinami  K, Leung  A, Williams  MV.  Interventions to reduce 30-day rehospitalization: a systematic review. Ann Intern Med. 2011;155(8):520-528.
PubMed   |  Link to Article
Li  Y, Glance  LG, Yin  J, Mukamel  DB.  Racial disparities in rehospitalization among Medicare patients in skilled nursing facilities. Am J Public Health. 2011;101(5):875-882.
PubMed   |  Link to Article
Fiscella  K, Franks  P, Gold  MR, Clancy  CM.  Inequality in quality: addressing socioeconomic, racial, and ethnic disparities in health care. JAMA. 2000;283(19):2579-2584.
PubMed   |  Link to Article
Smedley  BD, Stith  AY, Nelson  AR. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care (Institute of Medicine). Washington, DC: National Academies Press; 2003.
Mechanic  R, Tompkins  C.  Lessons learned preparing for Medicare bundled payments. N Engl J Med. 2012;367(20):1873-1875.
PubMed   |  Link to Article
Chandra  A, Dalton  MA, Holmes  J.  Large increases in spending on postacute care in Medicare point to the potential for cost savings in these settings. Health Aff (Millwood). 2013;32(5):864-872.
PubMed   |  Link to Article
Newhouse  JP, Garber  AM.  Geographic variation in Medicare services. N Engl J Med. 2013;368(16):1465-1468.
PubMed   |  Link to Article
American Hospital Association. Trendwatch: Maximizing the Value of Post-acute Care. Washington, DC: American Hospital Association; 2010.
Axon  RN, Williams  MV.  Hospital readmission as an accountability measure. JAMA. 2011;305(5):504-505.
PubMed   |  Link to Article
Brock  J, Mitchell  J, Irby  K,  et al; Care Transitions Project Team.  Association between quality improvement for care transitions in communities and rehospitalizations among Medicare beneficiaries. JAMA. 2013;309(4):381-391.
PubMed   |  Link to Article
van Walraven  C, Austin  P.  Administrative database research has unique characteristics that can risk biased results. J Clin Epidemiol. 2012;65(2):126-131.
PubMed   |  Link to Article

Figures

Place holder to copy figure label and caption
Figure.
Rate of Hospital Readmission Within 30 Days of Discharge From Rehabilitation

State-specific risk-standardized readmission rates were calculated using hierarchical generalized linear mixed models to account for clustering of patients within states. The models adjusted for 8 patient demographic and clinical variables: age, sex, race/ethnicity, living situation, rehabilitation impairment category, tier comorbidities, and admission motor and cognitive functioning. Final rates were obtained by taking the ratio of predicted to expected readmissions for each state and multiplying by the global unadjusted rate.

Graphic Jump Location

Tables

Table Graphic Jump LocationTable 1.  Sample Characteristics for Each Rehabilitation Impairment Categorya
Table Graphic Jump LocationTable 2.  Percentages of Patients Rehospitalized Within 30 Days of Discharge by Sample Characteristics for Each Rehabilitation Impairment
Table Graphic Jump LocationTable 3.  Case-Mix Group Distributions and 30-Day Readmission Rates for Each Rehabilitation Impairment Categorya

References

Patient Protection and Affordable Care Act, PL 111-148. US Dept of Health and Human Services website. http://www.hhs.gov/healthcare/rights/law/index.html. 2010. Accessed January 13, 2013.
Goodman  D, Fisher  E, Chang  C. The Revolving Door: A Report on U.S. Hospital Readmissions. Princeton, NJ: Robert Wood Johnson Foundation; 2013.
Jencks  SF, Williams  MV, Coleman  EA.  Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med. 2009;360(14):1418-1428.
PubMed   |  Link to Article
Medicare Payment Advisory Comission (MedPAC). Report to the Congress: Medicare Payment Policy. Washington, DC: Medicare Payment Advisory Commission; 2013.
Joynt  KE, Orav  EJ, Jha  AK.  Thirty-day readmission rates for Medicare beneficiaries by race and site of care. JAMA. 2011;305(7):675-681.
PubMed   |  Link to Article
Lichtman  JH, Leifheit-Limson  EC, Jones  SB,  et al.  Predictors of hospital readmission after stroke: a systematic review. Stroke. 2010;41(11):2525-2533.
PubMed   |  Link to Article
MedPAC. Report to the Congress: Medicare Payment Policy. Washington, DC: Medicare Payment Advisory Commission; 2012.
Centers for Medicare & Medicaid Services. Specifications for the All-Cause Unplanned Readmission Measure for 30 Days Post Discharge from Inpatient Rehabilitation Faculities. Washington, DC: Centers for Medicare & Medicaid Services; 2013.
Centers for Medicare & Medicaid Services (CMS), HHS.  Medicare program; inpatient rehabilitation facility prospective payment system for federal fiscal year 2014: final rule. Fed Regist. 2013;78(151):47859-47934.
PubMed
DeJong  G.  Bundling acute and postacute payment: from a culture of compliance to a culture of innovation and best practice. Phys Ther. 2010;90(5):658-662.
PubMed   |  Link to Article
Hussey P, Mulcahy A, Schnyer C, Schneider E. Bundled Payment Effects on Health Care Spending and Quality: Closing the Quality Gap: Revisiting the State of the Science. Rockville, MD: Agency for Healthcare Research and Quality; 2012. Evidence Report/Technology Assessment No. 208.
Chan  L, Koepsell  TD, Deyo  RA,  et al.  The effect of Medicare’s payment system for rehabilitation hospitals on length of stay, charges, and total payments. N Engl J Med. 1997;337(14):978-985.
PubMed   |  Link to Article
Inpatient Rehabilitation Facility Patient Assessment Instrument (IRF-PAI) Training Manual. Buffalo, NY: Uniform Data System for Medical Rehabilitation; 2001.
Granger  CV, Deutsch  A, Russell  C, Black  T, Ottenbacher  KJ.  Modifications of the FIM instrument under the inpatient rehabilitation facility prospective payment system. Am J Phys Med Rehabil. 2007;86(11):883-892.
PubMed   |  Link to Article
Ottenbacher  KJ, Hsu  Y, Granger  CV, Fiedler  RC.  The reliability of the functional independence measure: a quantitative review. Arch Phys Med Rehabil. 1996;77(12):1226-1232.
PubMed   |  Link to Article
Stineman  MG, Shea  JA, Jette  A,  et al.  The Functional Independence Measure: tests of scaling assumptions, structure, and reliability across 20 diverse impairment categories. Arch Phys Med Rehabil. 1996;77(11):1101-1108.
PubMed   |  Link to Article
Carter  GM, Relles  DA, Ridgeway  GK, Rimes  CM.  Measuring function for Medicare inpatient rehabilitation payment. Health Care Financ Rev. 2003;24(3):25-44.
PubMed
Federal Register: Quality Report Program for IRFs. Washington, DC: Department of Health and Human Services; 2011. Report No. 76.
Carter  G, Totten  M. Preliminary Analyses for Refinement of the Tier Comorbidities in the Inpatient Rehabilitation Facility Prospective Payment System.TR-201-CMS ed. Santa Monica, CA: Rand Corp; 2005.
McNutt  R, Johnson  TJ, Odwazny  R,  et al.  Change in MS-DRG assignment and hospital reimbursement as a result of Centers for Medicare & Medicaid changes in payment for hospital-acquired conditions: is it coding or quality? Qual Manag Health Care. 2010;19(1):17-24.
PubMed   |  Link to Article
Blyth  C, Still  H.  Binomial confidence intervals. J Am Stat Assoc. 1983;78:108-116.
Link to Article
Krumholz  HM, Merrill  AR, Schone  EM,  et al.  Patterns of hospital performance in acute myocardial infarction and heart failure 30-day mortality and readmission. Circ Cardiovasc Qual Outcomes. 2009;2(5):407-413.
PubMed   |  Link to Article
American Hospital Association. Issue Brief: Moving Towards Bundled Payment. Washington, DC: American Hospital Association; 2013.
Galloway  RV, Granger  CV, Karmarkar  AM,  et al.  The Uniform Data System for Medical Rehabilitation: report of patients with debility discharged from inpatient rehabilitation programs in 2000-2010. Am J Phys Med Rehabil. 2013;92(1):14-27.
PubMed   |  Link to Article
3M Health Information Systems. The Impact of Disability Measures on Expected Medicare Payments and Expected Provider Charges for Event-Based Episodes That Include Post-acute Care. Salt Lake City, UT: 3M Health Information Systems; 2013.
MedPAC. Report to the Congress: Medicare and the Health Care Delivery System. Washington, DC: Medicare Payment Advisory Commission; 2013.
Deutsch A, Mallinson T, Gage B. Analysis of Crosscutting Medicare Functional Status Quality Metrics Using Continuity and Assessment Record and Evaluation (CARE) Item Set. Washington, DC: RTI International; 2012. Report 0212050.
Chen  LM, Jha  AK, Guterman  S, Ridgway  AB, Orav  EJ, Epstein  AM.  Hospital cost of care, quality of care, and readmission rates: penny wise and pound foolish? Arch Intern Med. 2010;170(4):340-346.
PubMed   |  Link to Article
Joynt  KE, Jha  AK.  A path forward on Medicare readmissions. N Engl J Med. 2013;368(13):1175-1177.
PubMed   |  Link to Article
Hansen  LO, Young  RS, Hinami  K, Leung  A, Williams  MV.  Interventions to reduce 30-day rehospitalization: a systematic review. Ann Intern Med. 2011;155(8):520-528.
PubMed   |  Link to Article
Li  Y, Glance  LG, Yin  J, Mukamel  DB.  Racial disparities in rehospitalization among Medicare patients in skilled nursing facilities. Am J Public Health. 2011;101(5):875-882.
PubMed   |  Link to Article
Fiscella  K, Franks  P, Gold  MR, Clancy  CM.  Inequality in quality: addressing socioeconomic, racial, and ethnic disparities in health care. JAMA. 2000;283(19):2579-2584.
PubMed   |  Link to Article
Smedley  BD, Stith  AY, Nelson  AR. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care (Institute of Medicine). Washington, DC: National Academies Press; 2003.
Mechanic  R, Tompkins  C.  Lessons learned preparing for Medicare bundled payments. N Engl J Med. 2012;367(20):1873-1875.
PubMed   |  Link to Article
Chandra  A, Dalton  MA, Holmes  J.  Large increases in spending on postacute care in Medicare point to the potential for cost savings in these settings. Health Aff (Millwood). 2013;32(5):864-872.
PubMed   |  Link to Article
Newhouse  JP, Garber  AM.  Geographic variation in Medicare services. N Engl J Med. 2013;368(16):1465-1468.
PubMed   |  Link to Article
American Hospital Association. Trendwatch: Maximizing the Value of Post-acute Care. Washington, DC: American Hospital Association; 2010.
Axon  RN, Williams  MV.  Hospital readmission as an accountability measure. JAMA. 2011;305(5):504-505.
PubMed   |  Link to Article
Brock  J, Mitchell  J, Irby  K,  et al; Care Transitions Project Team.  Association between quality improvement for care transitions in communities and rehospitalizations among Medicare beneficiaries. JAMA. 2013;309(4):381-391.
PubMed   |  Link to Article
van Walraven  C, Austin  P.  Administrative database research has unique characteristics that can risk biased results. J Clin Epidemiol. 2012;65(2):126-131.
PubMed   |  Link to Article
CME


You need to register in order to view this quiz.

Multimedia

Supplement.

eAppendix 1. Number of Patients and ICD-9-CM Codes Associated With the Rehabilitation Impairment Categories

eAppendix 2. Top 25 MS DRGs Associated With Readmission After Discharge From Inpatient Rehabilitation Facilities for Six Rehabilitation Impairment Categories

Supplemental Content

Some tools below are only available to our subscribers or users with an online account.

Web of Science® Times Cited: 1

Related Content

Customize your page view by dragging & repositioning the boxes below.

Articles Related By Topic
Related Collections
PubMed Articles