Joint replacement Outcomes in IRFs and Nursing Treatment Sites

1

• Joint replacement Outcomesin IRFs and Nursing
  Treatment Sites

Practice-based Evidence for Post-acute Care
Policy The Case of Joint Replacement
Rehabilitation
American Congress of Rehabilitation
Medicine Washington, DC October 4, 2007
www.jointsstudy.net
2
Study Introduction and OverviewGrowth of Joint
ReplacementsFraming the IssuesStudy
ApproachStudy Facilities

• Gerben DeJong, PhD
• Center for Post-acute Studies
• National Rehabilitation Hospital
• Washington, DC

3
Limitations

• This presentation presents some of the first of
  many findings from the JOINTS Study. Some of the
  findings are nuanced and need to be interpreted
  in light of the overall evidence presented. The
  study also has several limitations, some by
  design and others by default. These limitations
  are outlined later in this presentation and
  should be taken into account when making practice
  and policy inferences.

4

• JOINTS I
• Content and processes of care
• Outcomes at discharge from rehabilitation
• Collaborators
• Center for Post-acute Studies, NRH
• Institute for Clinical Outcomes Research

• JOINTS II
• Outcomes from date of admission to 6 mos.
• Health care expenditures from date of admission
  to 6 mos.
• Collaborators
• Center for Post-acute Studies, NRH
• Institute for Clinical Outcomes Research
• IT HealthTrack
• Lewin Group

5
Study Investigators

• Gerben DeJong, PhD, PI
• NRH Center for Post-acute Studies
• Susan Horn, PhD, Co-PI
• Institute for Clinical Outcomes Research

6
Project Director Julie Gassaway, RN,
MSInstitute for Clinical Outcomes Research

• Project Manager
• Jean Hsieh, OT, PhD
• NRH Center for Post-acute Studies

7
Research Analysts

• Randall Smout, MS
• Institute for Clinical Outcomes Research
• Wengiang Tian, MD, PhD
• NRH Center for Post-acute Studies
• Roberta James, MStat
• Institute for Clinical Outcomes Research
• Koen Putman, PT, PhD
• Vrije Universiteit Brussel
• NRH Center for Post-acute Studies

8
Research Facilitators

• Mary Foley, CRN
• National Rehabilitation Hospital
• Elizabeth Newman, OTD
• National Rehabilitation Hospital
• Cathy Ellis, PT
• National Rehabilitation Hospital

9
Sponsors

• HealthSouth Corporation
• ARA Research Institute
• (AMRPA)
• Brooks Health
• American Hospital Association
• Federation of American Hospitals
• National Rehabilitation Hospital
• and many other individual organizations

10
JOINTS StudyJoint replacement Outcomesin IRFs
and Nursing Treatment Sites

• Partner
• Facilities
• 11 SNFs
• 11 IRFs
• from across the US

11
(No Transcript)
12
Participating Facilities

• SNF
• Brentwood Subacute Healthcare
• California Special Care Center
• Cedars at the JCA
• Crosslands Rehabilitation
• Frazier Rehab Institute
• Greenbriar Terrace Healthcare
• Harrison Health and Rehab
• Jewish Home of Rochester
• Laurel Baye of Greenville
• Mayo Clinic
• TCU Reading

• IRF
• Baptist Health Rehab
• Brooks Rehab Hospital
• Casa Colina
• HealthSouth Plano
• HealthSouth Scottsdale
• JFK Johnson Rehab
• Mayo Clinic
• Natl Rehab Hospital
• Rehab Institute of Chicago
• Rehab Hospital of CT
• Spalding Rehab Center

13
Policy Advisory Panel

• Trade associations
• AMRPA
• AHA
• AFH
• AASHA
• Professional associations
• AAPMR
• ACRM
• APTA
• AOTA
• ARN
• Consumer organizations
• Arthritis Foundation

• Government
• NCMRR
• AHRQ
• NIDRR (observer)
• CMS (observer)
• Accreditation
• JCAHO
• CARF
• Other
• Rand
• SeniorMetrix
• UDS

14
Total Knee and Total Hip Replacements1993-2005
Source AHRQ HCUP, NIS (1993-2005)
15
Total Knee and Total Hip ReplacementAcute Care
Length of Stay,1993-2005
Source AHRQ HCUP, NIS (1993-2005)
16
Total Knee and Total Hip Replacement Discharges
to Post-acute Settings, 1993-2005
Source AHRQ HCUP, NIS (1993-2005)
17
Where?

• ?Skilled nursing facility (SNF)?
• ?Inpatient rehabilitation facility (IRF)?
• Home health?
• Outpatient care?
• Inadequate evidence to select best venue for
  specific types of patients

18
Framing the Issue

• Comparing 2 black boxes.
• Never characterized differences in care between
  SNFs and IRFs.
• Contrasts made number of hours of treatment.
• Never fully characterized the type, timing,
  intensity, frequency, and duration of treatment
  in IRF and SNF.

19
Framing the Issue

• Without knowing what is in these black boxes,
  prudent purchasers, both government and health
  plans, cannot fully know what they are
  purchasing.

20
Framing the Issue

• In many ways, SNF vs. IRF is the wrong question.
• Instead, we need to ask Which patients do
  better in a SNF and which do better in an IRF?
• This requires adequate characterization of
• (1) the patient
• (2) the interventions associated with each
  setting of care.

21
Framing the Issue

• Today we will characterize what happens in the
  proverbial black box of SNF and IRF care.
• We will look at how patient characteristics and
  facility-level features are associated with
  outcomes.
• We will not discuss how rehabilitation activities
  and interventions are associated with outcomes
  and will not present a best practices analysis.

22
What You Will Hear Today

• Study approach and methods
• Study design
• Study outcome measures
• Study instruments
• Data collection protocols
• Facility selection and characteristics
• Patient selection and characteristics

23
What You Will Hear Today

• Characterizing the black box of joint
  replacement rehabilitation
• Therapy activities in SNFs and IRFs
• Therapy interventions in SNFs and IRFs
• Outcomes at discharge
• Patient and facility characteristics associated
  with outcomes, i.e., predictors of outcomes

24
The Larger JOINTS I Study Group

• 22 facilities
• ?11SNFs
• ?11 IRFs
• 2,384 patients with joint replacement
• Patient selection criteria
• 21 yrs or older
• Hip or knee replacement of any type for any
  reason
• Admitted from any source

25
JOINTS I Study Groupfor this presentation

• 19 facilities
• 11?8 SNFs
• -1 SNF Hybrid facility (N 266)
• -1 SNF excluded because of data quality issues
• -1 SNF with very small number (N7) and timing
  issues
• 11 IRFs

26
Hybrid Facility (HBF)

• Discovered one SNF was neither a conventional SNF
  nor an IRF
• Designed to be a different model of facility
• Hospital-based shared therapy staff with sister
  IRF
• Very short acute ALOS3.5 days
• Features of both SNF and IRF
• Similar to an IRF
• LOS (8.8 days)
• Admission severity
• Physician and pharmacy coverage
• Similar to a SNF
• Therapy intensity
• Removing this hybrid facility brought greater
  clarity to distinctions between SNFs and IRFs

27
JOINTS I Study Groupfor this presentation

• 1,892 patients with joint replacement (exclusive
  of those with hip fracture)
• 549 patients from SNFs
• 1,343 patients from IRFs

28
JOINTS I Facility Selection Factors

• Target number of pts 200 pts/facility
• To minimize transaction costs working with many
  facilities
• At least 2 facilities from each Census region
• Mix of freestanding and acute-hospital based
  facilities
• Mix of for-profit and non-profit facilities
• Payer mix

29
JOINTS I Facility Selection Factors

• Not a national probability sample
• Most important geographic diversity
• Had to relax criteria especially for SNFs
• Patient volumes were lower than expected
• More difficult to recruit
• Risk Attract better breed facilities of both
  types.
• A coalition of the willing

30
PTOT Intensity by Length of Stay
31
Study Approach

• Practice-based Evidence (PBE)
• Type of prospective observational cohort study
• Takes advantage of natural variation in
  rehabilitation practice
• Large numbers allow examination of differences
  among types of patients and patient subgroups

32
Study Facilities
33
Study Facilities
National IRF information from CMS POS file.
34
Study Facilities
35
Study Facilities
36
Study Facilities
37
Facility Joint Replacement Volume
38
Study Methods and Findings 1Special
ChallengesData CollectionPatient
CharacteristicsFunctional StatusLength of
StayTypes and Amounts of Therapy

• Julie Gassaway, RN, MPH
• Institute for Clinical Outcomes Research
• Salt Lake City, UT

39
Challenges to Open the Box

• Double challenge to provide evidence for the
  effectiveness of rehabilitation requires
• Identify individual elements of the
  rehabilitation process
• Demonstrate the link between specific
  interventions and positive outcomes

40
Outcomes Challenge

• Primary Outcomes
• Functional status at discharge
• Improvement in functional status from admission
  to discharge
• Discharge destination
• Joint replacement is an elective procedure and
  patients typically go home.

41
Outcomes Challenge

• Functional status
• IRFs use IRF-PAI FIM
• SNFs use MDS
• FIM and MDS use different time points
• FIM admission (3 days) and discharge (3 days)
• MDS 5th and 14th day
• Lack of crosswalk between FIM and MDS

42
Outcomes Challenge/Solution

• Solution
• Use FIM at admission and discharge in all
  facilities
• Needed to train SNF personnel to use FIM
• Enlisted the support of IT HealthTrack
• All users in SNFs and IRFs needed to score 100
  to assure reliability before enrolling patients
• Added another 2-month enrollment lag at SNFs

43
Detailed Data Needed

• Patient characteristic variables
• Severity of medical conditions, demographics, and
  psychosocial background
• Process variables
• All rehabilitation interventions (therapy,
  medical interventions, patient education)
• Outcome variables
• At Discharge functional independence, discharge
  to home, medical complications,
• At 6-Month Follow-up JOINTS II

44
JOINTS Signature Features

• Facility and clinical buy-in through use of its
  trans-disciplinary Clinical Practice Team
• Detailed characterization of the patient
  including the Comprehensive Severity Index (CSI)
• Detailed characterization of the care process
  including it point-of-care documentation (POC)
• High level of transparency for all stakeholders.

45
JOINTS Detailed Data Sources

• The medical record
• Patient characteristics, medical diagnoses,
  process steps, outcomes (including FIM)
• Point-of-care documentation
• Rehabilitation process details not in the record
• Six-Month Follow-up interviews
• JOINTS II Outcome data and post-discharge
  process

46
Data Collection Methods

• Point-of-care documentation
• Who?
• PT
• OT
• Nursing (varying compliance)
• Development
• Training
• 2-day training session

47
Point-of-Care Development

• Began at a face-to-face meeting with facility
  representatives, who identified clinical leads
  for each discipline (PT, OT, Nursing)
• Each discipline worked in weekly teleconferences
  for about 9 months
• Each discipline identified what they thought was
  critical to positive outcomes after joint
  replacement
• Discovered how things were done differently at
  different sites

48
PT Process Activitiesand Time Spent on Each

• Assessment/Eval
• Medical Monitoring
• Pain Management
• Edema Control
• Positioning
• Exercising
• Bed Mobility

• Transfers
• WC Mobility/Mgmt
• Pre-Gait/Standing
• Gait
• Community Mobility
• Home Program
• Team Communication

49
Interventions Associated with PT Activities

• Exercise Type
• Therapeutic Exercise
• Manual Therapy/Stretching
• Neuromuscular Reeducation
• Breathing
• Balance Training
• Aquatic Exercise
• Mobility Training
• Functional Endurance
• Bed to/from Chair transfer
• Shower/bath transfer
• Toilet transfer
• Car transfer
• Even surfaces
• Uneven Surfaces
• Stairs
• Curbs

• Education
• Orthopedic/Wt bearing precautions
• Family Training
• Body Mechanics
• Safety/Cueing
• Energy Conservation
• Orthotics
• Abductor Wedge/ Hip Abductor
• AFO
• Knee Immobilizer
• Modalities
• Cold/Hot Packs
• Electrotherapeutic modalities
• Ultrasound
• Non-Surgical Limb Treated

50
Nursing Process Variables

• Number of assists with transfers each shift
• Number of assists with ambulation each shift
• Orthotic devices used for transfer/ambulation
• Education/Reinforce Frequency each shift
• Assistive devices - Disease management
• Transfer techniques - ADLs
• Mobility - Nutrition
• Precautions - Bowel
• Safety - Bladder
• Pain management - Skin/Wound
• Medications

51
Patient Enrollment

• Patient Enrollment Period (JOINTS I)
• Feb 2006-Feb 2007
• Patients enrolled consecutively
• 21 yrs or older
• Hip or knee replacement of any type for any
  reason
• Admitted from any source
• IRFs, higher volumes, completed Oct 2006
• SNFs, lower volumes, later start, completed Feb
  2007

52
Study Group
53
Study Group Demographics Knee Replacement
54
Study Group Demographics Hip Replacement
55
Study Group Payer Mix Knee Replacement
p lt.001
56
Study Group Payer Mix Hip Replacement
p lt.001
57
Study Group CharacteristicsKnee Replacement
58
Study Group CharacteristicsHip Replacement
59
Study Group Health Status Knee Replacement
60
Study Group Health Status Knee Replacement
61
Study Group Health StatusHip Replacement
62
Study Group Health Status Hip Replacement
63
Acute and Post-acute CareKnee Replacement
64
Acute and Post-acute CareHip Replacement
65
IRF Patient Characteristics JOINTS vs. National
Data
National information comes from eRehab data,
calendar year 2006.
66
IRF Patient Characteristics JOINTS vs. National
Data
National information comes from eRehab data,
calendar year 2006.
67
Study Group Functional Status
Knee plt.001 Hip plt.001
Knee plt.001 Hip plt.001
68
Study Group Functional Status
Knee plt.001 Hip plt.001
69
Study Group Length of Stay
Knee plt.001 Hip plt.001
70
Definition of Joint Replacement CMGs
71
Length of Stay by CMGs
For CMGs 801-802 803-804 Knee plt.001 Hip
plt.001 For CMGs 805-806 Knee p.002 Hip plt.001
72
Study Group CMGsKnee Replacement
73
Study Group CMGsHip Replacement
74
Physical TherapyKnee Replacement
75
Physical TherapyHip Replacement
76
Occupational TherapyKnee Replacement
77
Occupational TherapyHip Replacement
78
PT OT Combined Knee Replacement
79
PT OT Combined Hip Replacement
80
PTOT Intensity by Length of Stay
81
Hours and Proportion of PT ActivitiesKnee
Replacement
82
Hours and Proportion of PT ActivitiesHip
Replacement
83
Hours and Proportion of OT ActivitiesKnee
Replacement
84
Hours and Proportion of OT Activities Hip
Replacement
85
Most Common PT/OT Interventions
86
What Makes this PBE Study Different?

• Powerful enough to answer big questions
• Complex enough to require almost 100 clinicians
  and researchers to plan the study
• Detailed enough to require comprehensive clinical
  participation

87
Study Methods and Findings 2Categorization of
Facility VolumesPatient Characteristics and
Outcomesby Facility VolumeMultivariate Analysis

• Susan D. Horn, PhD
• Institute for Clinical Outcomes Research
• Salt Lake City, UT

88
Facility Joint Replacement Volume
89
Demographic Differences by Facility Volume Knee
Replacement
Refers to those patients who lived alone prior
to surgery.
90
Demographic Differences by Facility VolumeHip
Replacement
Refers to those patients who lived alone prior
to surgery.
91
Admission CSI by Facility Volume Knee
Replacement
plt.001
92
Length of Stay by Facility Volume Knee
Replacement
plt.001
93
Admission CSI by Facility VolumeHip Replacement
plt.001
94
Length of Stay by Facility Volume Hip Replacement
plt.001
95
Discharge DestinationsKnee Replacement
p0.928
96
Discharge DestinationsHip Replacement
plt.001
97
Discharge Motor FIM by Facility Volume Knee
Replacement
plt.001
98
Change Motor FIM by Facility Volume Knee
Replacement
plt.001
99
Discharge Motor FIM by Facility Volume Hip
Replacement
plt.001
100
Change Motor FIM by Facility Volume Hip
Replacement
plt.001
101
JOINTS I Outcomes Knee Replacement
102
JOINTS I Outcomes Hip Replacement
103
Multivariate Analysis in Predicting Discharge
Motor FIM Score Knee Replacement
104
(No Transcript)
105
Discharge Motor FIM by Facility Vol.Knee
Replacement
1.38
-2.14
-3.59
106
Multivariate Analysis in Predicting Discharge
Motor FIM Score Hip Replacement
107
(No Transcript)
108
Discharge Motor FIM by Facility Vol. Hip
Replacement
2.38
-3.92
-6.04
109
JOINTS I SummarySummary of Key FindingsStudy
LimitationsNext Steps

• Gerben DeJong, PhD
• Center for Post-acute Studies
• National Rehabilitation Hospital
• Washington, DC

110
Summary of Key Findings to DateAcute Care

• A doubling of total joint replacements from from
  381,808 in 1993 to 774,764 in 2005
• A decline in acute care ALOS during same period,
  from about 7½-8 days to lt4 days
• 2.9 million joint replacements anticipated in
  2030
• About 2/3 discharged to some post-acute setting

111
Summary of Key Findings to DatePatient
Characteristics

• A typical knee replacement rehabilitation patient
  is an older (71.2 yrs) woman who is overweight
  (BMI 33) and participates in Medicare.
  (Similar for individuals with hip replacement.)
• SNF JR patients tended to be older and more
  female and were more likely to live alone and
  participate in Medicare.
• IRF JR patients present more medical acuity, more
  likely to have a bilateral replacement, and have
  higher admission pain scores.

112
Summary of Key Findings to DatePatient
Characteristics

• SNF and IRF patients were not materially
  different in their average rates of diabetes,
  hypertension, and ischemic heart disease.
• SNF and IRF patients had similar rates of
  revisions and acute LOSs

113
Summary of Key Findings to DatePractice Patterns

• Study SNFs exhibited a greater diversity of
  practice than IRFs with respect to LOS and
  intensity of daily therapy.
• The greater diversity of SNF practice patterns
  makes it more difficult to generalize to all SNFs
  in the same way one can generalize to other IRFs.

114
Summary of Key Findings to DatePractice Patterns

• Both SNFs and IRFs emphasize similar therapy
  activities, e.g., exercise and gait training, and
  interventions, albeit with different levels of
  intensity and duration.
• Over the entire length of stay, SNF and IRF
  patients receive about the same total amounts of
  PT IRF patients receive somewhat more OT over
  the course of their rehabilitation stay.

115
Summary of Key Findings to DateFunctional Status
and Discharge Outcomes

• SNF patients come in with higher functional
  scores and leave with higher functional scores.
• IRF patients, however, experience a higher net
  motor functional gain and achieve these gains in
  a shorter period of time. IRFs exhibited "higher
  LOS efficiency.
• Facility volume appears to be associated with
  discharge outcome.

116
Summary of Key Findings to DateFunctional Status
and Discharge Outcomes

• Controlling for patient differences, small-
  volume SNFs produced lower FIM gains for both
  knee replacement patients (-3.6 FIM pts) and hip
  replacement patients (-6.0 FIM pts).
• There were no small-volume IRFs in the study to
  determine if the same finding would apply to
  small-volume IRFs.
• Controlling for patient differences,
  medium-volume IRFs produce higher FIM gains for
  knee replacement patients (1.38 FIM pts) and hip
  replacement patients (2.38 FIM pts).

117
Study Limitations

• Selection issues. All facilities participated
  voluntarily. Likely to attract better breed
  facilities of both types.
• Small IRFs, especially units, were
  underrepresented in the study.
• Given the diversity/variability of SNF practice
  patterns, the study would have benefited from
  participation of more SNFs relative to IRFs.

118
Study Limitations

• While IRFs exhibit higher LOS efficiency, the
  analysis payment efficiency is still underway.
• These findings do not include outcomes at
  follow-up (6-months)
• Complications
• General health
• Functional status
• Rehospitalizations
• Rehabilitation services used
• Quality of life
• Community integration, employment

119
Some Next Steps

• Evaluate 6-month outcomes
• Examine payment and payment efficiency as well as
  LOS efficiency.
• Identify subgroups that do distinctively better
  in one setting or the other
• Examine associations between interventions and
  outcomes to determine best practices.

120
Study Contact Information

• Gerben DeJong, PhD
• Principal Investigator
• 202-877-1960
• Gerben.DeJong_at_MedStar.net
• Michael Brown
• Program Coordinator
• 202-877-1345
• Michael.Brown_at_MedStar.net
• Donal Lauderdale, MEE
• Dissemination Utilization Manager
• 202-877-1425
• Donal.Lauderdale_at_MedStar.net
• National Rehabilitation Hospital
• 102 Irving Street, NW
• Washington, DC 20010