The impact of severe obesity on post-acute rehabilitation efficiency, length of stay, and hospital costs.

Abstract

Background and Objective. The purpose of this retrospective observational study was to examine the influence of severe obesity on length of stay (LOS), rehabilitation efficiency, and hospital costs post-acute rehabilitation in a population-based, tertiary care, publicly-funded regional rehabilitation center. Participants. 42 severely obese subjects (mean age 53 y; mean BMI 50.9 kg/m(2)) and 42 nonobese controls (mean age 59 y; mean BMI 23.0 kg/m(2)) matched by sex and admitting diagnosis. Main Outcome Measures. Total LOS, rehab LOS, waiting for transfer LOS, Fuctional Independence Measure (FIM) efficiency, and hospital costs. Results. Compared to controls, severely obese subjects experienced longer total LOS (98.4 vs. 37.4 days; P = 0.03), rehabilitation LOS (55.8 vs. 37.4 days; P = 0.04), and waiting for transfer LOS (42.6 vs. 0 days; P = 0.006); increased hospital costs ($115,822 vs. $43,969; P = 0.03); and similar FIM efficiency (0.58 vs. 0.67; P = 0.27). Severe obesity was an independent predictor of total LOS (beta-coefficient 0.51; P = 0.03), rehab LOS (0.46; P = 0.02) but not FIM efficiency (-0.63; P = 0.06). Conclusion. Severe obesity adversely affects rehabilitation LOS and expenditures. Targeted interventions in severely obese individuals to optimize post-acute rehabilitation care delivery are needed.

Full text

Page 1

Hindawi Publishing Corporation
Journal of Obesity
Volume 2012, Article ID 972365, 7 pages
doi:10.1155/2012/972365
Research Article
TheImpact of SevereObesityonPost-Acute Rehabilitation
Efficiency,Lengthof Stay, and Hospital Costs
RajS.Padwal,1XiaomingWang,1AryaM. Sharma,1and DavidDyer2
1Department of Medicine, University of Alberta, Edmonton, AB, Canada T6G 2B7
2Glenrose Rehabilitation Hospital, Alberta Health Services, Edmonton, AB, Canada T5G 0B7
Correspondence should be addressed to Raj S. Padwal, rpadwal@ualberta.ca
Received 3 November 2011; Accepted 17 December 2011
Academic Editor: Francesco Saverio Papadia
Copyright © 2012 Raj S. Padwal et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background and Objective. The purpose of this retrospective observational study was to examine the influence of severe obesity
on length of stay (LOS), rehabilitation efficiency, and hospital costs post-acute rehabilitation in a population-based, tertiary care,
publicly-funded regional rehabilitation center. Participants. 42 severely obese subjects (mean age 53y; mean BMI 50.9kg/m2) and
42 nonobese controls (mean age 59y; mean BMI 23.0kg/m2) matched by sex and admitting diagnosis. Main Outcome Measures.
Total LOS, rehab LOS, waiting for transfer LOS, Fuctional Independence Measure (FIM) efficiency, and hospital costs. Results.
Compared to controls, severely obese subjects experienced longer total LOS (98.4 vs. 37.4 days; P = 0.03), rehabilitation LOS
(55.8 vs. 37.4 days; P = 0.04), and waiting for transfer LOS (42.6 vs. 0 days; P = 0.006); increased hospital costs ($115,822 vs.
$43,969; P = 0.03); and similar FIM efficiency (0.58 vs. 0.67; P = 0.27). Severe obesity was an independent predictor of total LOS
(beta-coefficient 0.51; P = 0.03), rehab LOS (0.46; P = 0.02) but not FIM efficiency (−0.63; P = 0.06). Conclusion. Severe obesity
adversely affects rehabilitation LOS and expenditures. Targeted interventions in severely obese individuals to optimize post-acute
rehabilitation care delivery are needed.
1.Introduction
Following acute medical illness, traumatic injury, or surgery,
patients often require rehabilitative care designed to improve
physical, cognitive, and functional status and reduce care-
giver burden [1]. Demand for this type of care, often
termed “post-acute rehabilitation” or PAR, is rising, in part
because the types of patients commonly referred for PAR
(elderly, frail, medically complex) are becoming increasingly
prevalent [1]. Thus, maximizing the efficiency and cost-
effectivenessofPARinthiseraofheightenedcompetition for
increasingly scarce health care resources is a major priority.
Severe or Class III obesity (body mass index or BMI ≥
40kg/m2) affects 3% of the adult Canadian population
and has tripled in prevalence over three decades [2, 3].
The long-term health consequences of Class III obesity are
well documented and include premature mortality; chronic
medical illness; increased health care costs; dramatically
reduced health-related quality of life (QOL) [4, 5]. In
contrast, the impact of severe obesity on PAR outcomes
has not been well studied even though obese individuals
are more likely than their normal weight counterparts to
be frail, functionally disabled, and medically complex [5–7].
A greater understanding of this relationship will help plan
service delivery needs if anticipated increases in the number
of severely obese patients accessing PAR occur.
Two previous studies examining PAR after hip and knee
arthroplasty reported higher costs and less efficient rehabil-
itation in severely obese subjects compared to either over-
weight or nonobese controls [8, 9]. However, these data have
limited generalizability because they were both performed in
private health care settings and focused solely on orthopedic
PAR. Another US study examining deconditioned general
medical patients reported no difference in rehabilitation effi-
ciency in severely compared to normal weight subjects [10].
The objective of this study was to examine the impact
of severe obesity on length of stay (LOS), functional status
change, and inpatient costs in a population-based, publicly
funded, tertiary care regional Canadian rehabilitation center.
Wehypothesizedthatsevereobesitywouldbeassociatedwith
reduced rehabilitation efficiency, greater lengths of stay, and
higher costs during PAR compared to controls.

Page 2

2Journal of Obesity
2.Methods
2.1. Setting. This retrospective analysis examined data from
the Glenrose Rehabilitation Hospital (GRH), the largest
freestanding academic tertiary-care rehabilitation center in
Canada. The GRH delivers comprehensive rehabilitation
services to patients from Alberta, Western Arctic, British
Columbia, and Northwestern Saskatchewan serving a catch-
ment area of over 9 million people and treating over 1,800
patients per year.
2.2. Rehabilitation Program. The Glenrose Rehabilitation
Hospital provides specialized comprehensive rehabilitation
for children, adults, and older adults with physical and
or cognitive disability. Coordinated services including as-
sessment, therapeutic interventions, medical interventions,
counseling, and patient and family education are all pro-
vided. Direct and indirect client services are provided by all
or a combination of 16 clinical disciplines. Direct care is
available six days per week and includes physical rehabili-
tation therapy delivered by physical therapists, occupational
therapists, speech language pathologists, and dietitians as
well as specialty trained rehabilitation nurses. Indirect care
includes services such as aiding and advising rehabilitation
clients and families on issues including discharge planning,
home modifications, and financial planning. The goals of
the programs are to optimize the client’s ability to regain
and maintain his or her functional abilities to resume their
desired lifestyle back in the community.
2.3. Data Collection. A prospectively collected electronic
registry for all GRH inpatients containing admission body
weight grouped into three categories (<90kg, 90–136kg,
and ≥137kg) was queried identify adult patients (age ≥ 18
years) admitted between April 1, 2004 and March 31, 2009
with a body weight over 137kg. Next, 1:1 controls with
body weights <90kg matched on sex, admission diagnosis
(orthopaedic, medical, or brain/spinal cord injury), and
admission date (within 2 years) were randomly selected.
Chart reviews were then performed to ascertain height and
calculate BMI. Cases with BMI levels <40kg/m2(n = 2) or
who died in hospital (n = 2) were excluded.
Additional sociodemographic and clinical data were col-
lectedforeachsubject,includingage,maritalstatus(married
or common law, single, or widowed), employment status
(employed,unemployedorsocialassistance,retired),andthe
following medical comorbidities (coded as binary variables):
hypertension, diabetes, dyslipidemia, stroke, heart disease,
peripheral vascular disease, chronic obstructive pulmonary
disease, asthma, sleep apnea, cancer, arthritis, dementia,
smoking, mental health, neurologic disease, renal failure,
venous thromboembolic disease, and presence of a pressure
ulcer.
2.4. Functional Status Assessment. Functional status was
assessed using the Functional Independence Measure (FIM)
tool [11]. The FIM is a valid, reliable, and sensitive measure
of disability severity and burden of care in the inpatient
clinical rehabilitation setting and is widely used to monitor
rehabilitation progress over the course of an admission
[12, 13]. Eighteen activities of daily living reflecting both
motor and cognitive disability are rated on a 7-point scale
ranging from fully independent (7) to fully dependent (1).
Summation of the scores across the 18 items reflects overall
burden of care [14]. Possible overall scores range from
18 to 126, with higher scores indicating greater functional
independence. The FIM can be subdivided into a 13-item
motor domain and a 5-item cognitive domain.
Trained and certified data collectors recorded and calcu-
lated FIM scores at admission and discharge in all patients.
FIM efficiency was calculated by dividing the change in
FIM score from admission to discharge by the rehabilitation
LOS (see below). Motor and cognitive subscores were also
calculated.
2.5. Length-of-Stay Measurement. Patient-specific identifiers
were used to perform data linkage with Canadian Institute
for Health Information (CIHI) administrative data to ascer-
tain LOS. CIHI collects data from inpatient facilities across
Canada (excluding Quebec) [15]. Extensive quality control
measures ensure high-quality data collection within CIHI
inpatient data; the frequency of missing data elements is less
than 0.001% [15].
Overall LOS was subdivided into the rehabilitation LOS
and waiting-transfer-of-service (WTS) LOS. Rehabilitation
LOS measured the duration of active rehabilitation and WTS
LOS, measured the time period between the completion
of active rehabilitation and discharge from the hospital.
PatientsaccumulateWTSLOSdaysiftheyareawaitinghome
environment modification prior to discharge or have been
unable to progress to independence and require placement
in a long-term care facility.
2.6. Costs. Total hospital costs were calculated for severely
obese subjects and controls by multiplying the average
overall length of stay for each study group by $1177 Cdn, the
average daily cost of an inpatient stay at GRH.
2.7. Statistical Analysis. The means and proportions of
baselinevariables,includingFIMscores,FIMefficiency,LOS,
and total costs were compared between cases and controls
using Wilcoxon rank sum tests. Results are reported for
the overall population and according to each of the three
admitting diagnostic groups.
Multivariable linear regression was used to determine
the independent predictors of overall LOS, rehab LOS, WTS
LOS, and FIM efficiency in the overall population. Log
transformation of each outcome was necessary to satisfy
model assumptions. Age and study group were first forced
into all models. As the patients were matched based on
sex and admission diagnosis, these were not entered into
the models. Additional covariates with a P value <0.25
on univariable analysis were also considered in the initial
model. Potential model covariates included age, sex, marital
status (married versus single/widowed), employment status

Page 3

Journal of Obesity3
Table 1: Baseline characteristics.
Variable Severe obesity (n = 42)
32 (76)
53 (15)
158.8 (25.3)
177.6 (10.2)
50.9 (8.3)
Controls
(N = 42)
32 (76)
59 (14)
69.3 (12.4)
173.1 (10.3)
23.0 (3.0)
P value
Sex, male, no (%)∗
Age, yr, mean (SD)
Weight, kg, mean (SD)
Height, cm, mean (SD)
Body mass index, kg/m2, mean (SD)
Reason for requiring rehabilitation, no (%)∗
Orthopaedic surgery
Acute medical illness
Stroke, brain or spinal cord injury
Married or common law
Single or widow
Employment status, no (%)
Employed
Unemployed or social assistance
Retired
Smoking, no (%)
Never
Current
Former
Hypertension
Diabetes, no (%)
Dyslipidemia, no (%)
Stroke, no (%)
Heart disease, no (%)
Peripheral vascular disease, no (%)
Chronic obstructive pulmonary disease, no (%)
Asthma, no (%)
Sleep apnea, no (%)
Cancer, no (%)
Arthritis, no (%)
Dementia, no (%)
Mental illness, no (%)
Neurological disease, no (%)
Renal failure, no (%)
Pressure ulcer, no (%)
Venous thromboembolic disease, no (%)
∗Subjects were matched on sex and admitting diagnosis.
—
0.06
<0.0001
0.05
<0.0001
26 (62)
8 (19)
8 (19)
18 (43)
24 (57)
26 (62)
8 (19)
8 (19)
19 (45)
23 (55)
—
—
—
0.83
0.83
13 (31)
13 (31)
16 (38)
17 (40)
11 (26)
14 (33)
0.36
0.63
0.65
18 (43)
5 (12)
19 (45)
39 (92)
36 (86)
19 (45)
9 (21)
16 (38)
6 (14)
5 (12)
3 (7)
18 (43)
0 (0)
16 (38)
4 (10)
20 (48)
10 (24)
12 (29)
9 (21)
4 (10)
15 (36)
13 (31)
14 (33)
19 (45)
12 (29)
12 (29)
7 (17)
8 (19)
5 (12)
11 (26)
0 (0)
0 (0)
3 (7)
8 (19)
6 (14)
12 (29)
3 (7)
4 (10)
7 (17)
3 (7)
0.50
0.03
0.26
<0.0001
<0.0001
0.11
0.58
0.05
0.75
0.10
0.08
<0.0001
0.08
0.05
0.50
0.07
0.03
0.03
0.58
0.69
(employed, unemployed or social assistance, retired), reason
for requiring rehabilitation (orthopedic surgery, acute med-
icalillness,strokeorbrain/spinalcordinjury),smoking(nev-
er, former, current), and medical comorbidities (Table 1).
The final model was created using a stepwise backwards
selection method to determine which of these additional
covariates contributed to the model at a Wald Chi-square P
value of 0.25. SAS (Version 9.2, Cary, NC) was used for all
analyses. P values less than 0.05 were considered statistically
significant.
2.8. Ethics Approval. Ethics approval was obtained from the
University of Alberta Research Ethics Board.
3.Results
3.1. Baseline Characteristics. The baseline sociodemographic
characteristics of all 84 patients are detailed in Table 1. Sixty-
two percent of subjects were admitted after orthopaedic
surgery, 19% after admission for acute medical illness, and
19% following stroke, brain, or spinal cord injury. Subjects

Page 4

4Journal of Obesity
Table 2: Length of stay, functional independence measure, and cost outcomes.
Outcome
Length of Stay
Total LOS, days, mean (SD)
Rehab LOS, days, mean (SD)
WTS LOS, days, mean (SD)
Functional independent measure
Total FIM score admission, mean (SD)
Total FIM score discharge, mean (SD)
Overall FIM efficiency, mean (SD)
FIM motor score on admission, mean (SD)
FIM motor score on discharge, mean (SD)
Change in motor score, mean (SD)
FIM cognitive score on admission, mean (SD)
FIM cognitive score on discharge, mean (SD)
Change in cognitive score, mean (SD)
Costs
Total costs, mean (SD)
Severe obesity (n = 42)Controls (N = 42)
P value
98.4 (186.3)
55.8 (42.66)
42.6 (167.00)
37.4 (20.9)
37.4 (20.9)
0 (0)
0.03
0.04
0.006
85.5 (16.6)
107.2 (12.5)
0.58 (0.53)
54.1 (14.3)
74.4 (11.6)
20.2 (14.7)
31.3 (5.9)
32.9 (3.8)
1.5 (3.7)
85.7 (23.4)
106.8 (19.9)
0.67 (0.54)
55.7 (18.3)
76.2 (13.9)
20.0 (13.3)
29.7 (8.4)
30.6 (7.4)
1.0 (2.2)
0.48
0.54
0.27
0.41
0.20
0.90
0.50
0.17
0.53
115,822 (219,314)43,969 (24,603)0.03
LOS: length of stay; FIM: functional independence measure; SD: standard deviation; WTS: waiting transfer of service.
were 76% male. Severely obese subjects were younger
(53 versus 59 years old), heavier (BMI 50.9 versus
23.0kg/m2) and had a significantly higher prevalence of
hypertension (93 versus 45%), diabetes (86 versus 29%),
sleep apnea (43 versus 0%), neurological disease (24 versus
7%),andrenalfailure(29versus10%)butalowerprevalence
of current smoking (12 versus 31%). The range of BMI
values was 41–76kg/m2in the severely obese patients and
17–27kg/m2for controls.
3.2. LOS, FIM, and Cost Outcomes. Severely obese subjects
hadasignificantlygreatermeanoverallLOS(98.4versus37.4
days; P = 0.03), rehab LOS (55.8 versus 37.4 days P = 0.04),
and WTS LOS (42.6 versus 0 days; P = 0.006) compared
to controls (Table 2). Admission and discharge FIM scores
and FIM score changes were similar between study groups
(Table 2).FIMefficiencywaslowerinseverelyobesesubjects;
this difference was not statistically significant (0.58 versus
0.67; P = 0.27). FIM motor and cognitive subscore changes
were nearly identical between severely obese subjects and
controls (Table 2).
Total daily hospital costs were $115,822 in the severely
obese compared to $43,969 in controls (P = 0.03).
LOS and FIM outcomes stratified by admitting diagnosis
are presented in Table 3. LOS outcomes were consistently
higher in the severely obese, but results were statistically
significant only in the orthopedic group. FIM efficiency was
not significantly different between the severely obese and
controls in any of the subgroups.
3.3. Linear Regression Analysis. The results of separate cova-
riate-adjusted linear regression models examining log overall
LOS, log rehab LOS, and log FIM efficiency are detailed
in Table 4. Severe obesity was an independent predictor of
greater overall LOS (beta-coefficient 0.51; P = 0.03) and
rehab LOS (0.46; P = 0.02). Severe obesity was associated
with reduced FIM efficiency, but this association was not
statistically significant (−0.63; P = 0.06).
4.Discussion
In summary, severely obese subjects undergoing PAR in
a population-based, publically funded rehabilitation center
experienced increased lengths of stay and incurred greater
hospital costs but had similar rehabilitation efficiency com-
paredtononobesecontrols.Toourknowledge,thisisthefirst
study to examine PAR outcomes in severely obese patients
cared for within a publically funded health care system.
Our findings should prompt clinicians and decision makers
within our rehabilitation program and similarly structured
programs to better understand the reasons for increased LOS
and costs and develop strategies to mitigate these effects.
Few other studies have examined PAR outcomes in the
severely obese and all have taken place within the US health
care system. Two controlled observational studies of post-
orthopedic-surgery patients reported higher costs and lower
FIM efficiency in severely obese subjects compared to non-
obese controls [8, 9]. LOS results in these two studies
were inconsistent. In 177 patients undergoing PAR following
hip arthroplasty at a university-affiliated teaching center,
severely obese subjects had longer hospital stays compared
to overweight patients (12.3 versus 11.8 days; P < 0.05) [9].
Conversely, in 5428 patients across 15 sites undergoing PAR
after total knee replacement, LOS was shorter in the severely
obese compared to nonobese subjects (9.7 versus 10.3 days;
P < 0.05)[8].Astudyof1077deconditionedgeneralmedical
patients reported that FIM score gains were not significantly
different between severely obese and normal weight subjects

Page 5

Journal of Obesity5
Table 3: Length of stay and functional independence measure outcomes by admitting diagnosis subgroup.
Outcome
After orthopaedic surgeryMedical illness Stroke/brain/spinal injury
Severe
obesity
(n = 26)
120.5
(225.3)
Controls
(n = 26)
P value
Severe
obesity
(N = 8)
46.5 (25.9)
Controls
(n = 8)
P value
Severe
obesity
(N = 8)
78.6
(126.7)
Controls
(n = 8)
P value
Total LOS, day,
mean (SD)
Rehab LOS, day,
mean (SD)
WTS LOS, day,
mean (SD)
FIM score
admission, mean
(SD)
FIM score
discharge, mean
(SD)
FIM efficiency,
mean (SD)
37.3 (22.4)0.02 35.0 (19.2)0.43
39.8 (19.9)0.83
62.3 (47.3)37.3 (22.4)0.02 42.5 (15.7)35.0 (19.2)0.43 48.1 (44.8)39.8 (19.9)0.83
58.2
(207.0)
0.0 (0.0)0.02 4.0 (11.3)0.0 (0.0) 0.3230.5 (86.3)0.0 (0.0) 0.32
88.9 (15.3) 94.4 (15.2)0.21 84.8 (16.4)88.3 (13.0)0.56 75.9 (18.5) 54.9 (28.8)0.14
109.0
(10.7)
114.0 (4.1)0.12107.3 (7.8)112.1 (5.2) 0.34
103.0
(19.7)
78.4 (32.9) 0.03
0.49 (0.49) 0.66 (0.62)0.28 0.56 (0.40) 0.78 (0.47) 0.210.88 (0.68) 0.62 (0.32)0.34
LOS: length of stay; FIM: functional independence measure; SD: standard deviation; WTS: waiting transfer of service.
Table 4: Linear regression analysis examining the effect of severe obesity on log LOS and log FIM efficiency.
ModelVariable Beta-coefficient (SE)
P value
Overall LOS
model R2= 0.1798
Intercept4.44 (0.46)
<0.0001
0.03 Severe obesity0.51 (0.23)
Retired (versus employed) 0.67 (0.26) 0.01
Age (per one year increase)
−0.02 (0.009)
−0.26 (0.19)
−0.26 (0.22)
4.14 (0.37)
0.04
smoking former
0.17
Diabetes
0.25
Rehab LOS
model R2= 0.1693
Intercept
<0.0001
0.02 Severe obesity0.46 (0.18)
Retired (versus employed)0.47 (0.21) 0.03
Diabetes
−0.32 (0.18)
−0.01 (0.01)
−0.21 (0.15)
−2.12 (0.59)
−0.63 (0.33)
−0.84 (0.27)
−0.60 (0.34)
0.65 (0.25)
0.08
Age (per one year increase)
0.09
Mental illness
0.16
FIM efficiency
model R2= 0.3133
Intercept
0.0006
Severe obesity
0.06
Unemployed (versus employed)
0.003
Heart Disease
0.08
Mental illness 0.01
Asthma 1.01 (0.63) 0.11
Current smoker (versus never)0.60 (0.30) 0.05
Age0.02 (0.01) 0.07
HTN0.52 (0.32)0.10
Venous thromboembolic disease0.68 (0.42)0.13
Dementia
−0.66 (0.42)
−0.55 (0.40)
0.37 (0.30)
0.13
PVD
0.18
DM0.23
LOS: length of stay; FIM: functional independence measure.

Page 6

6Journal of Obesity
[10]. Other PAR studies have not specifically focused on
severe obesity per se but, rather, have included patients
with lesser degrees of adiposity. These studies have reported
smaller FIM score gains in patients with excess adiposity
undergoing PAR after spinal cord or burn injuries [16, 17].
Thus,whenourresultsandtheresultsofpriorstudiesare
considered in aggregate, PAR costs are consistently increased
intheseverelyobese,butPARefficiencyandLOSfindingsare
contradictory. Differences between studies in rehabilitation
programs, underlying reasons for requiring PAR, prevalence
of comorbid conditions and health care delivery models
may account for some of these contradictory results. We
also note that the BMI range selected for the control group
differs between studies: normal (BMI 18.5–24.9kg/m2),
nonoverweight (BMI < 25kg/m2), and overweight (BMI
25.0–29.9kg/m2) controls have all been used and this may
partly account for the discrepant findings. The BMI level(s)
associated with optimal values for PAR efficiency and LOS
also requires further clarification. When examined across the
range of BMI levels, the relationships between BMI and LOS,
BMI and FIM efficiency, and BMI and costs were curvilinear
after orthopedic surgery [9]. That is, LOS and costs were
higher and FIM efficiency was lower in patients with either
very low or very high BMI and optimal values for each of
theseoutcomeswerefoundatBMIlevelsofaround30kg/m2.
In deconditioned medical patients, FIM score gains were
greatest in the group with BMI levels between 30 and
35kg/m2. We hypothesize that the presence of sicker, frailer
and nutritionally compromised patients within the lower
BMI categories may result in worse PAR outcomes, although
further study is required to clarify this issue.
Our LOS and FIM score change results were consis-
tent when stratified by admitting diagnosis, but only the
orthopedic surgery subgroup LOS results achieved statistical
significance and this may have resulted from a lack of power
due to the low sample sizes in the other two subgroups.
We also found an increase in WTS LOS in the severely
obese; in contrast, increased WTS LOS was not found in the
control group. No prior study has reported this outcome. We
suspect that the increased WTS LOS in the severely obese is a
consequence of the patient’s inability to gain independence
following rehabilitation. In our experience, these subjects
cannot return home and due to a lack of suitable alternative
discharge destinations, often wait in hospital for transfer to
a nursing home. Further study is required to determine the
extent to which this issue contributes to a “bed blocking”
effect (preventing use of these beds for other patients
requiring PAR) and to reductions in cost-efficient healthcare
delivery.
Targeted interventions to optimize PAR delivery and
improve PAR efficacy in the severely obese may poten-
tially lead to result in earlier discharge, reduced LOS, and
decreased costs. Specifically, tailored interventions designed
to improve the motor impairment, mobility disability, and
frailty commonly found in the severely obese may improve
outcomes [6, 7]. One might consider modifying rehabilita-
tion equipment so that traditional components of PAR can
be safely delivered in the severely obese. Employing activities
such as water-based exercises that “cushion” the excess bulk
found in these subjects may make physical exercise easier
and less stressful on joints. Multidisciplinary case manage-
ment, with specific attention paid to the challenges and
complexities found in severely obese subjects may also help
to optimize care and preempt complications and setbacks.
However there are very limited published data on
bariatric-specific PAR interventions and this deficiency
was recently recognized at a multidisciplinary consensus
conference [18]. Many potential barriers to developing
effective rehabilitation strategies in bariatric patients were
identifiedbythisexpertpanel,includingthelackofbariatric-
specificrehabilitationprogramsinbothacuteandpost-acute
rehabilitation, a paucity of standards of care specific to obese
patients,weightbiasonthepartofhealthcarestaff,andalack
of training programs and research consortia specializing in
bariatric rehabilitation [18]. The higher prevalence of medi-
cal complexity, mental health impairment, and psychosocial
dysfunction in the severely obese were also recognized as
potential barriers to effective rehabilitation and factors that
may increase the likelihood of acute illness relapse during
PAR [5, 18].
A major strength of this study is the low risk of selection
bias, as the GRH serves as the primary rehabilitation center
foralargecatchmentareaand,incontrasttoprivatelyfunded
rehabilitation centers, admission to the GRH is not influ-
enced by payment or reimbursement factors. In addition, we
did not limit our analysis to a specific type of PAR program;
rather, we examined all severely patients admitted for PAR
over a five-year span. Unlike other studies, we also were
able to breakdown overall LOS into separate components to
examine the effect of severe obesity on rehabilitation and
nonrehabilitation LOS. However, a limitation of this study is
that the sample size of this study was not large and this may
have limited the statistical power of the study, particularly
when performing subgroup analyses. A second limitation
was that we only abstracted data on severely obese patients
and nonobese controls and are were unable to compare the
severely obese patients to patients with lesser degrees of
obesity.
The need for PAR in severely obese patients will
undoubtedly increase given the rising prevalence of severe
obesity and expected increases in the need for PAR [1].
Data from this study will help clinicians and adminis-
trators estimate the effect of severe obesity on PAR in
their institution allocate resources accordingly. A proactive
approach to PAR in the severely obese is recommended and
further research should focus on developing interventions to
maximize efficient PAR delivery in this patient population.
Conflict of Interests
The authors declare no conflict of interests with respect to
this work.
Acknowledgments
This study was unfunded. R. S. Padwal had full access to
the data and takes responsibility for the integrity of the data
and accuracy of the data analysis. R. S. Padwal and A. M.

Page 7

Journal of Obesity7
Sharma are supported by an alternative funding plan from
the Government of Alberta and the University of Alberta. A.
M. Sharma is supported by an Alberta Health Services Chair
in Obesity Research and Management.
References
[1] A. W. Heinemann, “State of the science on postacute rehabil-
itation: setting a research agenda and developing an evidence
base for practice and public policy: an introduction,” Assistive
Technology, vol. 20, no. 1, pp. 55–60, 2008.
[2] M. Shields, M. Carroll, and C. Ogden, “Adult obesity preva-
lenceinCanadaandtheUnitedStates,”NCHSDataBriefs,vol.
56, pp. 1–8, 2011.
[3] M. Tjepkema, “Measured obesity. Adult obesity in Canada:
measured height and weight,” in Nutrition: Findings from the
Canadian Community Health Survey 2004, Issue I (cat no 82-
620-MWE2005001), 2008, http://www.statcan.ca/english/re-
search/82-620-MIE/2005001/articles/adults/aobesity.htm.
[4] A. Peeters, J. J. Barendregt, F. Willekens et al., “Obesity in
adulthoodanditsconsequencesforlifeexpectancy:alife-table
analysis,” Annals of Internal Medicine, vol. 138, no. 1, pp. 24–
32, 2003.
[5] D. D. Hensrud and S. Klein, “Extreme obesity: a new medical
crisis in the United States,” Mayo Clinic Proceedings, vol. 81,
supplement 10, pp. S5–S10, 2006.
[6] R. E. Hubbard, I. A. Lang, D. J. Llewellyn, and K. Rockwood,
“Frailty,bodymassindex,andabdominalobesityinolderpeo-
ple,” Journals of Gerontology A, vol. 65, no. 4, pp. 377–381,
2010.
[7] D. T. Villareal, M. Banks, C. Siener, D. R. Sinacore, and S.
Klein, “Physical frailty and body composition in obese elderly
men and women,” Obesity Research, vol. 12, no. 6, pp. 913–
920, 2004.
[8] H. K. Vincent and K. R. Vincent, “Obesity and inpatient reha-
bilitation outcomes following knee arthroplasty: a multicenter
study,” Obesity, vol. 16, no. 1, pp. 130–136, 2008.
[9] H. K. Vincent, J. P. Weng, and K. R. Vincent, “Effect of
obesity on inpatient rehabilitation outcomes after total hip
arthroplasty,” Obesity, vol. 15, no. 2, pp. 522–530, 2007.
[10] N. B. Jain, S. Al-Adawi, A. S. S. Dorvlo, and D. T. Burke,
“Association between body mass index and functional inde-
pendence measure in patients with deconditioning,” American
JournalofPhysicalMedicine&Rehabilitation,vol.87,no.1,pp.
21–25, 2008.
[11] Uniform Data System for Medical Rehabilitation website,
2011, http://www.udsmr.org.
[12] T. A. Dodds, D. P. Martin, W. C. Stolov, and R. A. Deyo, “A
validation of the functional independence measurement and
its performance among rehabilitation inpatients,” Archives of
Physical Medicine and Rehabilitation, vol. 74, no. 5, pp. 531–
536, 1993.
[13] M. E. Cohen and R. J. Marino, “The tools of disability
outcomes research functional status measures,” Archives of
Physical Medicine and Rehabilitation, vol. 81, no. 12, part B,
pp. S21–S29, 2000.
[14] M. G. Stineman, J. A. Shea, A. Jette et al., “The functional in-
dependence measure: tests of scaling assumptions, structure,
and reliability across 20 diverse impairment categories,”
Archives of Physical Medicine and Rehabilitation, vol. 77, no.
11, pp. 1101–1108, 1996.
[15] Canadian Institute for Health Information, “Discharge Ab-
stract Database,” 2011, http://www.cihi.ca//CIHI-ext–portal/
pdf/internet/DAD EXECUTIVE SUM 09 10 EN.
[16] K. W. Stenson, A. Deutsch, A. W. Heinemann, and D. Chen,
“Obesity and inpatient rehabilitation outcomes for patients
with a traumatic spinal cord injury,” Archives of Physical
Medicine and Rehabilitation, vol. 92, no. 3, pp. 384–390, 2011.
[17] R. T. Farrell, R. L. Gamelli, R. F. Aleem, and J. M. Sinacore,
“The relationship of body mass index and functional out-
comes in patients with acute burns,” Journal of Burn Care and
Research, vol. 29, no. 1, pp. 102–108, 2008.
[18] M. Forhan, Y. Bhambhani, D. Dyer, X. Ramos-Salas, M.
Ferguson-Pell, and A. Sharma, “Rehabilitation in bariatrics:
opportunities for practice and research,” Disability and Reha-
bilitation, vol. 32, no. 11, pp. 952–959, 2010.