You are on page 1of 6

t h e s u r g e o n 1 6 ( 2 0 1 8 ) 4 0 e4 5

Available online at www.sciencedirect.com


ScienceDirect
The Surgeon, Journal of the Royal Colleges
of Surgeons of Edinburgh and Ireland
www.thesurgeon.net

A comparison of outcomes in morbidly obese,


obese and non-obese patients undergoing primary
total knee and total hip arthroplasty

Angela H. Deakin*, Aghimien Iyayi-Igbinovia, Gavin J. Love


Department of Orthopaedics, Golden Jubilee National Hospital, Agamemnon Street, Clydebank, G81 4DY, United
Kingdom

article info abstract

Article history: Background: Obesity is a growing public health issue with the prevalence of morbid obesity,
Received 28 January 2016 (Body Mass Index (BMI)  40 kg/m2) increasing. There is some evidence these patients have
Received in revised form more peri- and post-operative complications and poorer outcomes when undergoing
10 October 2016 arthroplasty procedures. This audit aimed to determine and compare the outcomes of non-
Accepted 24 October 2016 obese, obese and morbidly obese patients undergoing arthroplasty at our institution.
Available online 27 January 2017 Method: This was a retrospective audit of patients from our institution who had undergone
total knee (TKA) or total hip arthroplasty (THA) in 2009. Data collected were: age, gender,
Keywords: BMI, length of stay (LOS), Oxford knee or hip score (OKS/OHS), satisfaction and complica-
Morbid obesity tions up to two years post operation. Patients were divided into three groups: BMI < 30, BMI
Total knee arthroplasty 30e40 and BMI > 40. Outcomes for each BMI group were compared.
Total hip arthroplasty Results: 1014 TKA and 906 THA operations were included. When compared to obese and
Patient satisfaction non-obese patients, morbidly obese patients undergoing TKA had a mean LOS one day
Complications longer, a mean OKS four points lower and higher rates of postoperative problems, 37% vs.
21%. For THA patients there was no difference in LOS, OHS score was two points lower for
each increasing BMI category and postoperative problems increase from 25% for non-obese
to 31% for obese and 38% for morbidly obese patients.
Conclusion: These results will be useful in informing obese patients of their potential out-
comes following TKA or THA. These patients can then make a more informed choice before
proceeding with arthroplasty.
© 2016 Royal College of Surgeons of Edinburgh (Scottish charity number SC005317) and
Royal College of Surgeons in Ireland. Published by Elsevier Ltd. All rights reserved.

I ¼ 30e34.99, Class II ¼ 35e39.99, Class III ¼ 40.1 The defini-


Introduction tion of severe or morbid obesity has varied, with NIH defining
it as BMI > 35 in 19912 and NICE defining it as BMI > 40 in 2002.3
The World Health Organization defines obesity as a Body Mass Within the UK morbid obesity is more usually used to refer to
Index (BMI) 30 kg/m2 and lists three classes of obesity; Class BMI > 40. The prevalence of obesity in the United Kingdom is

* Corresponding author.
E-mail address: angela.deakin@gjnh.scot.nhs.uk (A.H. Deakin).
http://dx.doi.org/10.1016/j.surge.2016.10.005
1479-666X/© 2016 Royal College of Surgeons of Edinburgh (Scottish charity number SC005317) and Royal College of Surgeons in Ireland.
Published by Elsevier Ltd. All rights reserved.
t h e s u r g e o n 1 6 ( 2 0 1 8 ) 4 0 e4 5 41

rising, with a trebling in the number of obese people in En- hospital in Scotland up to two years after surgery were ob-
gland and Wales over the last 20 years.4 Recent meta-analyses tained from NHS Information Services Division (ISD). The
have shown complications are higher for obese patients after following complications related to surgery were identified:
total knee arthroplasty (TKA)5,6 and that in total hip arthro- Acute Myocardial Infarction (AMI); Cerebrovascular accident
plasty (THA) there were more complications and worse (CVA); death; Deep Vein Thrombosis (DVT)/Pulmonary Em-
functional outcomes.7 bolism (PE); dislocation (THA only); infection; revision surgery.
However much of the literature has been carried out on: Readmissions for other reasons were grouped under “other”.
historical cohorts with a wide range of operation dates Patients were divided into three groups: those with
meaning that they also reflect substantial changes in practice BMI < 30 (non-obese), patients with BMI of 30e40 (obese) and
and variations of proportions of obese patients within the those with BMI > 40 (morbidly obese). Outcomes for each
population during the study; cohorts where there are only group were compared using appropriate statistical tests (Chi-
small numbers of morbidly obese patients or comparisons square for categorical/ordinal data, median test for non-
have not been made across all obesity groups i.e. morbidly parametric continuous data and ANOVA for parametric
obese are compared to non-obese; only TKA or THA cohorts or continuous data).
these have been analysed as total joint arthroplasty so there is
little information comparing TKA and THA arthroplasty from
the same practice. Therefore the aims of this audit were to Results
determine the function, satisfaction and complications within
two years of operation for all patients undergoing arthroplasty Total knee arthroplasty
at our institution within a single calendar year. Knees and
hips would be treated separately and the outcomes compared A total of 1014 TKA operations were included. Fifty-four
based on obesity groups (non-obese, obese and morbidly percent of the cohort was female, mean age was 68.6 years
obese). This would provide information on outcomes and (SD 8.8, range 41e96) and mean BMI was 32.0 (SD 5.4, range
complications for obese and morbidly obese patients and 16e55). When grouped by BMI 35.8% of patients were not
highlight any difference between knee and hip patients in obese (BMI < 30), 54.7% were obese (BMI of 30e40) and 7.7%
order to guide future practice. were morbidly obese (BMI > 40). Patient in the morbidly obese
group were significantly younger and more likely to be female
Methodology compared to those in other BMI groups (p < 0.001). The com-
parison of demographics across BMI groups is given in Table 1.
This was a retrospective audit carried out under the Clinical The median LOS was 5 days in both the non-obese and
Governance procedures of our institution, including Caldicott obese groups but was significantly longer at 6 days in the
Guardian approval for the use of patient data. As this study morbidly obese group (median test p ¼ 0.042) (Table 2). At one
was purely a retrospective review of data already collected as year follow-up the mean OKS was 38 in both non-obese and
part of each patient's clinical treatment it did not require obese groups, but was 34 in the morbidly obese group which
ethical review. The selection criteria for the study were having was significantly lower (ANOVA p < 0.001) (Table 2). In the
had either an elective primary total knee arthroplasty (TKA) or morbidly obese group fewer patients were very satisfied
primary total hip arthroplasty (THA) carried out by any of the although this was only approaching significance (72% vs, 84%
nine consultant orthopaedic surgeons working at our insti- and 84%, Chi-squared p ¼ 0.054) but there was no significant
tution in 2009, giving an unselected consecutive series. Uni- difference in the numbers unsure or dissatisfied (3% vs, 6%
condylar knee replacements, emergency procedures and and 6%, Chi-squared p ¼ 0.635) (Table 2).
revision operations were excluded. In the morbidly obese group 37% of patients had a
A list of patients meeting the selection criteria was collated complication (requiring readmission) or problem (dealt with
from both the hospital and departmental databases. All data through outpatient clinic) related to surgery, which was
for any events or appointments up to two years following significantly more than the non-obese group, 21%, and the
operation were reviewed. For each patient the following data obese group, 21% (Chi-square p ¼ 0.004) (Table 3). The majority
were collected either from hospital databases or case notes: of these problems were in the immediate post-operative
age, gender, BMI, length of stay (LOS e patients were admitted period and resolved satisfactorily. Within the first two years
on day prior to operation), Oxford knee scores (OKS) or Oxford after operation 14% of morbidly obese patients (11 patients)
hip scores (OHS) (0e48 scale, 48 being best outcome) and pa- were readmitted to hospital for any reason compared to 8% of
tient satisfaction, both from the one year follow up, which was non-obese and 8% of obese patients. For readmissions related
the standard clinical care review point when these data were to surgery these were 12% (9 patients) in the morbidly obese
recorded. All post-operative appointments up to two years
post-surgery were reviewed for any indication of problems
relating to surgery. This included any problems reported by
Table 1 e Demographics of the different BMI groups for
patients, those identified by the independent Arthroplasty
TKA; Data presented as mean (SD) [range].
Service which carried out both routine and addition follow-up
BMI  30 30 < BMI  40 BMI > 40
appointments and those identified via consultant review.
Standard appointments were at six weeks and one year post- n 381 555 78
operative for TKAs, and three months and one year for THAs. Gender 47% female 55% female 78% female
Age 71.3 (8.7) [42e96] 67.5 (8.6) [41e86] 63.7 (7.5) [48e77]
In addition all complications leading to readmission to any
42 t h e s u r g e o n 1 6 ( 2 0 1 8 ) 4 0 e4 5

The median LOS was 5 days in all BMI groups (median test
Table 2 e Outcomes of BMI groups for TKA. Data
p ¼ 0.583) (Table 5). At one year follow-up the mean OHS were
presented as median [IQR] {range} or mean (SD) [range].
significantly different between groups, being 43 for the non-
BMI  30 30 < BMI  40 BMI > 40
obese group, 41 in the obese group and 39 in the morbidly
n ¼ 381 n ¼ 555 n ¼ 78
obese group (ANOVA p ¼ 0.001) (Table 5). In the morbidly
Length of stay 5 [3] {3e31} 5 [2] {2e59} 6 [3] {3e18} obese group 87% of patients were very satisfied and 3% were
(days)
unsure or dissatisfied which was not significantly different
1 year OKS 38 (8) [48e12] 38 (8) [48e5] 34 (10) [48e4]
1 year satisfaction (%)
when compared to 90% and 2%, and 91% and 2% in the non-
Very satisfied 83.8 84.3 72.1 obese and obese groups respectively (Chi-squared p ¼ 0.955)
Satisfied 9.8 9.8 24.6 (Table 5).
Unsure 3.7 4.9 3.3 Significantly more patients in the morbidly obese group,
Dissatisfied 2.8 1.1 0.0 38%, had a complication or problem compared to 25% in the
non-obese group and 31% in the obese group (Chi-square
p ¼ 0.047) (Table 6). The majority of these problems were in the
Table 3 e Number of patients with problems and early post-operative period and resolved satisfactorily. Within
complications related to their TKA operation by BMI the first two years after operation 15% (6 patients) from the
group. morbidly obese group were readmitted to hospital for any
BMI  30 30 < BMI  40 BMI > 40 reason compared to 9% of non-obese and 10% of obese pa-
n ¼ 381 n ¼ 555 n ¼ 78 tients. For readmissions related to surgery in the morbidly
Complication or problem within first two years obese group this was 8% (3 patients) compared to 7% for non-
Total 80 (21%) 117 (21%) 29 (37%) obese and 8% for obese patients which was not significantly
Post-operative problems dealt with via out-patient clinic different (Chi-square p ¼ 0.914) (Table 4).
Wound/SSSI 14 (4%) 17 (3%) 3 (4%)
Stiffness 4 (1%) 5 (1%) 3 (4%)
Haematoma 0 (0%) 0 (0%) 1 (1%)
Swelling 5 (1%) 12 (2%) 7 (9%) Discussion
Pain 27 (7%) 37 (7%) 5 (6%)
Other 4 (1%) 4 (1%) 1 (1%) This retrospective audit examined outcomes for patients with
Major complications related to surgery requiring re-admission different levels of obesity in patients who had TKA or THA
AMI 1 (<1%) 3 (<1%) 0 (0%) carried out at our institution. We found that high BMI patients
CVA 0 (0%) 0 (0%) 0 (0%)
were more likely to be younger and female, which has already
Death 0 (0%) 0 (0%) 0 (0%)
been widely reported.8e13 For those undergoing TKA we found
DVT/PE 2 (<1%) 10 (2%) 2 (3%)
Infection 6 (2%) 7 (1%) 3 (4%) morbidly obese patients (BMI > 40) had poorer outcomes than
Revision 4 (1%) 10 (2%) 1 (1%) obese and non-obese, which had very similar outcomes.
Other 13 (4%) 12 (2%) 3 (4%) However for those undergoing THA there were differences in
outcomes between all groups, with each increasing BMI group
doing slightly worse.
group compared to 7% for non-obese and 8% for obese pa-
tients (Chi-square p ¼ 0.358) (Table 3).
Total knee arthroplasty

Total hip arthroplasty The literature varies as to whether higher BMI patients have a
longer post-operative stay after TKA with some authors
There were 906 THA operations included. Sixty-three percent showing an increase14 and others not8,15,16 although Napier
of the cohort was female, mean age was 68 years (SD 10, range et al.15 used the lower cut off of BMI > 35 and in Krushell and
34e91) and mean BMI was 29.7 (SD 5.3, range 16e49). When Fingeroth16 patients were not discharged directly to home.
grouped by BMI 57.0% were not obese (BMI < 30), 38.6% were This audit supports the former with the median length of stay
obese (BMI of 30e40) and 4.4% were morbidly obese (BMI > 40). for morbidly obese patients being a day longer than others. We
Patients in the morbidly obese group were significantly
younger (p < 0.001) and more likely to be female, although this
only approached statistical significance (p ¼ 0.077). The Table 5 e Outcomes of BMI groups for THA. Data
comparison of demographics across BMI groups is given in presented as median [IQR] {range} or mean (SD) [range].
Table 4. BMI  30 30 < BMI  40 BMI > 40
n ¼ 516 n ¼ 350 n ¼ 40
Length of stay 5 [3] {2e71} 5 [3] {3e31} 5 [4] {3e57}
Table 4 e Demographics of the different BMI groups for (days)
THA; Data presented as mean (SD) [range]. 1 year OHS 43 (7) [48e9] 41 (8) [48e7] 39 (8) [48e15]
1 year satisfaction (%)
BMI  30 30 < BMI  40 BMI > 40
Very satisfied 90.3 91 86.5
N 516 350 40 Satisfied 7.4 7.0 10.8
Gender 65% female 59% female 72% female Unsure 1.8 1.3 2.7
Age 70 (10) [38e91] 66 (10) [34e 87] 63 (6) [38e79] Dissatisfied 0.5 0.7 0.0
t h e s u r g e o n 1 6 ( 2 0 1 8 ) 4 0 e4 5 43

meant to be joint specific score it may well be affected by


Table 6 e Number of patients with problems and
overall health and therefore very high BMI patients may report
complications related to their THA operation by BMI
group. a lower score due to general difficulties with movement rather
than their knee replacement having poorer function. A more
BMI  30 30 < BMI  40 BMI > 40
in-depth study looking at actual knee function would be
n ¼ 516 n ¼ 350 n ¼ 40
required to ascertain this.
Complication or problem within first two years
Our institution is a high volume dedicated elective lower
Total 129 (25%) 110 (31%) 15 (38%)
Post-operative problems dealt with via out patient clinic
limb arthroplasty centre and in general our patients report
Wound/SSSI 21 (4%) 33 (9%) 4 (10%) high levels of satisfaction. However this study has shown that
Stiffness 5 (1%) 2 (<1%) 0 (0%) even within this environment morbidly obese patients are less
Swelling 8 (2%) 6 (2%) 2 (5%) likely to be very satisfied which has been shown before.16
Pain 13 (2%) 16 (5%) 1 (2%) However obese patients show similar levels of satisfaction to
Clicking or 5 (1%) 2 (<1%) 0 (0%)
non-obese. Further work looking specifically at what leads
clunking
patients to be very satisfied is required to understand why the
Trendelenburg 17 (3%) 7 (2%) 1 (2%)
gait very high BMI patients show lower satisfaction, however it
Heterotopic 1 (<1%) 2 (<1%) 0 (0%) could be due to the higher numbers of post-operative prob-
ossification lems seen in this group. However in this cohort dissatisfaction
Leg length 12 (2%) 9 (3%) 3 (7%) with TKA does not appear to be related to BMI.
discrepancy In this audit, we found a significantly higher rate of post-
Nerve palsy/foot 5 (1%) 2 (<1%) 0 (0%)
operative problems in the morbidly obese group compared
drop
to non-obese and obese patients, although we could not show
Other 6 (1%) 4 (1%) 1 (2%)
Major complications related to surgery requiring re-admission a difference in major complications, probably due to the lack
AMI 4 (1%) 0 (0%) 0 (0%) of power. The literature shows increased complications in
CVA 1 (<1%) 0 (0%) 0 (0%) both obese11,20 and morbidly obese patients11,15,16,18,19
Death 0 (0%) 0 (0%) 0 (0%) although one smaller study did not find this in BMI > 35 pa-
Dislocation 9 (2%) 5 (1%) 0 (0%) tients.9 It should also be noted that different studies use
DVT/PE 6 (1%) 6 (2%) 1 (2%)
different and sometimes more restrictive list of complica-
Infection 3 (<1%) 6 (2%) 1 (2%)
Revision 9 (2%) 9 (3%) 0 (0%)
tions. This might explain the different results seen as higher
Other 4 (1%) 1 (<1%) 1 (2%) BMI does seem to lead to more bleeding, wound and infection
problems and revisions6,16,20,21 but differences are not seen in
PE or mortality.20,21 When counselling morbidly obese pa-
believe this is the first cohort to look at length of stay in high tients before TKA the higher risks of post-operative problems
BMI patients where an enhanced recovery after surgery pro- should be discussed.
gramme was in place for all patients.17 The longer median
length of stay in the morbidly obese may indicate that these Total hip arthroplasty
patients are unable to benefit as much from the early mobi-
lisation and the more rapid rehabilitation that an enhanced As with TKA for THA the literature varies as to whether higher
recovery programme provides. The fact that the high BMI BMI patients have a longer post-operative stay with some
patients were younger might lead to expecting a shorter LOS authors showing an increase14 and others not12,22 although
but we did not find this indicating that the BMI was a stronger Michalka et al. used the lower cut off of BMI > 35.22 This audit
factor.14 supports the latter with the median length of stay for all pa-
A number of small studies have reported no significant tient groups being five days. We believe this is the first cohort
difference in outcome scores (such as OKS or Knee Society to look at length of stay in high BMI patients where an
Score) for TKA patients with higher BMIs15,18,19 but many enhanced recovery after surgery programme was in place for
indicate that obese6,9,20 and morbidly obese10,16 patients have all patients.17 This shows that, in contrast to the TKA patients,
lower (worse) outcomes although the clinical significance of an enhanced recovery program can provide the same results
some of the small changes seen for obese patients is ques- in higher BMI patients in terms of LoS. This may be due to the
tioned.6 Our findings would support the view that it is the very THA cohort generally having a lower BMI.
high BMI patients who are more likely to have poor outcomes, Our OHS scores showing a two point decrease between
rather than all the obese. There may be a number of reasons each group are very similar to those reported in the litera-
for the lower scores. There is evidence that high BMI patients ture12,13,23,24 and studies using the Harris Hip Score also
have on average a lower pre-operative functional level.10,16 showed worse outcomes for obese patients.7,25 Although
Therefore even if they gain the same improvement their these are low changes in OHS they do reach clinical signifi-
post-operative function will be less.10 Unfortunately we were cance of four points between the non-obese and morbidly
unable to assess improvement within our cohort. However obese. Also our data and other literature seem to indicate a
maybe caution should be exercised when relating post- linear trend to worse scores with increasing BMI.7,24 This
operative function between different ranges of BMIs. Ideally could indicate that, even though the THA population has a
pre-operative function should be known along with the pa- lower mean BMI, many patients may benefit from a reduction
tient's expectations in order to more accurately predict a in BMI before surgery unlike in TKA where it seems that obese
satisfactory post-operative outcome. Also although the OKS is patients have very similar outcomes to non-obese. It has been
44 t h e s u r g e o n 1 6 ( 2 0 1 8 ) 4 0 e4 5

reported that for short term follow up no difference in OHS are of pre-operative scores which makes it impossible to assess
seen but that obese patients had better improvement in and comment on changes in scores post-operatively. Finally
scores.22 As we did not have pre-operative OHS we were not there are variations in the literature in the categorisation of
able to confirm whether or not high BMI patients had the same morbid obesity as either BMI > 35 kg/m2 or BMI > 40 kg/m2.
or greater levels on improvement. This makes interpretation and comparison of results more
Again in contrast to the TKA patients we found similar difficult.
rates of satisfaction across all BMI groups which concur with This study has provided information about how outcomes
published results.22 and complications vary with obesity in a large unselected
In this audit, we found significant changes in the rate of cohort treated in a large specialist arthroplasty centre over a
post-operative problems in the different BMI groups with the single calendar year with the same standardised enhanced
morbidly obese group having the highest rates, although we recovery peri-operative care relevant to modern joint
did not show a difference in major complications, which may replacement surgery. It therefore has high external validity,
be due to the lack of power. The literature shows increased particularly compared to other study designs, such as RCTs
complications in both obese7,23 and morbidly obese pa- and matched cohorts. Due to this it may provide more rele-
tients.13,21,25 One large study showed no difference in com- vant information as to current real world outcomes than some
plications but they assessed each complication individually of the historical literature. We found high BMI patients tend to
giving the study low power to find any differences and another be younger than the general TKA or THA arthroplasty popu-
small study also showed no short term differences but again lation and are more likely to be female. This study also high-
would have had low power.12,22 It should also be noted that lighted the differences between TKA and THA. In this cohort
different studies use different and sometimes more restrictive TKA patients had a higher BMI than THA patients and the
list of complications. This might explain the different results results would seem to indicate that increasing BMI has a
seen as higher BMI does seem to lead to more bleeding, wound negative effect on THA patients, moving from non-obese to
problems, infections, dislocations, DVT/PE and revisions but obese to morbidly obese whereas for TKA patients differences
not other complications.7,13,23,25 When counselling obese pa- are only really seen in the very high BMI patients. Due to the
tients for THA the higher likelihood of post-operative prob- variation in outcomes it would be useful to carry out a cost
lems should be discussed. benefit analysis on treating high BMI patients but this was
outside the remit of our audit. Our results will be useful in
informing obese patients of their potential outcomes
Summary following TKA or THA. These patients can then make a more
informed choice before proceeding with arthroplasty.
This study was a retrospective audit which stratified pa-
tients by BMI. This is an unselected consecutive series. The
audit did not collect information on the number of patients
Conflict of interest
who were declined surgery due to being medically unfit (as
defined by a consultant anaesthetist), or the BMI for those
The authors have no conflicts of interest to declare.
patients. However during the time of the audit there was no
No financial support was received for this study.
selection for joint replacement based on BMI so this cohort
should be an accurate representation of the whole popula-
tion who are able to undergo a joint replacement operation.
We have analysed TKA and THA separately whereas some Acknowledgements
studies have combined total joint arthroplasties.14,21 This
has given lower numbers of morbidly patients in each The authors would like to thank Nadia Sciberras, Mohammed
cohort (78 TKA and 40 THA) so reducing power to investigate Almustafa and Marie Anne Smith for their help with the data
these patients but has allowed us to show the differences collection and Mohammad Nickdel for his help with preparing
with BMI for each operation. The use of data from one cal- the manuscript.
endar year from a single institution also reduces the power
of this audit to see statistically significant differences,
particularly in the THA group. However it does allow the references
comparison of outcomes where all patients have received
the same standardised peri-operative care and there have
not been large variations in practice within the cohort being 1. World Health Organisation. Global database on body mass index
analysed. As we found that the higher BMI patients tended e BMI classification. 2016 [Accessed 31 March 2016], http://
to be younger and female the differences in outcomes seen apps.who.int/bmi/index.jsp?introPage¼intro_3.html.
may be in part due to this. However due to the small 2. National Institute for Health. Gastrointestinal surgery for
numbers it was not possible to carry out multivariable lo- severe obesity. Consensus statement. NIH Consens Dev Conf
1991;9(1) [Accessed 31 March 2016], https://www.google.co.uk/
gistic regression to isolate these effects. However as high
url?sa¼t&rct¼j&q¼&esrc¼s&source¼web&cd¼2&ved¼0ahU
BMI patients have these characteristics the findings are
KEwj-xJi9_erLAhXBuw8KHenlBw8QFggpMAE&url¼https%3A%
representative of this group even if the underlying cause is 2F%2Fconsensus.nih.gov%2F1991%2F1991GISurgeryObesity
also strongly linked to a different demographic character- 084PDF.pdf&usg¼AFQjCNFka7YTyMwVHO9DwKihmBba9e
istic. A further limitation of our study was the unavailability rog&bvm¼bv.118353311,d.ZWU.
t h e s u r g e o n 1 6 ( 2 0 1 8 ) 4 0 e4 5 45

3. Clegg A, Sidhu MK, Colquitt J, Royle P, Loveman E, Walker A. analysis from the New Zealand joint registry. J Arthroplast
Clinical and cost effectiveness of surgery for people with morbid 2014;29:1884e8.
obesity. 2011 [Accessed 31 March 2016], https://www.google.co. 14. Bradley BM, Griffiths SN, Stewart KJ, Higgins GA, Hockings M,
uk/url?sa¼t&rct¼j&q¼&esrc¼s&source¼web&cd¼3&ved¼0 Isaac DL. The effect of obesity and increasing age on operative
ahUKEwjS67Xw5OrLAhXIoA4KHQL7BeIQFggrMAI&url¼ time and length of stay in primary hip and knee arthroplasty.
https%3A%2F%2Fwww.nice.org.uk%2Fguidance J Arthroplast 2014;29:1906e10.
%2FTA46%2Fdocuments%2Fclinical-and-cost-effectiveness- 15. Napier RJ, O'Brien S, Bennett D, Doran E, Sykes A, Murray J,
of-surgery-for-people-with-morbid-obesity2&usg¼AFQjC et al. Intra-operative and short term outcome of total knee
NFvm69-UEXIHPWYwwf6qKkmPFbPWQ. arthroplasty in morbidly obese patients. The Knee
4. NHS Health and Social Care Information Centre. Health survey 2014;21:784e8.
for England 2004: updating of trend tables to include 2004 data. 2005 16. Krushell RJ, Fingeroth RJ. Primary total knee arthroplasty in
[Accessed 10 December 2014], http://www.hscic.gov.uk/ morbidly obese patients: a 5- to 14-year follow-up study. J
catalogue/PUB01152/hea-surv-upd-tren-tab-eng-2004-rep.pdf. Arthroplast 2007;22:77e80.
5. Chen J, Cui Y, Li X, Miao X, Wen Z, Xue Y, et al. Risk factors for 17. Removed for blinding.
deep infection after total knee arthroplasty: a meta-analysis. 18. Amin AK, Clayton RAE, Patton JT, Gaston M, Cook RE,
Arch Orthop Trauma Surg 2013;133:675e87. Brenkel IJ. Total knee replacement in morbidly obese
6. Kerkhoffs GM, Servien E, Dunn W, Dahm D, Bramer JA, patients: results of a prospective, matched study. J Bone Jt Surg
Haverkamp D. The influence of obesity on the complication Br Vol 2006;88-B:1321e6.
rate and outcome of total knee arthroplasty: a meta-analysis 19. Dewan A, Bertolusso R, Karastinos A, Conditt M, Noble PC,
and systematic literature review. J Bone Jt Surg Am Parsley BS. Implant durability and knee function after total
2012;94:1839e44. knee arthroplasty in the morbidly obese patient. J Arthroplast
7. Haverkamp D, Klinkenbijl MN, Somford MP, Albers GHR, van 2009;24. 89,94.e3.
der Vis HM. Obesity in total hip arthroplastyddoes it really 20. Si HB, Zeng Y, Shen B, Yang J, Zhou ZK, Kang PD, et al. The
matter? Acta Orthop 2011;82:417e22. influence of body mass index on the outcomes of primary
8. Lozano LM, Tio M, Rios J, Sanchez-Etayo G, Popescu D, total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc
Sastre S, et al. Severe and morbid obesity (BMI >/¼ 35 kg/m(2)) 2015;23:1824e32.
does not increase surgical time and length of hospital stay in 21. Friedman RJ, Hess S, Berkowitz SD, Homering M.
total knee arthroplasty surgery. Knee Surg Sports Traumatol Complication rates after hip or knee arthroplasty in morbidly
Arthrosc 2015;23:1713e9. obese patients. Clin Orthop Relat Res 2013;471:3358e66.
9. Collins RA, Walmsley PJ, Amin AK, Brenkel IJ, Clayton RA. 22. Michalka PKR, Khan RJK, Scaddan MC, Haebich S,
Does obesity influence clinical outcome at nine years Chirodian N, Wimhurst JA. The influence of obesity on early
following total knee replacement? J Bone Jt Surg Br outcomes in primary hip arthroplasty. J Arthroplast
2012;94:1351e5. 2012;27:391e6.
10. Rajgopal V, Bourne RB, Chesworth BM, MacDonald SJ, 23. Jameson SS, Mason JM, Baker PN, Elson DW, Deehan DJ,
McCalden RW, Rorabeck CH. The impact of morbid obesity on Reed MR. The Impact of body mass index on patient reported
patient outcomes after total knee arthroplasty. J Arthroplast outcome measures (PROMs) and complications following
2008;23:795e800. primary hip arthroplasty. J Arthroplast 2014;29:1889e98.
11. Werner BC, Evans CL, Carothers JT, Browne JA. Primary total 24. Judge A, Batra RN, Thomas GE, Beard D, Javaid MK,
knee arthroplasty in super-obese patients: dramatically Murray DW, et al. Body mass index is not a clinically
higher postoperative complication rates even compared to meaningful predictor of patient reported outcomes of
revision surgery. J Arthroplast 2015;30:849e53. primary hip replacement surgery: prospective cohort study.
12. Andrew JG, Palan J, Kurup HV, Gibson P, Murray DW, Beard DJ. Osteoarthr Cartil 2014;22:431e9.
Obesity in total hip replacement. J Bone Jt Surg Br Vol 2008;90- 25. Chee YH, Teoh KH, Sabnis BM, Ballantyne JA, Brenkel IJ.
B:424e9. Results of a prospectively matched study: total hip
13. Murgatroyd SE, Frampton CMA, Wright MS. The effect of body replacement in morbidly obese patients with osteoarthritis. J
mass index on outcome in total hip arthroplasty: early Bone Jt Surg Br Vol 2010;92-B:1066e71.