R e s e a rc h

An enhanced recovery after surgery program for

hip and knee arthroplasty
The ERAS steoarthritis is the leading A b s tra c t
program had
a small but
significant
O cause of pain and disability
am ong the elderly and af­
fects 15% of the population. Despite a
range of treatments for osteoarthritis
Objective: To in stitu te and evaluate th e benefits o f an enhanced recovery
a fte r surgery (ERAS) program across three hosp itals in Victoria.
Design, setting and participants: We used a b e fo re -a n d -a fte r q u a lity
im p rove m e nt stu dy design consisting o f three phases: pre-ERAS
(OA), joint replacement rem ains the program d ata colle ction fro m March to S eptem ber 2012; ERAS training
effect on main treatment option for patients in and im p le m e n ta tio n during S e ptem be r 2012; and change perform ance
hospital stay, whom the disease has progressed.1 m easurem ent fo llo w in g ERAS im p le m e n ta tio n fro m O ctober 2012 to May
2013.
particularly In Victoria, m ore th a n 20000 hip
an d knee joint replacem ents are Main outcome measures: The prim ary end p oint w as duration o f h osp ital
for knee stay a fte r knee or hip arthroplasty. Secondary end points w ere adherence to
now perform ed each year, reflect­
replacement ing orthopaedic practices globally.
th e ERAS bundle, and process and p a tie n t recovery characteristics.

patients The prevalence of OA and the need Results: We enrolled 412 p a tie n ts to th e pre-ERAS (existing -practice )
phase and com pared th e m w ith 297 patie nts in the ERAS phase. For ERAS
for joint replacem ent are likely to
patients, com pared w ith existing-practice patients, h o sp ita l stay was
increase because of a com bination reduced (geom etric mean, 5.3 [SD, 1.6] v 4.9 [SD, 1.6] days; P < 0.001) and
of increasing risk factors (age, obesity) there w as a sign ifican t im p rove m e nt in th e p roportion o f p a tie n ts ready
and im proved surgical and anaes­ for discharge on Day 3 a fte r surgery (41% v 5 9% ; P< 0.001). The m ost
thetic techniques that make surgery co m m o n reason fo r delayed discharge w as p a tie n ts w a itin g fo r review or
possible for more people.1 access to reh ab ilita tion services. There were m arkedly im proved indicators
o f processes and outco m e s o f care, including im proved p a tie n t education,
Across health services, there is wide reduced fa sting tim es, less blood loss, b e tte r analgesia, earlier a m b u la tio n
variation in hospital length of stay and im proved overall q u a lity o f recovery.
N ic h o la s C h ris te lis
FFPMRCA, FANZCA,
for patients receiving hip and knee Conclusion: We foun d th a t an ERAS program could be successfully
FFPMANZCA12 rep lacem en ts. T his is p ro b ab ly im p lem en te d in elective jo in t arthroplasty, leading to a shorter d uration o f
S o p h ie W a lla c e
independent of casemix and more h osp ital stay. We recom m end this orth op a ed ic ERAS pathway.
BHSc(Nurs), reflective of varying health service
G radC ertlnfectC ont, MPH'
practices. Surgical injury, p ain ,
C la ir e E S a g e stress-induced catabolism, impaired perioperative care (the fast-track interventions used in a multimodal,
BAppS ci(N urs), MNurs'
organ function and im paired cog­ m ethodology) reduces the need integrated clinical care pathw ay to
U a t e B a b itu nitive fu n ctio n m ay contribute to for prolonged hospitalisation and achieve im proved functional out­
MB BS, FANZCA3
com plications, p ro lo n g ed h o sp i­ convalescence, and reduces m or­ comes and rapid recovery.6 ERAS
S u s a n L ie w talisatio n , p o sto p erativ e fatigue, bidity, w ith subsequent economic pathways have led to reduced hos­
M B B S (H o n s). FRACS(O rth)’
delayed convalescence and the need savings.2'6 Enhanced recovery after pital stays after hip or knee replace­
Jam es D ugal for reh ab ilitatio n . O p tim isa tio n su rg ery (ERAS) p ro g ram s are a m ents — as short as 3 days in m any
MBBS, FRACS3
of in d iv id u al care com ponents in care package of evidence-based centres.5'8
Ib o ly a N y u la s i
MSc, G radD ipM gt12

1Sixteen predefined enhanced recovery after surgery items for hip or knee arthroplasty
N o r a M u t a l im a
BSc.M BChB, D P I # • Nurse coordinator counselling in the orthopaedic or preadmission clinic
• Preadmission review by a physiotherapist and/or dietitian
T o n T ra n
MBBS, FRACS, FAOA5 • Minimal fasting preoperatively, defined as clear oral fluids up to 2 hours before surgery
• Preoperative oral carbohydrate loading
P a u l s M y le s
MD. FRARCSI. FANZCA' • No sedative premedication (benzodiazepines, opioids or neuroleptics)
• Pre-emptive analgesia w ith paracetamol and gabapentinoids according to protocols
1 Th e A lfre d H o spital,
M elbourne, VIC.
• Spinal anaesthesia (not epidural)
2 M onash U niversity, • Local anaesthesia technique (surgeon-delivered local infiltration o f analgesia or anaesthetic fem oral nerve block)
M elbourne. VIC. • Minimal (« 1 0 mg) intravenous morphine intraoperatively
3 Bendigo H o spital,
• Intraoperative avoidance of excessive intravenous fluids (knee, >1L; hip, > 2 L; both: subtracting blood loss)
Bendigo, VIC.

4 D a ndenong H o spital,
• Active intraoperative warming (forced air warm ing and/or warm ed intravenous fluids)
M elbourne, VIC. • Antiem etic prophylaxis
5 M onash H ealth, • M ultim odal oral analgesia for s*3 days postoperatively, to include a non-steroidal anti-inflam m atory drug or
M elbourne, VIC.
cyclooxygenase-2 inhibitor
n .c h ris te lis @ • Early postoperative (recovery room) oral carbohydrate supplementation
a lf re d .o rg .a u
• Mobilisation within 24 hours
• Early hospital discharge (« 5 days) ♦
doi: 1 0 .5 6 9 4 /m ja l4 .0 0 6 0 1

MJA 202 (7) • 20 April 2015

363
R e s e a rc h

Phase 1: over the 6 m onths before

2 Patient demographic and perioperative characteristics*
im plem entation of the ERAS pro­
Existing practice ERAS
Variable (n = 412) (n = 297) P gram, we recorded perioperative data
Sex, m ale
for all eligible patients undergoing
164 (4 0 % ) 113 (3 8 % ) 0 .6 4
surgery (the existing-practice cohort).
M ean age, years (SD) 68(11) 6 7 (1 0 ) 0.22
Mean w eight, kg (SD) 8 4 (1 9 ) 8 7 (2 0 ) 0.0 92 Phase 2: training of staff m anaging
M edical h istory orthopaedic surgical patients. For
1 month, the evidence-based back­
C urrent sm oker 4 6 (11%) 3 0 (10% ) 0.65
ground to ERAS was promulgated to
H ypertension 2 8 4 (6 9 % ) 194 (6 5 % ) 0.31
all surgical, anaesthetic and nursing
Coronary artery disease 70 (17%) 41 (14% ) 0.25 staff. This was done in various forms
Stroke 2 0 (5 % ) 10 (3 % ) 0.33 including lectures, workshops, meet­
H eart failure 19 (5 % ) 15 (5 % ) 0 .8 6 ings and written instructions.
Peripheral vascular disease 13 (3 % ) 8 (3 % ) 0.72
Phase 3: change perform ance. We
Diabetes 81 (2 0 % ) 73 (2 5 % ) 0.12 undertook a repeat audit following
COPD 8 8 (21%) 67 (2 3 % ) 0.70 the implementation of the ERAS care
Preoperative a naem ia 3 6 (9 % ) 22 (7 % ) 0.52 package (the ERAS cohort).
Usual m edications
O ur study received ethics approval
O pioid 106 (2 6 % ) 5 8 (2 0 % ) 0.053 as an audit project w ith a w aiver
Aspirin w ith in 5 days 5 8 (14% ) 5 4 (18% ) 0.14 for specific patient consent (Alfred
C lopidogrel w ith in 7 days 2 (< 1 % ) 0 0.23 Human Research Ethics Committee,
W arfarin w ith in 7 days 2 2 (5 % ) 6 (2 % ) 0.025 EC 92/12).
NSAID/COX-2 inhibitor 109 (2 6 % ) 109 (37 % ) 0 .0 0 4 The pre-ERAS phase ran from March
ACE inhibitor/A R B 235 (57 % ) 168 (5 7 % ) 0.90 to September 2012. Training of staff
Beta blocker 71 (17%) 51 (17%) 0.98 took place over September 2012. The
S ta tin 141 (3 4 % ) 8 9 (3 0 % ) 0.37 ERAS phase ran from October 2012
C alcium channel blocker 9 8 (2 4 % ) 73 (2 5 % ) 0.81
to May 2013.
Diuretic 120 (2 9 % ) 79 (27% ) 0.45 Patient health status w as q u anti­
O ral h ypoglycaem ic 57 (14% ) 6 0 (2 0 % ) 0.0 24 fied using the American Society of
Insulin 9 (2 % ) 11 (4 % ) 0.25 A nesthesiologists physical status
LMW H 2 0 (5 % )
classification, ranging from 1 (healthy
8 (3 % ) 0.15
patient) to 5 (moribund patient not
ASA physical s ta tu s 0.571
expected to survive w ithout the
1 10 (2 % ) 10 (3 % )
operation). Patient quality of recov­
2 223 (5 4 % ) 161 (5 4 % ) ery was assessed using the patient-
3 172 (4 2 % ) 122 (41% ) centred, 15-item quality-of-recovery
4 7 (2 % ) 3 (1% ) score,9and the 12-item World Health
Disease 0.22 Organization Disability Assessment
O ste o arth ritis 3 6 7 (8 9 % ) 275 (9 3 % )
Schedule 2.0 score.10 Indicators of
patient satisfaction were assessed at
R heum atoid a rth ritis 4 (1% ) 0
30 days after surgery using a 5-item
A vascular necrosis 14 (3 % ) 7 (2 % )
L ikert scale (0 = strongly agree,
O ther 27 (7 % ) 15 ( 5 % ) 5 = strongly disagree).
Previous PONV or m o tio n sickness 124 ( 3 0 % ) 81 (2 7 % ) 0.41
Actual hospital stay was tim ed from
ACE = a n g io te ns in -c o nv e rtin g enzym e. ARB = ang io te n s in -re c e p to r blocker. ASA = A m erican
S o cie ty o f A n esthesiologists. COPD = c hronic o b s tru c tiv e p u lm o n a ry disease. the beginning of surgery until dis­
COX = cyclooxygenase. ERAS = enhanced recovery a fte r surgery, LM W H = lo w m o le c u la r w e ig h t charge. We evaluated readiness for
heparin. NSAID = n o n -s te ro id a l a n ti-in fla m m a to ry drug. PONV = pos to pe ra tiv e nausea and
discharge on postoperative Day 3,
vo m itin g . * D ata are no. ( % ) o f p a tie n ts unless o th e rw is e s pecified. f P value derived fro m x 2 te s t
fo r trend. ♦ defined by whether patients were eat­
ing and drinking, had no drain tubes
or u rin ary catheters, were w eight­
We aim ed to assess the extent to Methods bearing, and had well controlled pain
which a predefined ERAS program scores (visual analogue scale with a
for orthopaedic surgical patients We used a before-and-after study range of 0 to 10) at rest and on move­
could be achieved, and to evaluate design consisting of three phases. ment of less than 3 and 5, respectively.
improvements in quality of care and Public health services involved in A successful ERAS implementation
patient outcome across three public the study were the Alfred, Bendigo required at least 11 of 16 prespecified
hospitals in Victoria. and Monash hospitals. ERAS items (Box 1).

364 MJA 202 (7) ■ 20 A pril 2015

 R e s e a rc h

O n com pletion of our study, there

3 P eriop erative and surg ica l care *
was a concern raised by the surgical
team at the lead institution regarding Existing practice
an apparent increased incidence of Variable (n = 412) ERAS (n = 297) p
acute kidney injury (AKI). In view Preadmission clinic, seen by:
of the widespread use of local anaes­ Nurse1 40 6 (9 9 % ) 297 (100% ) 0.037
thetic infiltration with a solution that
Anaesthetist 405 (9 8 % ) 297(100% ) 0.024
included ketorolac 30 mg, we chose to
investigate this more formally at the Surgeon 4 0 6 (9 9 % ) 297 (100% ) 0.037

lead institution. We retrospectively Physiotherapist1 331 (8 0 % ) 224 (75%) 0.12

retrieved all perioperative creatinine Occupational therapist 324 (79% ) 249 (84 % ) 0.077
data for the study cohorts. AKI was Dietitian1 0 61 (21%) <0.001
defined according to AKI Netw ork11
Day of surgery
and RIFLE (risk, injury, failure, loss,
end-stage kidney disease)12 criteria. Admission on day of surgery 4 0 4 (9 8 % ) 294 (9 9 % ) 0.32
We did not include urine output or Shower w ith antibiotic soap 399 (97% ) 238 (8 0 % ) <0.001
oliguria in the definitions of AKI, in Preoperative skin wipes 4 (1%) 91 (31%) <0.001
part because we did not collect these
Oral (clear) fluids given1 2 (< 1%) 180 (61%) <0.001
data, b u t prim arily because u rin e
output is an unreliable indicator of Oral carbohydrate drink1 0 248 (84 % ) <0.001
renal function in the perioperative Gabapentin premedication1 21 (5% ) 172 (58 % ) <0.001
setting. Type o f surgery

To account for the restrictive intra­ Hip 214 (52% ) 129 (43% ) 0.025
venous (IV) fluid regim en used in Knee 198 (48 % ) 168 (57%) 0.025
the ERAS cohort (which m ay have Revision 26 (6 % ) 13 (4% ) 0.26
artificially elevated serum creatinine
Tubes
because of the avoidance of a dilu-
tional effect from excessive IV fluids Urinary catheter 200 (49% ) 107 (36 % ) 0.001
increasing body water), we calculated Drain tube(s) 105(25% ) 59 (20 % ) 0.080
the adjusted creatinine concentration Type o f anaesthesia
by first estimating the volume of dis­
General (ire g io n a l) 266 (65% ) 164(5 5% ) 0.014
tribution for creatinine as equal to
Spinal1 236 (57% ) 205 (6 9 % ) 0.001
total body water (assumed to be 60%
of body weight, expressed in mL), and Epidural or CSE 16 (4% ) 3 (1%) 0.019
assum ing that 50% of IV fluid was Postoperative regional analgesia
still accumulated as tissue oedema at Nerve block used 135 (33% ) 44 (15%) <0.001
the time of postoperative creatinine
LA infiltration1 214 (52% ) 222 (75% ) <0.001
measurements:13
PONV prophylaxis1 233 (57%) 202 (6 8 % ) 0.002
adjusted creatinine concentra­ Mean to ta l IV fluids, mL (SD)1 1756 (767) 1446 (687) <0.001
tion = serum creatinine con­
Active (forced air) warming1 392 (95% ) 285 (9 6 % ) 0.34
centration x (l + [0.5 x IV fluid
balance/total body water]) Mean lowest temperature, °C (SD) 35.7(0.5) 36.2 (0.4) 0.039
Mean duration of surgery, hours (SD) 2.0 (0.9) 1.9 (0.6) 0.33
A nalysis of the data show ed that
there was no increased incidence of CSE = combined spinal and epidural. ERAS = enhanced recovery after surgery. IV =intravenous. LA = local anaesthesia.
PONV = postoperative nausea and vomiting. * Data are no. (%) of patients unless otherwise specified, t Key ERAS
AKI in the ERAS group (Appendix 1). implementation points. ♦
The prim ary end point of the study
w as d u ra tio n of h o sp ita l stay.
Secondary end points were adher­ reported as mean (SD) or median and stay. Therefore, we log-transformed
ence to the ERAS bundle (defined as interquartile range (IQR). Numerical hospital stay data to enable valid
& 11 items), and a num ber of patient data were first tested for normality comparison using the t test; in addi­
outcome m easures. A sam ple size using the Kolmogorov-Smirnov test tion, we report median (IQR) length
calculation based on a change in and then compared using the Student of stay an d results of W ilcoxon
hospital stay from a m ean of 7 days t test or Wilcoxon rank-sum test, as ran k -su m testing. Patients u n d er­
(SD, 4 days) to 6 days (SD, 3 days), with appropriate. Rates were com pared going each type of surgery were also
an a value of 0.05 and a |3 value of using x 2or Fisher exact test, as appro­ analysed as subgroups. A P value of
0.2, required at least 380 patients to priate. Hospital stay was expected to less than 0.05 was considered statisti­
be enrolled, but we included a larger be skewed to the right because of a cally significant. Data were analysed
sample in view of the planned sub­ small proportion experiencing com­ w ith SPSS version 20.0 for Windows
group analyses. Continuous data are plications and a protracted hospital (SPSS Inc).

MJA202 (7) ■ 20 April 2015

365
R e s e a rc h

su rg ery , th e ERAS p ro g ra m w as
4 Recovery p ro file an d h o s p ital sta y
associated w ith a red u ced h o sp ital
Existing pra c tic e ERAS stay (geom etric m ean, 5.3 [SD, 1.6] v
V a ria b le (n = 4 12) (n = 2 9 7 ) P
4.5 [SD, 1.5] days [P = 0.001]; m edian,
R ecovery room 5.0 [IQR, 4.0-6.7] v 4.1 [IQR, 3.0-6.0]
Median pain score (IQR ), a t rest* 0 (0 -5 ) 0 (0 -4 ) 0.047 days [P=0.005]); and a greater propor­
Median pain score (IQR ), on m o v e m e n t* 0 (0 -7 ) 0 (0 -4 ) <0.001 tion of patien ts w ere m ore likely to
be d isch arg ed by Day 5 (64% v 52%;
A dm ission te m perature, °C (SD) 36.1 (0.5 ) 36.0 (0 .6 ) 0 .0 0 6
P = 0.019). T here w as no c h a n g e in
P o s to p e ra tiv e , a t 2 4 hours
m ed ian hospital stay for h ip replace­
M edian pain score (IQR ), a t rest* 5 (3 -7 ) 4 (2 -5 ) <0.001 m e n t p a tie n ts in th e ERAS g ro u p
M edian pain score (IQR ), on m o v e m e n t* 6 (4 -8 ) 5 (3 -7 ) <0.001 com pared w ith the existing-practice
Mean q u a lity o f recovery score (5D ) (range, 0 -1 5 0 ) 103 (19) 106 (2 0 ) 0 .0 56 group (median, 5.0 [IQR, 3.5-70] v 5.0
[IQR, 4.0-6.9] days; P = 0.99). O verall,
T otal knee replacem ent
th e 75th cen tile for le n g th of stay
Mean knee flexion (SD ), degrees 51 (19) 57 (2 4 ) 0 .0 2 6
decreased from 6.8 to 6.0 days.
M edian quadriceps stren g th (IQR ), Nm 3 (2 -3 ) 2 (2 -3 ) 0.11
We fo u n d h ig h rates of com pliance
P o s to p e ra tiv e , a t 4 8 hours
w ith nearly all ERAS item s (Box 3).
M edian pain score (IQR), a t rest* 4 (2 -6 ) 3 (1 -5 ) < 0.001 T here w as in c re ase d u se of sp in a l
M edian pain score (IQR), on m o v e m e n t* 6 (4 -8 ) 5 (2 -7 ) 0.001 anaesthesia. The use of femoral nerve
T ota l knee replacem ent block (with or w ithout a catheter) was
su b stitu te d b y fav o u rin g su rg eo n -
Mean knee flexion (SD ), degrees 72 (19) 78 (14) 0 .0 0 9
delivered local an aesth etic in filtra­
M edian quadriceps stren g th (IQR ), Nm 3 (2 -3 ) 3 ( 2 -3 ) 0.90
tio n in 75% of cases; th is c h a n g e
R ecovery p a ra m e te rs , m e d ia n hours (IQ R ) in p ractice v a rie d across th e th re e
W eight bearing 1.1 (1.0—2.0) 1.0 (0 .9 -2 .0 ) 0.001 hospitals (98%, 37% an d 98%). There
O ral flu id intake 3.2 (2 .0 -5 .0 ) 2.7(17-4.1) 0.016 w ere im proved dynam ic p ain scores
an d quality of recovery (Box 4). There
O ral fo o d intake 7.0 (4.3-15) 6.3 (3 .2 -7 9 ) 0 .0 0 4
were im provem ents in other recovery
R em oval o f drain tu b es 27 (2 4 -4 2 ) 25 (2 3 -2 7 ) 0 .0 0 2
p aram eters (early feeding, am b u la­
Rem oval o f urinary ca th ete r 4 8 (4 2 -7 6 ) 33 (1 7 -6 0 ) < 0.001 tion an d rem oval of tubes). Patients
B lood tra n s fu s io n in h o s p ita l, no. o f p a tie n ts ( % ) 58 (14% ) 31 (10% ) 0.24 u n d e rg o in g k n ee rep lacem en t h ad
R e tu rn to th e a tre , no. o f p a tie n ts ( % ) 14 (3 % ) 10 (3 % ) 0.76 im p ro v ed flexion o n p o sto p erativ e
Days 1 an d 2.
Le n g th o f stay, days

G eom etric m ean (SD) 5 3 (1 .6 ) 4.9 (1.6) < 0.001 T he p ro p o rtio n of p a tie n ts rea d y
M edian (IQR) 5.0 (4 .0 —6.8) 5.0 (3 .8 -6 .2 ) 0.10
for d isc h a rg e on D ay 3 afte r s u r ­
gery w as significantly higher in the
ERA S = enhanced recovery after surgery. IQ R = interquartile range. * Visual analogue scale: 0 = none, 10 = severe. ♦
ERAS g ro u p c o m p a re d w ith th e
existing-practice group: 59% v 41%,
Results The ERAS p ro g ram led to a signifi­ respectively; relative risk, 1.35 (95%
cantly higher rate of successful imple­ Cl, 1.18-1.53); P < 0.001 (Box 6 an d
We enrolled 709 patients into the pro­ m e n tatio n of th is clinical p ath w a y A ppendix 1).
ject; 412 in the existing-practice cohort (2% v 81%; P < 0.001) (Box 3, Box 4, The 6-week com plication rates w ere
an d 297 in the ERAS cohort (Box 2). Box 5 a n d A p p en d ix 2). T he p o st­ sim ilar an d there w as no increase in
We achieved 100% patient follow-up im plem entation cohort h ad a signif­ the rate of hospital readm ission. Pain
to hospital discharge and 90% follow­ icantly in c re ase d n u m b e r of ERAS levels w ere sim ilar an d there w as a
u p at 6 w eeks; th ere w ere 41 (10%) in te rv e n tio n s c o m p a re d w ith th e h igher level of patien t satisfaction at
and 25 (8%) patients m issing in each existing-practice gro u p (m edian, 12 6 w eeks after surg ery (Box 5).
cohort, respectively. C om parison of [IQR, 10-13] v 8 [IQR, 7-10]; P < 0.001).
T he incid en ce of AKI w as co m p a­
data from the three hospitals show ed O v erall, th e re w a s a sig n ific a n t rable betw een groups (A ppendix 1).
that the patients w ere sim ilar dem o- red u ctio n in h o sp ital stay (geom et­ The fin al p lasm a cre atin in e values
graphically as w ell as having sim ilar ric m ean, 5.3 [SD, 1.6] v 4.9 [SD, 1.6] w ere slig h tly h ig h e r in th e ERAS
rates of physical functioning an d co­ days [P < 0.001]) (Box 4), w ith aro u n d group, b u t th is could be accounted
m orbidity. The existing-practice and half of the patients being discharged for by higher baseline (preoperative)
ERAS cohorts, w ith the exception of from hospital w ith in 5 days of su r­ values; ERAS p atien ts h a d a m e an
som e m edications, w ere com parable g ery (ERAS group, 60% v existing- change in creatinine from 78 m m ol/L
(Box 2), w hich allow ed u n ad ju ste d p ractice g ro u p , 52%; P = 0.086). For (SD, 24 m m o l/L ) preo p erativ ely , to
analyses b etw een groups. those u n dergoing knee replacem ent a fin al re a d in g of 79 m m o l/L (SD,

MJA 202 (7) • 2 0 A p ril 2015

366
 R e s e a rc h

24m m ol/L) postoperatively. There

5 Recovery p ro file a t 6 w e eks a fte r surgery*
were no cases of renal failure.
Existing practice ERAS
Variable (n = 412) (n = 297) P
Discussion W ound infection 21 (5 % ) 13 (4 % ) 0.99
Prosthesis infection 5 (1%) 2 (< 1 % ) 0 .6 0
Our study results indicate that imple­
menting an ERAS program may be P rosthetic jo in t dislocation 2 (< 1%) 3 (1%) 0.31
beneficial for other Victorian public P eriprosthetic fracture 0 0 > 0 .9 9
hospitals. The ERAS program had T hro m b o e m b o lis m 13 (3 % ) 10 (3 % ) 0.59
a sm all b u t significant effect on
Urinary tra c t infection 8 (2 % ) 2 (1% ) 0.22
hospital stay, particularly for knee
Death 2 (< 1 % ) 1 (< 1 % ) 0.85
replacem ent patients. A pertinent
finding from our study was that a W orst pain rating in past 24 hours1 2 (0 -4 ) 2 (0 -3 ) 0.01
higher proportion of patients m an­ E xtent o f d is a b ility in past 24 hours1 2 (0 -3 ) 1(0 -2 ) 0.37
aged through the ERAS care pathway P atient s a tis fa c tio n 1
compared w ith the existing-practice
W as surgery w o rth w h ile ? 1 (1-2 ) 1 (1-1) < 0.001
group (59% v 41%, respectively) were
deemed ready for discharge on post­ Did surgery im prove your d aily life? 1 (1-2 ) 1 d -2 ) 0.015

operative Day 3. Do you fe e l b e tter? 1 (1-2 ) 1 d -2 ) < 0.001

The limited effect on actual hospital Do you have tro u b le sleeping? 3 (2 -4 ) 3 (2 -4 ) 0.67
stay in this project is likely to be due H o s p ita l readm ission 25 (6 % ) 15 (5 % ) ■ 0.87
to one key factor: despite an effec­ ERAS = enhanced recovery after surgery. * Data are no. (%) of patients or median score (IQR). t Visual analogue scale:
tively implemented ERAS program, 0 = none, 10 = severe. $5-point Likert scale: 0 = strongly agree, 5 = strongly disagree. ♦
there were entrenched hospital prac­
tices that prevented earlier hospital
discharge even though patients were improvement in hospital stay for the
majority of patients. 6 Proportion of patients ready
deemed ready for discharge; that is, for discharge on Day 3 after
discharge p la n n in g practice was A dm inistrative and traditional pat­ surgery*
m ostly unchanged. This is due to terns of clinical practice limit oppor­ ioo% -
established w ard practices, includ­ tunities for change and are common
ing a repetitive requirement for many causes of delayed discharge from £ 80% -

joint replacement patients to undergo hospital.14'16 Perhaps specific fast-

0J

"to
their initial rehabilitation program track arthroplasty units that have Q. 60% - ----------------- — i

as an inpatient (delaying their dis­ o

evidence-based and protocolised
charge). Patients referred for reha­
rapid recovery pathw ays can opti­
bilitation often w ait for some tim e
mise cost-efficient quality outcomes
(hours or days) before being reviewed
after hip and knee replacement sur­
by rehabilitation services. Further, if
gery.15 This could reduce hospital
surgery occurred on a Thursday or
costs, im prove patient satisfaction
Friday, patients had m inim al access
w ith care and potentially reduce
to physiotherapy over the weekend. ERAS = enhanced recovery a fte r surgery.
perioperative morbidity. * P < 0 .0 01 . ♦
These aspects offer opportunities for
improvement. The challenge is clear: We demonstrated that an ERAS pro­
to convert improvement in care (and gram for orthopaedic joint replace­ We clearly dem onstrated th at we
outcome) into shorter hospital stay. ment can be achieved. We markedly could successfully im plem ent a
O ur p rim ary end point w as d u ra­ im proved m ost indicators of pro­ p re d e fin e d ERAS p ro g ram for
tion of hospital stay. As we have cesses related to an ERAS program. orthopaedic surgical patients in
done previously,14 we used a log- These included preadmission patient public hospitals, and that doctors
transform ation and com pared geo­ education, reduced fasting times, and nurses could follow such a regi­
metric means to account for skewness clear oral fluids, w ritten in stru c­ men to improve outcome parameters.
of our data (a sm all p roportion of tions (including expected day of dis­ There was high uptake of nearly all
patients staying in hospital for very charge), less blood loss, better pain ERAS items. This led to clinically
long times distorts central tendency relief, earlier ambulation and better im p o rtan t im provem ents in care,
— a well know n phenom enon for overall quality of recovery. Similar a small reduction in hospital stay
m any types of surgery). A second­ success has been reported in other for knee replacement patients, and
ary non-param etric com parison of countries.8,17-18 Medical teams can be an overall im provem ent in some
median stays was not statistically sig­ trained to deliver an ERAS program aspects of patient satisfaction. There
nificant. Therefore, we reported the and this clearly improves the quality was evidence of improved quality of
75th centiles to illustrate the observed of care. recovery. There was no effect on most

M JA 202 (7) ■ 20 A p ril 2015

367
R e s e a rc h

complications and no adverse effect transfusion was reduced from 23% to an ERAS program, we did not under­
on hospital readmission rates. 9.8% (P< 0.001). There was a trend of take any health costing analyses.
a reduced rate of 30-day myocardial
O ur 6-week complication rates are We found a high level of general
infarction (from 0.8% to 0.5%; P = 0.2)
lower than those reported in m ost acceptance and uptake of the ERAS
and stroke (from 0.5% to 0.2%; P = 0.2).
studies, the largest of which included program and, on the whole, it had a
There was no measurable effect on
4500 consecutive unselected hip and positive effect on patients and staff.
deep vein thrombosis (0.8% v 0.6%;
knee replacem ents.5 In that study, We can recommend this orthopaedic
P =0.5) or pulmonary embolism (1.2%
the first 3000 patients represented ERAS pathway.
v 1.1%; P = 0.9).
existing or traditional practice and a
further 1500 patients underw ent an There were some limitations of our Acknowledgements: We thank the research and clinical
staff o f each of the three hospitals.
ERAS protocol similar to that used study. It was not a randomised trial
in our study. This group reported a and there may have been some imbal­ Competing interests: This project received funding from
the Victorian Department of Health (Chris Potter, Senior
decrease in length of stay from 6 days ance betw een the groups that we Policy Officer). This included a payment to Paul Myles for
to 3 days (P < 0.001), as well as a reduc­ did not account for. The study was protocol development, analysis and writing of a report.

tion in 30-day m ortality (from 0.5% unblinded and we compared numer­ Received 22 Apr 2014, accepted 22 Oct 2014. ■
to 0.1%; P = 0.02) and 90-day mortality ous secondary end points. Although
(from 0.8% to 0.2%; P = 0.01). Blood there are likely to be cost benefits of References are available online at www.mja.com.au.

C o m m e n ts
Share ■ D iscuss • P articipate • D ebate ■ Inform

This issue's suggested co m m e n t question is:

What are your views on the immunisation exemption
conscientious objector forms?
See page 351 fo r th is issue’s com m ents.

Readers are invited to e m a il us a brief co m m e n t (no

m ore than 100 w ord s) on any current health -re la te d
m atter, such as th e sta te o f our hospitals, ju n io r doctors, Applying for NHMRC project grants =
Indigenous health, general practice, m e n ta l health, INSANITY
Medicare, rural m edicine, drugs, em ergency de p a rtm e n ts,
obesity or general MJA them es.

The best published MJA C om m ent each m o n th , as judged Patients should be asked for their views
by th e Editor-in-C hief, w ill receive $ 5 0 .0 0 . on all aspects of end-of-life care including
organ and tissue donation.
P ublication is a t th e E ditor’s discretion; you w ill be n otified
if your co m m e n t is selected. C o m m e n ts m ay be edited
and w ill no t be sent back to th e a u th o r fo r approval.
Doctors across all specialties must take time
Please include your fu ll name, discipline and state
to familiarise themselves with regulations
o f residence, and disclose any relevant info rm a tio n
(especially permits) regarding the prescription
or affilia tio n s th a t m ay a ffe ct inte rp re ta tio n o f your of the entire list of Schedule 8 drugs.
com m ents.

If you have a q u e stio n /to p ic you w o u ld like your

colleagues to co m m e n t on, please e m a il us. Given the high risk of contamination by body fluids
associated with deliveries and surgeries it is extremely
Email your comments to:
difficult to fully protect staff from infection.
[email protected]

MJA 202 (7) • 20 April 2015

368
Copyright of Medical Journal of Australia is the property of Australasian Medical Publishing
Company and its content may not be copied or emailed to multiple sites or posted to a listserv
without the copyright holder's express written permission. However, users may print,
download, or email articles for individual use.