Professional Documents
Culture Documents
doi: 10.1093/bja/aew362
Review Article
Abstract
Enhanced recovery after surgery (ERAS) protocols produce significant clinical and economic benefits in a range of surgical
subspecialties. There is a long tradition of applying clinical pathways to the perioperative care of joint arthroplasty patients.
Enhanced recovery after surgery represents the next step in the evolution of standardized care. To date, reports of full ERAS
pathways for hip or knee arthroplasty are lacking. In this narrative review, we present the evidence base that can be usefully
applied to constructing ERAS pathways for hip or knee arthroplasty. The history and rationale for applying ERAS to joint
arthroplasty are explained. Evidence demonstrates improved outcomes after joint arthroplasty when a standardized ap-
proach to care is implemented. The efficacy of individual ERAS components in hip or knee replacement is considered,
including preoperative education, intraoperative anaesthetic techniques, postoperative analgesia, and early mobilization
after joint arthroplasty. Interventions lacking high-quality evidence are identified, together with recommendations for fu-
ture research. Based on currently available evidence, we present a model ERAS pathway that can be applied to perioperative
care of patients undergoing hip or knee arthroplasty.
Enhanced recovery after surgery (ERAS) is rapidly becoming fa- consistently been associated with superior recovery, decreased
miliar to anaesthetists. Perioperative programmes incorporating morbidity, reduced hospital length of stay, and cost savings.2
multimodal, evidence-based interventions have come to be The clinical and economic gains that can be achieved with ERAS
known as fast-track or ERAS pathways. Enhanced recovery after have been demonstrated across a range of surgical specialties,
surgery was pioneered by Henrik Kehlet in the 1990s as an effort including gynaecological oncology, urology, vascular, and thor-
to improve recovery after colorectal surgery.1 Kehlet hypothe- acic surgery.3
R
sized that surgical stress, metabolic and endocrine derange- According to the ERASV Society, there are 20 components of
ments, and prolonged immobilization contribute to organ care that influence the stress response and enhance recovery. It
dysfunction: pain, nausea, vomiting, ileus, fatigue, and cogni- can quickly be appreciated that designing and implementing an
tive disturbance. The extent of organ dysfunction thus deter- ERAS protocol is a multidisciplinary endeavour. Accordingly, an
mines recovery. According to ERAS concepts, it is unlikely that ERAS team should comprise expertise from surgery, anaesthesia,
any single surgical technique, anaesthetic intervention, or nursing, physical therapy, and nutrition. The team has primary
medication can significantly reduce organ dysfunction. Rather, responsibility for reviewing the available literature and for for-
recovery is facilitated by a multimodal approach directed to- mulating and delivering the protocol. Once implemented, a key
wards global modulation of the surgical stress response. plank in any ERAS platform is audit, with continuous evaluation
Enhanced recovery after surgery in colorectal surgery has of component efficacy and institutional compliance (Fig. 1).
C The Author 2016. Published by Oxford University Press on behalf of the British Journal of Anaesthesia.
V
All rights reserved. For Permissions, please email: journals.permissions@oup.com
iii62
ERAS for hip and knee arthroplasty | iii63
AUDIT
Fig 1 A recommended protocol for ERAS for total joint arthroplasty. This diagram highlights the multimodal, multidiscliplinary nature of ERAS protocols. Audit is
required and is a key driver of continuous evaluation and refining of the components of care. CSE, combined spinalepidural; ERAS, enhanced recovery after sur-
gery; LIA, local infiltration anaesthesia; NSAID, non-steroidal anti-inflammatory; PNB, peripheral nerve block; PONV, postoperative nausea and vomiting.
iii64 | Soffin and YaDeau
Table 1 Features distinguishing clinical pathways from ERAS protocols. The ERAS protocols encompass the entire perioperative phase, in-
clude a broad, coordinated care team, and emphasize assessing the delivery of best practice. ERAS, enhanced recovery after surgery
6316 patients.6 Compared with standard care, patients on a clin- attempt to present major areas of disagreement and suggest
ical pathway had significantly fewer postoperative complica- reasonable ways forward. Future audit and research could prof-
tions [specifically, deep vein thrombosis (DVT), pulmonary itably investigate ERAS protocols for joint arthroplasty based on
embolism (PE), requirement for manipulation, superficial and these components, both to document outcomes and to refine
deep infections, and decubitus ulcers], shorter hospital length of the choice of components.
2 h-clear/6 h-solids fast is also upheld in obese and morbidly A body of evidence demonstrates a higher incidence of post-
obese cohorts,30 and in patients with uncomplicated diabetes operative complications when primary hip or knee arthroplasty
mellitus.31 A more conservative fast is recommended for dia- is performed under general compared with neuraxial/regional
betic patients with gastroparesis.32 In addition to liberalized anaesthesia. Pulmonary compromise, pulmonary embolus,
fasting guidelines, ERAS protocols recommend that patients need for transfusion, renal injury, infection, length of stay, and
consume up to 300 ml of a clear carbohydrate-rich drink 23 h 30 day mortality are all significantly lower after neuraxial anaes-
before surgery, with the goal of presenting the patient for sur- thesia.4244 Others suggest that outcomes are equivalent be-
gery in a metabolically fed state.29 A patient in an anabolic tween techniques.45 46 These results are difficult to reconcile
state undergoes less postoperative nitrogen and protein loss33 because there are multiple sources of potential bias. Registry-
and maintains more lean muscle mass.34 These concepts are based data are subject to selection bias, and older, outdated
supported by several randomized controlled trials, indicating modes of general anaesthesia could contribute to the appear-
accelerated recovery after colorectal surgery when patients re- ance of neuraxial superiority. In contrast, randomized con-
ceive preoperative carbohydrate loading.35 36 trolled trials (and the resultant meta-analyses) have their own
There are limited data specifically exploring the role of meta- sources of bias, including difficulty in generalizing results from
bolic state and recovery after joint arthroplasty. In a recent pilot (often) healthy study subjects to the much less selected patients
study, 32 patients undergoing primary hip arthroplasty were found in everyday clinical practice.
randomized to usual care or a package of nutritional interven- A recent study comparing total i.v. anaesthesia vs spinal an-
tions, including preoperative supplementation and carbohy- aesthesia for total knee arthroplasty found a 6 h reduction in
drate drink.37 Length of hospital stay and C-reactive protein
peripheral nerve block, acetaminophen, and non-steroidal anti- The role of additional oral adjuvants for multimodal anal-
inflammatory drugs feature prominently. General anaesthesia gesia is unclear. A meta-analysis found opioid sparing from
and subsequent monotherapy with i.v. opioids are not corre- gabapentin for knee arthroplasty.81 The anticonvulsant prega-
lated with optimal analgesia after total knee replacement.56 balin has been studied, with mixed evidence that perioperative
Non-steroidal anti-inflammatory drugs57 58 and acetamino- pregabalin reduces opioid consumption and prevents chronic
phen59 60 are opioid sparing and promote efficient postoperative neuropathic pain after total knee82 83 or total hip arthroplasty.84
analgesia, but on their own are typically insufficient. There are However, in conjunction with an epidural, adductor canal block,
several reasonable choices for initial postoperative analgesic and multimodal oral analgesia after knee arthroplasty, pregaba-
therapy: patient-controlled epidural analgesia, peripheral nerve lin was associated with increased sedation, lower patient satis-
block (either single injection or continuous), and peri-articular faction, and no benefit on pain scores.85 There is recent interest
injection/local infiltration analgesia. All three of these choices in gabapentinoids as adjuncts to prevent PONV after joint
have ardent advocates; some institutions favour combinations arthroplasty by virtue of the opioid-sparing effect. Some meta-
of all. Unfortunately, many studies compare an advanced anal- analyses show that gabapentinoids reduce nausea after hip83 or
gesic modality with systemic opioids; direct head-to-head com- knee86 replacement, although there was no effect on the inci-
parisons are less common.61 dence of vomiting in either study. Another meta-analysis failed
Excellent analgesia can be achieved by continuous epidural to find any protective PONV effect of gabapentin after knee
local anaesthetic with or without opioids after knee or hip re- arthroplasty.87 These benefits need to be balanced against the
placement.62 It is important to avoid high concentrations of epi- risk of sedation or dizziness from gabapentinoids.
replacement seems to be a less effective strategy and did not for reduced postoperative infection, organ complications, and
prevent intraoperative hypothermia in hip or knee risk of blood transfusion in the GDFT group compared with con-
replacement.102 trol. Unlike other surgical subspecialties, in elective joint re-
One argument against maintaining normothermia is the placement the role, type, and volume of i.v. fluids have not been
possible neuroprotective effect of mild hypothermia. Short- investigated. The effects of i.v. fluid on postsurgical outcomes
term postoperative cognitive impairment has been associated are likewise unknown. These topics are ripe for future
with warmer temperatures in elderly patients after knee re- investigation.
placement.103 The long-term effects of mild hypothermia for
joint replacement, who might benefit, and how to risk stratify
patients are intriguing avenues for research. In the interim, Blood conservation
there are sufficiently described benefits of normothermia to rec-
A conservative blood management strategy is crucial to the suc-
ommend a normothermic goal as part of ERAS protocols.
cess of an ERAS programme. Allogeneic blood transfusion is
associated with immunomodulation and volume overload.111
112
On the contrary, transfusion and anaemia are both associ-
Antimicrobial prophylaxis ated with increased incidence of infection, increased length of
Infection after joint arthoplasty is a serious complication that stay, and higher mortality after joint arthroplasty.111113 Risk
can be difficult to treat.104 There is currently no defined factors associated with allogeneic transfusion after knee re-
guideline for antibiotic/antiseptic prophylaxis for joint replace- placement include age >75 yr, male sex, hypertension, and
death as the primary outcomes of importance after hip or knee improvements seen in other surgical subspecialties, further
replacement. Some authors have criticised the premise that evaluation and adoption of ERAS pathways should be a priority
venographic evidence of DVT is equivalent to symptomatic, for joint arthroplasty care teams.
fatal PE, because there does not appear to be a direct correlation
between DVT and PE.140 For example, DVT is two- to three-fold
more common after knee replacement compared with hip re-
Authors contributions
placement, yet total knee arthroplasty is associated with a E.M.S. and J.T.Y. contributed equally to conceiving the idea, col-
reduced incidence of PE, compared with total hip arthro- lecting data, and drafting and revising the manuscript.
plasty.141 Additionally, all-cause mortality is higher in patients
taking potent anticoagulants (LMWH or DOAC) than in patients
receiving aspirin or warfarin, and clinically significant PE
Declaration of interest
occurred despite the use of anticoagulants.142 None declared.
Other data suggest that DOACs might be superior to LMWH
after hip or knee arthroplasty. For example, RECORD (Regulation
of Coagulation in ORthopaedic surgery to prevent Deep vein
Funding
thrombosis and pulmonary embolism) consisted of four double- This work was supported by the Department of Anesthesiology,
blind, randomized studies comparing oral rivaroxaban (factor Hospital for Special Surgery, New York, NY.
Xa inhibitor) with LMWH.143 Two studies were performed in pa-
study of 4500 consecutive primary hip and knee replace- no effect on gastric fluid volume and pH in fasting and non-
ment. Acta Orthop 2011; 82: 57781 fasting obese patients. Can J Anaesth 2004; 51: 1115
13. Jrgensen CC, Kehlet H. On behalf of the Lundbeck 31. Breuer JP, von Dossow V, von Heymann C, et al.
Foundation Centre for Fast-track hip and knee replacement Preoperative oral carbohydrate administration to ASA III-IV
collaborative Group. Role of patient characteristics for fast- patients undergoing elective cardiac surgery. Anesth Analg
track hip and knee arthroplasty. Br J Anaesth 2013; 110: 2006; 103: 1099108
97280 32. Kong MF, Horowitz M. Diabetic gastroparesis. Diabet Med
14. Starks I, Wainwright TW, Lewis J, Lloyd J, Middleton RG. 2005; 22(Suppl 4): 138
Older patients have the most to gain from orthopaedic 33. Svanfeldt M, Thorell A, Hausel J, et al. Randomized clinical
enhanced recovery programmes. Age Ageing 2014; 43: 6428 trial of the effect of preoperative oral carbohydrate treat-
15. Aasvang EK, Luna IE, Kehlet H. Challenges in postdischarge ment on postoperative whole-body protein and glucose
function and recovery: the case of fast-track hip and knee kinetics. Br J Surg 2007; 94: 134250
arthroplasty. Br J Anaesth 2015; 115: 8616 34. Yuill KA, Richardson RA, Davidson H, Garden OJ, Parks RW.
16. Winther SB, Foss OA, Wik TS, et al. 1-year follow-up of 920 The administration of an oral carbohydrate-containing
hip and knee Arthroplasty patients after implementing fluid prior to major elective upper-gastrointestinal surgery
fast-track. Acta Orthop 2015; 86: 7885 preserves skeletal muscle mass postoperatively: a rando-
17. Goon AK, Dines DM, Craig EV, et al. A clinical pathway for mised clinical trial. Clin Nutr 2005; 24: 327
total shoulder arthroplasty A pilot study. HSS Jl 2014; 10: 35. Nygren JT, Thorell AL, Ljungqvist O. Preoperative oral
hip arthroplasty: a randomised controlled trial. Acta 65. Ilfeld BM. Continuous peripheral nerve blocks: an update of
Anaesthesiol Scand 2015; 59: 298309 the published evidence and comparison with novel, alter-
48. McCartney CJL, Choi S. Does anaesthetic technique really native analgesic modalities. Anesth Analg 2016; in press
matter for total knee arthroplasty? Br J Anaesth 2013; 111: 66. Barrington MJ, Olive D, Low K, Scott DA, Brittain J, Choong P.
3313 Continuous femoral nerve blockade or epidural analgesia
49. Liu SS, Buvanendran A, Rathmell JP, et al. Predictors for after total knee replacement: a prospective randomized
moderate to severe acute postoperative pain after total hip controlled trial. Anesth Analg 2005; 101: 18249
and knee replacement. Int Orthop 2012; 36: 22617 67. Elkassabany NM, Antosh S, Ahmed M, et al. The risk of falls
50. Liu SS, Buvanendran A, Rathmell JP, et al. A cross-sectional after total knee arthroplasty with the use of a femoral nerve
survey on prevalence and risk factors for persistent post- block versus an adductor canal block: a double-blinded
surgical pain 1 year after total hip and knee replacement. randomized controlled study. Anesth Analg 2016; 122:
Reg Anesth Pain Med 2012; 37: 41522 1696703
51. Van den Bosch JE, Bonsel GJ, Moons KG, Kalkman CJ. Effect 68. Shah NA, Jain NP. Is continuous adductor canal block better
of postoperative experiences on willingness to pay to avoid than continuous femoral nerve block after total knee
postoperative pain, nausea, and vomiting. Anesthesiology arthroplasty? Effect on ambulation ability, early functional
2006; 104: 10339 recovery and pain control: a randomized controlled trial. J
52. Apfel CC, Kranke P, Eberhart LH, Roos A, Roewer N. Arthroplasty 2014; 29: 22249
Comparison of predictive models for postoperative nausea 69. Machi AT, Sztain JF, Kormylo NJ, et al. Discharge readiness
81. Han C, Li XD, Jiang HQ, Ma JX, Ma XL. The use of gabapentin 98. Schmied H, Kurz A, Sessler DI, Kozek S, Reiter A. A mild
in the management of postoperative pain after total knee hypothermia increases blood loss and transfusion require-
arthroplasty: a PRISMA-compliant meta-analysis of ments during total hip arthroplasty. Lancet 1996; 347:
randomized controlled trials. Medicine 95: e3883 28992
82. Buvanendran A, Kroin JS, Della Valle CJ, Kari M, Moric M, 99. Winkler M, Akca O, Birkenberg B, et al. Aggressive warming
Tuman KJ. Perioperative oral pregabalin reduces chronic reduces blood loss during hip arthroplasty. Anesth Analg
pain after total knee arthroplasty: a prospective, random- 2000; 91: 97884
ized, controlled trial. Anesth Analg 2010; 110: 199207 100. Benson EE, McMillan DE, Ong B. The effects of active warm-
83. Mao Y, Wu L, Ding W. The efficacy of preoperative adminis- ing on patient temperature and pain after total knee arthro-
tration of gabapentin/pregabalin in improving pain after plasty. Am J Nurs 2012; 112: 2633
total hip arthroplasty: a meta-analysis. BMC Musculosk 101. Kim YS, Jeon YS, Lee JA, et al. Intra-operative warming with
Disord 2016; 17: 373 a forced-air warmer in preventing hypothermia after tour-
84. Clarke H, Page GM, McCartney CJ, et al. Pregabalin reduces niquet deflation in elderly patients. J Int Med Res 2009; 37:
postoperative opioid consumption and pain for 1 week after 145764
hospital discharge but does not affect function at 6 weeks 102. Deren ME, Machan JT, DiGiovanni CW, Ehrlich MG,
or 3 months after total hip arthroplasty. Br J Anaesth 2015; Gillerman RG. Prewarming operating rooms for prevention
115: 90311 of intraoperative hypothermia during total knee and hip
85. YaDeau JT, Lin Y, Mayman DJ, et al. Pregabalin and pain arthroplasties. J Arthroplasty 2011; 26: 13806
115. Pierson JL, Hannon TJ, Earles DR. A blood-conservation al- 131. Wei Z, Liu M. The effectiveness and safety of tranexamic
gorithm to reduce blood transfusions after total hip and acid in total hip or knee arthroplasty: a meta-analysis of
knee arthroplasty. J Bone Joint Surg Am 2004; 86: 15128 2720 cases. Transfus Med 2015; 25: 15162
116. Pola E, Papaleo P, Santoliquido A, Gasparini G, Aulisa L, De 132. Zhang H, Chen J, Chen F, Que W. The effect of tranexamic
Santis E. Clinical factors associated with an increased risk acid on blood loss and use of blood products in total knee
of perioperative blood transfusion in nonanemic patients arthroplasty: a meta-analysis. Knee Surg Sports Traumatol
undergoing total hip arthroplasty. J Bone Joint Surg Am 2004; Arthrosc 2012; 20: 174252
86: 5761 133. Yang ZG, Chen WP, Wu LD. Effectiveness and safety of
117. Cuenca JI, Garca-Erce JA, Martnez F, Cardona R, Perez- tranexamic acid in reducing blood loss in total knee
Serrano L, Mun ~ oz M. Preoperative haematinics and transfu- arthroplasty: a meta-analysis. J Bone Joint Surg Am 2012; 94:
sion protocol reduce the need for transfusion after total 11539
knee replacement. Int J Surg 2007; 5: 8994 134. He P, Zhang Z, Li Y, Xu D, Wang H. Efficacy and safety of
118. Spahn D. Anaemia and patient blood management in hip tranexamic acid in bilateral total knee replacement: a
and knee surgery. Anaesthesiology 2010; 113: 48295 meta-analysis and systematic review. Med Sci Monit 2015;
119. Sharrock NE, Mineo R, Urquhart B, Salvati EA. The effect of 21: 363442
two levels of hypotension on intraoperative blood loss dur- 135. Fu DJ, Chen C, Guo L, Yang L. Use of intravenous tranexamic
ing total hip arthroplasty performed under lumbar epidural acid in total knee arthroplasty: a meta-analysis of random-
anesthesia. Anesth Analg 1993; 76: 5804 ized controlled trials. Chin J Traumatol 2013; 16: 6776