Rationale and impact
- Shared decision making and information for people offered hip, knee or shoulder replacement
- Decision aids for elective joint replacement
- Preoperative rehabilitation
- Anaesthesia and analgesia for hip replacement
- Anaesthesia and analgesia for knee replacement
- Anaesthesia and analgesia for shoulder replacement
- Tranexamic acid to minimise blood loss
- Preventing infections
- Avoiding implant selection errors
- Partial and total knee replacement
- Patella resurfacing
- Surgical approaches for primary elective hip replacement
- Shoulder replacement for osteoarthritis with no rotator cuff tear
- Shoulder replacement for pain and functional loss for people with a previous proximal humeral fracture
- Inpatient rehabilitation
- Outpatient rehabilitation
- Long-term care
These sections briefly explain why the committee made the recommendations and how they might affect practice. They link to details of the evidence and a full description of the committee's discussion.
The committee agreed that decision aids can be a useful way of helping people offered joint replacement surgery understand their options and make decisions about their care. Evidence from studies of decision aids for joint replacement showed that their content varied widely, and this led the committee to question what the components of a decision aid should be. Their view is that a decision aid should not simply be a means of providing information but should actively help people to participate in making decisions about their care. Because of the wide variation in the decision aids used in the studies, it was not possible to compare them with each other. The committee were therefore unable to recommend any particular decision aid for joint replacement. To investigate the question of what defines a decision aid for elective joint replacement, the committee made a recommendation for research.
The committee agreed, based on their experience, that preoperative rehabilitation helps to prepare people for surgery, increases their ability to manage any complications of surgery, promotes understanding and engagement with postoperative rehabilitation and prepares people for life with a joint replacement.
Evidence showed that preoperative rehabilitation reduces the length of hospital stays for people having a hip or knee replacement, although this was for intensive rehabilitation programmes and was from settings where hospital stays are usually longer than in the NHS.
Based on the evidence and their own experience, the committee agreed that preoperative rehabilitation for people having hip or knee replacement should include advice on exercises before and after surgery, lifestyle and ways to maximise independence and quality of life.
There was no evidence on preoperative rehabilitation for people having shoulder replacement. The committee noted that preoperative exercises to improve muscle function in the affected limb are often severely limited by pain for people having shoulder replacement, and agreed that the benefits seen in people having hip and knee replacement might not apply to those having shoulder replacement. They included shoulder replacement in their recommendation for research on preoperative rehabilitation.
Current practice varies widely, ranging from no preoperative rehabilitation to comprehensive individualised preoperative rehabilitation programmes. However, most services offer preoperative rehabilitation advice to everyone having hip or knee replacement, so this recommendation is not expected to lead to a substantial change in practice. For some services, providing information, exercise and lifestyle advice may increase the time needed from the orthopaedic multidisciplinary team. However, this cost can be expected to be offset by reductions in the length of hospital stays.
Evidence showed that regional and general anaesthesia are equally effective for people having hip replacement surgery, so the committee recommended that a choice should be offered. There was no evidence to support a combination of regional and general anaesthesia.
Based on their experience, the committee agreed that using local or regional analgesic techniques in combination with regional or general anaesthesia reduces postoperative pain. Clinical evidence showed that, when used with general or regional anaesthesia, both local infiltration analgesia (LIA) and nerve blocks are beneficial and there was no clinical evidence to suggest that either is more beneficial than the other. However, the committee noted that some nerve blocks impair motor function and agreed that these should be avoided in hip replacement because they can delay mobilisation.
Economic evidence showed that LIA is cost effective. For nerve blocks, the evidence on costs suggested that they are cost effective if they do not delay surgery by more than 5 minutes.
All orthopaedic units currently offer a choice of general or regional anaesthesia. Most augment this with either LIA or a nerve block. Although the cost of nerve blocks varies, it is not expected that services currently offering LIA will change to nerve blocks. This recommendation is unlikely to lead to significant changes in practice.
Evidence showed that regional and general anaesthesia are equally effective for people having knee replacement surgery, so the committee recommended that a choice should be offered. There was no evidence to support a combination of regional and general anaesthesia.
Evidence also showed that adding local infiltration analgesia (LIA) or a nerve block to regional or general anaesthesia is beneficial, and that adding both LIA and a nerve block to regional anaesthesia is more beneficial than adding either on its own, although this benefit was less pronounced with general anaesthesia.
The committee noted that some nerve blocks impair motor function and agreed that these should be avoided because they can delay mobilisation. Adductor canal nerve blocks allow for a better range of movement sooner after surgery and are now more commonly used than femoral blocks, but the evidence on them was limited.
Economic evidence showed that LIA is cost effective. For nerve blocks, the evidence on costs suggested that they are cost effective if they do not delay surgery by more than 5 minutes. Because of the uncertainty about the cost effectiveness of nerve blocks added to LIA to augment anaesthesia in knee replacement, the committee made a recommendation for research.
In current practice, regional anaesthesia for knee replacement surgery is usually augmented with LIA, a nerve block, or both. The recommendation might lead services that currently augment anaesthesia with nerve blocks, either together with LIA or on their own, to change to augmenting with LIA only or to arrange services so that administering the nerve block does not delay surgery. Services that currently augment anaesthesia with an LIA only are not expected to see a substantial change in practice.
There was not enough evidence to support recommendations on specific types of anaesthesia for shoulder replacement. Because of this uncertainty, the committee stressed the importance of discussing the options with people having this type of joint replacement. Although small benefits were seen in studies combining general anaesthesia with LIA and regional anaesthesia with LIA, they were offset by phrenic nerve palsy events. The committee made recommendations for research on supplementary analgesia or anaesthesia and regional, general, or regional with general anaesthesia in shoulder replacement. They noted that using regional anaesthesia alone has the potential to increase day-case shoulder replacement surgery.
This recommendation is not expected to change current practice.
Good evidence showed that, in people having primary elective hip or knee replacement, topical (intra-articular) tranexamic acid in combination with intravenous tranexamic acid reduces the number of blood transfusions needed when compared with topical or intravenous tranexamic acid alone. Although 1 study suggested that combining topical with oral tranexamic acid is the most clinically and cost-effective administration method, this evidence was not strong enough to support a recommendation for this combination.
Evidence in people having primary elective shoulder replacement also showed a benefit from tranexamic acid but did not address combined administration. However, the committee reasoned that the benefits seen in hip and knee replacement could also apply in shoulder replacement. They agreed that, although there may not be the same benefits in terms of reduced blood transfusions in shoulder replacement surgery, tranexamic acid reduces bleeding, which lessens fatigue and nausea. The committee noted that tranexamic acid is an inexpensive treatment.
The BNF advises a reduced dose of intravenous tranexamic acid for people with renal impairment. Because the absorption is uncertain when tranexamic acid is administered topically, the committee agreed that it should be given only intravenously to people with renal impairment.
Although the use of tranexamic acid is widespread in current practice, the method of administration varies. In the committee's experience, topical (intra-articular) tranexamic acid is commonly used in combination with intravenous tranexamic acid in hip and knee replacements, but not in shoulder replacements. Increased use of this combination in shoulder replacements might increase doses and the use of disposables. However, the associated costs are expected to be more than offset by the savings produced by a reduced need for blood transfusions in hip and knee replacements and reduced bleeding in shoulder replacements.
No evidence was found on adding antibiotic or antiseptic agents to saline wound wash-out solution to reduce surgical site infections in people having primary elective joint replacement. The committee acknowledged that washing the wound with saline is common practice and is used to improve visibility of the operative site for the surgeon. They noted that the use of antibiotic and antiseptic agents in wash-out solutions varies across the NHS. They were concerned about the risk of increasing antimicrobial resistance through the use of these agents. They agreed that, because of this risk, other means of preventing infection in joint replacement surgery, such as prophylactic antibiotics and ultra-clean air ventilation in operating theatres, should be used, and included a cross reference to the NICE guideline on surgical site infections.
There was little good evidence on the use of ultra-clean air ventilation in operating theatres. Evidence from randomised controlled trials supported ultra-clean air ventilation, but these trials may not fully reflect current practice. Evidence from observational studies supported conventional air ventilation systems, but it was unclear whether these studies followed up participants for more than 2 years, which the committee agreed is the minimum follow‑up period needed to produce an accurate picture of infection rates. It was also unclear whether the registry data used in the studies produced an accurate record of the number of infections over the longer term, and whether prophylactic antibiotics were used in all of the observational studies.
Although the committee noted the limitations in the evidence, they agreed that ultra-clean air ventilation is likely to be more effective at reducing surgical site infections than conventional turbulent air ventilation. They agreed that patient safety is the primary consideration and that infection after a joint replacement is a serious complication. Because of this, and given the uncertainty of the evidence, the committee agreed to recommend that current practice be maintained.
These recommendations are expected to reduce the routine use of antibiotic or antiseptic agents in wash-out solutions. They are not expected to affect the use of prophylactic antibiotics and ultra-clean air ventilation in operating theatres, which are current practice.
The committee's recommendations were based on their experience and expertise. They reasoned that 2 'stop moments', when theatre staff stop other activity and formally inspect each implant component, would ensure that all components are compatible. The second stop moment provides an extra opportunity to correct an implant selection error before closure. The committee agreed that intraoperative real-time data entry could be considered as a further means of ensuring that mismatched components are identified before implantation. They also agreed that technological solutions might help avoid implant selection errors and made a recommendation for research.
Intraoperative 'stop moments' to check implant components before implantation are common and are not expected to change current practice. Intraoperative real-time data entry is not current practice and, if implemented, is likely to increase theatre time.
One large trial showed no difference between partial and total knee replacement in quality of life, patient-reported outcome measures or revision surgery at 5 years. Partial knee replacement was shown to be more cost effective. However, the trial only reported outcomes after 5 years and the committee were aware that the benefits of total knee replacement may be seen after this. Two studies that compared partial with total knee replacement had limited relevance because they looked at implants that are no longer in use or were restricted to people who had both knees replaced.
The committee agreed that there are advantages and disadvantages to both procedures and this was broadly supported by the evidence. The outcomes for each type of surgery are thought to be similar although recovery after partial knee replacement tends to be faster, with a shorter stay in hospital and less pain during the recovery period. Complications such as infections, blood clots, heart attacks or stroke are rare for both procedures, but are thought to be rarer after partial than total knee replacement. Partial knee replacement leaves more of the original knee intact, but the remaining parts of the knee could develop arthritis and may need to be replaced in the future. The committee were aware that National Joint Registry data indicate a greater likelihood of revision surgery within 10 years after partial knee replacement.
The committee agreed, based on their experience, that different people weigh these relative risks or benefits differently, depending on their personal circumstances and preferences. They therefore recommended that people should be offered a choice of partial or total knee replacement after a discussion of the benefits and risks of each.
This recommendation may result in an increase in the number of partial knee replacements undertaken. It is expected that all orthopaedic services will need to provide both partial and total knee replacement surgery. The committee noted that total and partial knee replacement are very different types of procedure, and surgeons need to ensure they perform a sufficient number of each procedure every year to ensure good surgical outcomes.
Total knee replacements make up the majority of current practice, so offering a choice of partial or total knee replacement is likely to increase the number of partial knee replacements. The economic evidence largely suggested that partial knee replacements are cost effective compared with total knee replacements. Therefore, increasing the proportion of partial knee replacements is likely to be cost saving.
The committee looked at 3 options: resurfacing, no resurfacing and selective resurfacing. There was not enough clinical evidence to indicate whether any of the options was more beneficial than the others. However, strong economic evidence showed that resurfacing is cost effective compared with no resurfacing over a 10‑year time horizon because of reduced hospital readmissions. Because of the lack of clinical evidence, the committee also made a recommendation for research on selective resurfacing in knee replacement.
Current practice varies, with resurfacing carried out in around 35% to 40% of knee replacements. This recommendation can be expected to increase the number of knee replacement operations with patella resurfacing. There may be an initial increase in costs because of more costly hospital stays for resurfacing, but this is expected to be more than offset by reduced numbers of hospital readmissions in the long term.
The committee looked at evidence on 5 surgical approaches for hip replacement: posterior, anterolateral, direct anterior, direct superior and supercapsular percutaneously assisted (SuperPATH). The evidence did not indicate that any of these approaches was more beneficial than any other. The National Joint Registry for 2017 reported that 97% of hip replacements were done using the posterior or anterolateral approach. They concluded that either of these 2 established approaches could be considered, with the choice of approach based on the knowledge and experience of the surgeon and individual patient characteristics.
There was limited evidence on the newer approaches (direct anterior, direct superior and SuperPATH) and the committee made a recommendation for research to investigate these approaches.
The recommendation reflects most current practice and is not expected to lead to substantial changes.
Evidence showed that conventional total shoulder replacement provides more overall benefit than humeral hemiarthroplasty. The recommendation is limited to people with adequate glenoid bone because this is necessary for conventional total shoulder replacement to be considered. For people without adequate glenoid bone, another solution, such as reverse shoulder replacement or other surgery, is needed.
The committee agreed that the type of implant should not be specified in the recommendation but should be part of shared decision making between the person having surgery and the surgeon.
Conventional total shoulder replacement is increasingly being offered to people aged under 60 as confidence grows in its long-term durability. There is a lack of evidence in this age group, so the committee made a recommendation for research to compare conventional total shoulder replacement with humeral hemiarthroplasty.
The committee were unable to make a recommendation for practice on reverse total shoulder replacement in this context because of the lack of evidence and their uncertainty about its effectiveness compared with other procedures. The committee noted that although reverse total shoulder replacement was originally designed for people with a rotator cuff tear, it is being used more widely for people with no rotator cuff tear to obviate the need for early revision surgery after rotator cuff failure. The committee made a recommendation for research to compare reverse total shoulder replacement with conventional total shoulder replacement.
The recommendation reflects most current practice and is not expected to lead to substantial changes.
Shoulder replacement for pain and functional loss for people with a previous proximal humeral fracture
The committee looked at 3 types of procedures for people with a previous proximal humeral fracture: reverse total shoulder replacement, humeral hemiarthroplasty and conventional total shoulder replacement. They were unable to make recommendations for practice because of a lack of evidence. They made a recommendation for research on procedures for shoulder replacement for people with a previous proximal humeral fracture.
Evidence in people who have had primary elective hip or knee replacement showed that rehabilitation within 24 hours of surgery, including mobilisation, reduces length of hospital stays. The committee agreed that early discharge improves quality of life and is likely to be cost saving. They acknowledged concern about increased pain from early mobilisation, but noted the evidence showing that, for most people, the benefits outweigh any adverse effects.
The committee agreed that the first contact with the person should be made or led by a physiotherapist or occupational therapist who can assess whether the person is medically unwell or has specific needs. They may delay rehabilitation if clinically necessary. The committee agreed that some aspects of rehabilitation can be provided by a member of the physiotherapy or occupational therapy team with suitable training and support.
In the committee's experience, rehabilitation, including mobilisation, is best started on the day of surgery, but they acknowledged that this is not always possible if the operation is finished late in the day.
There was no evidence on inpatient rehabilitation after shoulder replacement. However, in the committee's experience, the benefits are similar to those seen after hip or knee replacement. They agreed that people who have had shoulder replacement should ambulate within 24 hours of surgery, but mobilisation of the shoulder depends on the orthopaedic team's clinical assessment. The committee noted that the timing of shoulder mobilisation varies widely in practice, with some services advising use of a sling for 10 days and others advising it for 6 weeks. There was no evidence available on when the shoulder should be mobilised so the committee made a recommendation for research.
The recommendation largely reflects current practice and is not expected to result in substantial changes. Starting inpatient rehabilitation within 24 hours of surgery might mean that some hospitals will need to reorganise or increase physiotherapy and occupational therapy services to ensure they are available throughout weekends for people who have surgery on a Friday or Saturday. Most hospitals will already have physiotherapy or occupational therapy staff present at weekends; however, in some hospitals they may not be seeing elective hip and knee replacement patients as part of current practice. For those hospitals that do need to take on additional staff, these costs are expected to be offset by a reduction in the length of hospital stays.
The committee agreed that outpatient rehabilitation after hip, knee or shoulder replacement is essential. Evidence suggested that, for people who have had hip or knee replacement, self-directed rehabilitation and supervised rehabilitation are similarly effective. Compared with self-directed rehabilitation, supervised rehabilitation is very costly. The committee agreed that, in their experience, self-directed rehabilitation is effective for most people after hip or knee replacement if undertaken with advice, and ongoing support if needed, from the physiotherapy or occupational therapy team.
There was no evidence to enable the committee to compare self-directed with supervised outpatient rehabilitation for people who have had shoulder replacement, so they recommended that advice may be given on either self-directed or supervised rehabilitation. They also made a recommendation for research.
The committee agreed, based on their experience, that provision needs to be made for people with additional needs that make self-directed outpatient rehabilitation difficult or who find that it is not meeting their rehabilitation goals, and who would benefit from supervised group or individual rehabilitation. They noted the lack of evidence in this area and made a recommendation for research on supporting rehabilitation for people with additional needs.
Although the proportions of people having self-directed or supervised rehabilitation after elective joint replacement are not known, it is likely that the recommendations will increase the proportion having self-directed rehabilitation and decrease rehabilitation costs.
There was no evidence available to inform recommendations on long-term follow‑up and monitoring after joint replacement surgery. The committee were aware of an ongoing study to investigate follow‑up after hip and knee replacement surgery. That study does not include people who have had shoulder replacement, so the committee made a recommendation for research on follow-up after shoulder replacement.
The committee agreed that, in the absence of recommendations on follow‑up and monitoring after hip, knee or shoulder replacement surgery, a recommendation is needed to ensure that people who have problems with their joint replacement are referred to an orthopaedic surgical service. Primary care practitioners are expected to use their clinical judgement to determine the urgency of the referral.
The recommendation reflects current practice and is not expected to result in changes.