Referral for specialist services

National Clinical Guideline Centre (UK)

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Clinical Guideline Centre (UK). Osteoarthritis: Care and Management in Adults. London: National Institute for Health and Care Excellence (UK); 2014 Feb. (NICE Clinical Guidelines, No. 177.)

A new version of this title is available

Update information: December 2020: in the recommendation on adding opioid analgesics NICE added links to other NICE guidelines and resources that support discussion with patients about opioid prescribing and safe withdrawal management. For the current recommendations, see www.nice.org.uk/guidance/CG177/chapter/recommendations.

Update information: December 2020: in the recommendation on adding opioid analgesics NICE added links to other NICE guidelines and resources that support discussion with patients about opioid prescribing and safe withdrawal management. For the current recommendations, see www.nice.org.uk/guidance/CG177/chapter/recommendations.

Osteoarthritis: Care and Management in Adults.

Show details

Contents

< Prev Next >

11Referral for specialist services

11.1. Referral criteria for surgery

11.1.1. Clinical introduction

Prosthetic joint replacement is the removal of articular surfaces from a painful joint and their replacement with synthetic materials, usually metal and plastic (although a variety of surfaces are now in widespread use including ceramic and metal). It has been successfully performed for over 40 years and is now one of the commonest planned surgical procedures performed. Over 120,000 are performed annually in the UK accounting for 1% of the total healthcare budget. It is performed in the vast majority of cases for pain which originates from the joint, limits the patient’s ability to perform their normal daily activities, disturbs sleep and does not respond to non-surgical measures. Joint replacement is very effective at relieving these symptoms and carries relatively low risk both in terms of systemic complications and suboptimal outcomes for the joint itself. Joint replacement allows a return to normal activity with many patients able to resume moderate levels of sporting activity including golf, tennis and swimming.

Successful outcomes require:

careful selection of patients most likely to benefit
thorough preparation in terms of general health and information
well performed anaesthesia and surgery
appropriate rehabilitation and domestic support for the first few weeks

For most patients the additional risk of mortality as a consequence of surgery, compared to continuing conservative treatment is small. The recovery from joint replacement is rapid with patients commencing rehabilitation the day following surgery and normal activities within 6 – 12 weeks, although knee recovery may be slower than hip; 95% of hip and knee replacements would be expected to continue functioning well into the second decade after surgery with the majority providing lifelong pain free function. However, around one in five patients are not satisfied with their joint replacements and a few do not get much improvement in pain following joint replacement.

Joint replacement is one of the most effective surgical procedures available with very few contraindications. As a result the demand from patients for these treatments continues to rise along with the confidence of surgeons to offer them to a wider range of patients in terms of age, disability and co-morbidities.

11.1.2. Methodological introduction: indications for joint replacement

We looked for studies that investigated the indications for referring osteoarthritis patients for total/partial joint replacement surgery. Due to the large volume of evidence, studies were excluded if they used a mixed arthritis population of which <75% had osteoarthritis or if population was not relevant to the UK.

Seven expert opinion papers¹^,⁹¹^,¹²⁸^,²¹³^,²⁷⁶^,³²⁹^,³⁶⁷, 1 cross-sectional study²²⁷, 1 observational study¹²¹ and 1 observational-correlation study¹⁷⁹ were found.

The 7 expert opinion papers consisted of surveys and consensus group findings from rheumatologists, orthopaedic surgeons and other clinicians and their opinions of the indications for referral for joint replacement surgery.

The cross-sectional study²²⁷ studied patients suitable for toal knee arthroplasty (TKA) and assessed their willingness to undergo TKA surgery. The observational study¹²¹ assessed criteria that surgeons used as indications for total hip arthroplasty (THA) surgery. The observational-correlation study¹⁷⁹ assessed the willingness of patients (from low-rate and high-rate surgery areas) to undergo arthroplasty.

All studies are hierarchy level of evidence 3 or 4.

11.1.3. Methodological introduction: predictors of benefit and harm

We looked for studies that investigated the patient centred factors that predict benefits and harms from osteoarthritis related surgery. Due to the large volume of evidence, studies were excluded if they used a mixed arthritis population of which <75% had osteoarthritis or if population was not relevant to the UK. Additionally, studies were categorised into groups of predictive factors and for each category, the largest trials and those that covered each outcome of interest were included.

2 cohort studies (level 2+)⁵⁸^,³³³, 2 case-control studies (level 2+)⁹^,⁴²⁶ and 20 case-series’ (level 3)⁶¹^,¹⁰⁴^,¹⁰⁷^,¹³⁵^,¹³⁷^,¹⁶⁴^,¹⁶⁸^,¹⁷⁷^,²¹¹^,²¹⁷^,²²¹^,²²²^,²³⁷^,²⁶⁶^,³⁰⁰^,³⁰³^,³⁸⁶^,³⁹⁴^,⁴⁰⁷^,⁴²³ were found focusing on factors that predict the outcome of joint replacement surgery.

The 2 cohort studies⁵⁸^,³³³ were methodologically sound and differed with respect to osteoarthritis/surgery site, trial size and follow-up time. The first cohort study⁵⁸ investigated N=100 patients who had either TKA or THR compared to N=46 controls, with a follow-up time of 6 months. The second cohort study³³³ investigated N=184 patients who had THR compared to N=2960 controls, with a follow-up time of 6 and 12 months.

The 2 case-control studies⁹^,⁴²⁶ were methodologically sound and both assessed the effect of knee replacement surgery on Knee Society Score and Survival of the prosthesis in obese and non-obese patients.

11.1.4. Evidence statements: indications for joint replacement

Age

Four studies¹^,⁹¹^,²²⁷^,²⁷⁶ looked at the effect of age on indications for surgery in knee osteoarthritis patients and found that age was associated with the decision to perform surgery.

Three studies¹²¹^,²⁷⁶^,³⁶⁷ looked at the effect of age on indications for surgery in hip osteoarthritis patients and found that age was associated with the decision to perform surgery.

1 study ¹⁷⁹ looked at the effect of age on indications for surgery in hip or knee osteoarthritis patients and found that age was associated with the decision to perform surgery.

Table 269

Effect of gender on attitudes towards surgery for OA.

Gender

Two studies⁹¹^,²²⁷ looked at the effect of gender on indications for surgery in knee osteoarthritis patients and found that gender was not associated with the decision to refer for surgery but was associated with the patient’s willingness to undergo surgery.

One study¹²¹ looked at the effect of gender on indications for surgery in hip osteoarthritis patients and found that gender was associated with priority to undergo surgery.

One study¹⁷⁹ looked at the effect of gender on indications for surgery in hip or knee osteoarthritis patients and found that gender was not associated with willingness to undergo surgery.

Gender outcome	Reference	Outcome / Effect size
Knee osteoarthritis
Patient’s willingness to undergo surgery	1 cross-sectional study²²⁷(N=26,046)	OR 0.60, 95% CI 0.49 to 0.74 Favours men (more willing)
Referral for surgery	1 study of expert opinions⁹¹( N=244 Family Physicians and N=96 Rheumatologists)	Age <55 years: 52% FP’s = less likely and 35% = more likely to refer Age >80 years: >70% of FPs who treated more patients with severe knee osteoarthritis = less likely to refer
Hip
Priority for surgery	1 observational study¹²¹ (N=74 patients, N=8 surgeons)	RR 1.41, 95% CI 1.03 to 1.91 Favours women (Higher priority)
Hip or knee
Definite willingness to undergo arthroplasty	1 observational-correlation study¹⁷⁹(N=1027)	No association

Weight/BMI

Two studies ¹^,²⁷⁶ looked at the effect of weight on indications for surgery in knee osteoarthritis patients and found that weight was associated with the decision against surgery.

Three studies¹²¹ ²¹³^,²⁷⁶ looked at the effect of weight on indications for surgery in hip osteoarthritis patients and found that obesity was associated with the decision against surgery in 2 studies but was not associated with decision for surgery in 1 study.

Table 270

Effect of weight/BMI on attitudes towards surgery for OA.

Smoking/Drugs/Alcohol

Three studies¹^,⁹¹^,²⁷⁶ looked at the effect of smoking, drugs or alcohol on indications for surgery in knee osteoarthritis patients. 2 studies found that drug and/or alcohol use was associated with the decision against surgery, however 1 study found that smoking data was insufficient to make a conclusion.

One study ²⁷⁶ looked at the effect of smoking, drugs or alcohol on indications for surgery in knee osteoarthritis patients. 2 studies found that alcohol use was associated with the decision against surgery.

Table 271

Effect of smoking/drugs/alcohol on attitudes towards surgery for OA.

Co-morbidities

Three studies¹^,⁹¹^,²⁷⁶ looked at the effect of comorbidities on indications for surgery in knee osteoarthritis patients. Overall, all 3 studies found that comorbidities were associated with the decision against surgery.

Two studies²¹³^,²⁷⁶ looked at the effect of comorbidities on indications for surgery in hip osteoarthritis patients. 1 study found that comorbidities were associated with the decision against surgery, in the second study experts were not sure about the role of comorbidities.

Table 272

Effect of comorbidities on attitudes towards surgery in OA.

Structural features

One study²⁷⁶ looked at structural features as indications for surgery in knee osteoarthritis patients and found that destruction of joint space was an indication for surgery.

Four studies ¹²¹^,³⁶⁷ ²⁷⁶ ¹²⁸ looked at structural features as indications for surgery in hip osteoarthritis patients. Overall, all 3 studies found that joint space damage/high x-ray scores were required as an indicator for surgery. 1 study found bone quality was not an indication for surgery.

Table 273

Effect of structural features on attitudes to surgery for OA.

Symptoms, Function, Global assessment, QoL

Five studies ¹²¹^,¹²⁸^,²¹³^,²⁷⁶^,³⁶⁷ looked at osteoarthritis symptoms and function as indications for surgery in hip osteoarthritis patients and found mixed results, however pain was found by most studies to be an important requirement for surgery.

Hip or Knee

One study¹⁷⁹ looked at osteoarthritis symptoms as indications for surgery in hip or knee osteoarthritis patients and found no association between WOMAC disease severity and willingness to undergo surgery.

Table 274

Effect of symptoms, function and quality of life on attitudes to surgery for OA.

Osteoarthritis Grade

Two studies ²²⁷ ⁹¹ looked at osteoarthritis grade as indications for surgery in knee osteoarthritis patients. Both studies found that patients with more severe disease were more willing to undergo surgery and were more likely to be referred for surgery.

Two studies ³²⁹ ²¹³ looked at osteoarthritis grade as indications for surgery in hip osteoarthritis patients. Both studies found that more severe disease was a more important indicator for surgery.

Table 275

Effect of grade of OA on attitudes towards surgery for OA.

Willingness

One study²²⁷ looked at willingness of knee osteoarthritis patients to undergo surgery and found that approximately one third of patients would not accept surgery if offered and they were concerned wit the risks and benefits of surgery.

One study¹⁷⁹ looked at willingness of hip or knee osteoarthritis patients in high and low-rate surgery areas to undergo surgery and found that patients in high rate arthroplasty areas were more willing to undergo surgery.

Table 276

Willingness to undergo surgery for OA.

Use of assistive devices

One study²⁷⁶ looked at the effect of usage of assistive devices by knee osteoarthritis patients on the decision to undergo surgery and found that assistived device use did not affect the decision to perform surgery.

One study²⁷⁶ looked at the effect of usage of assistive devices by hip osteoarthritis patients on the decision to undergo surgery and found that overall, assistive device use did not affect the decision to perform surgery.

Table 277

Effect of assistive devices on attitude towards surgery for OA.

Patient psychological factors (including expectations)

Three studies¹^,⁹¹^,²⁷⁶(N=13 experts) looked at the effect of psychological factors on indications for surgery in knee osteoarthritis patients and all studies found that psychological factors were important indicators affecting the decision to perform surgery.

One study²⁷⁶ looked at the effect of psychological factors on indications for surgery in hip osteoarthritis patients and all studies found that psychological factors were important indicators affecting the decision to perform surgery.

Table 278

Effect of psychological factors in attitudes towards surgery for OA.

Post-operative care and Physician advice

One study²⁷⁶ looked at the effect of home care on the decision to perform surgery in knee osteoarthritis patients and found that limited home care did not affect the decision to perform surgery.

Two studies²⁷⁶^,³⁶⁷ looked at the effect of limited home care and previous nonsurgical treatment and surgical risk on indications for surgery in hip osteoarthritis patients and found that limited home care did not affect the decision to perform surgery but previous nonsurgical treatment and surgical risk significantly affected the decision.

One study¹⁷⁹ looked at the effect of interaction with their physician on willingness to undergo surgery in patients with hip or knee osteoarthritis and found mixed results.

Table 279

Effect of postoperative care and physician advice on attitudes to surgery for OA.

11.1.5. Evidence statements: predictors of benefit and harm

11.1.5.1. Age

Knee osteoarthritis

Peri-operative complications / hospital stay

One case-series²²² (N=454) found that for TKA patients:

There was NS difference between younger and older patients for length of stay in the acute care setting or rehabilitation facilities and in-hospital complications;
Older age group were more likely to be transferred to rehabilitation facilities regardless of joint type replaced (Older patients with TKA = 83%, younger patients 40%).

One case-series⁴²³ (N=124) found that:

Older age (71–80 years or ≥81 years versus 65–70 years) was a significant predictor of AEs;
Patients at low risk of AEs included those with fewer than 2 of the following risk factors: age >70 years, male gender, 1 or more comorbid illnesses:
Age 71–80 years: OR 1.3 (95% CI 1.0 to 1.6);
Age 81–95 years: OR 1.6 (95% CI 1.1 to 2.4).

One case-series¹⁶⁴ (N=3048) found that older patients had a much higher mortality rate post TKR:

Patients aged <65 years: mortality rate 0.13% (N=1 out of N=755 patients)
Patients aged ≥85 years: mortality rate 4.65% (N=4 out of N=86 patients)
Risk ratio was 14 times higher in patients aged ≥85 years than the rest of the patients (OR 13.7, 95% CI 3.0 to 44.8).

Long-term survival of prosthesis

One case-series¹⁷⁷ (N=35, 857) found that for TKA:

Cumulative revision rate for TKA due to:
- any cause was higher in younger patients (<60 years old) than the older group (≥60 years old) at 8.5 years post-surgery (13% and 6% respectively);
- loosening of components was higher in younger patients (<60 years old) than the older group (≥60 years old) at 8.5 years post-surgery (6% and 2.5% respectively).
While for TKA patients regression analysis showed that risk for revision due to:
- any cause was significantly lower (risk ratio 0.49, 95% CI 0.38 to 0.62, p=0.0000) in the older patients (≥60 years) compared to younger patients (<60 years);
- loosening of components was significantly lower (risk ratio 0.41, 95% CI 0.27 to 0.62, p=0.0000) in the older patients (≥60 years) compared to younger patients (<60 years);
- any cause attributable to year of surgery decreased each year (risk ratio 0.92, 95% CI 0.89 to 0.96, p=0.0000) in the older patients (≥60 years) compared to younger patients (<60 years);
- loosening of components attributable to year of surgery decreased each year (risk ratio 0.87, 95% CI 0.82 to 0.94, p=0.0001) in the older patients (≥60 years) compared to younger patients (<60 years);
- infection attributable to year of surgery decreased each year (risk ratio 0.91, 95% CI 0.85 to 0.96, p=0.0015) in the older patients (≥60 years) compared to younger patients (<60 years)
- And that there was no significant difference between the older (≥60 years) and younger patients (<60 years), for risk of revision due to infection.

The same case-series¹⁷⁷ (N=35, 857) found that for unicompartmental KA cumulative revision rate due to:

any cause was higher in younger patients (<60 years old) than the older group (≥60 years old) at 9.2 years post-surgery (22% and 14% respectively);
loosening of components was higher in younger patients (<60 years old) than the older group (≥60 years old) at 9.5 years post-surgery (8% and 6.5% respectively).
Whilst regression analysis showed that for unicompartmental KA patients:
- risk for revision due to any cause was significantly lower (Risk ratio 0.55, 95% CI 0.45 to 0.65, p=0.0000) in the older patients (≥60 years) compared to younger patients (<60 years);
- risk for revision due to loosening of components was significantly lower (Risk ratio 0.63, 95% CI 0.48 to 0.83, p=0.0012) in the older patients (≥60 years) compared to younger patients (<60 years);
- there was no significant difference between the older (≥60 years) and younger patients (<60 years), for risk of revision due to infection;
- risk for revision (due to any cause) attributable to year of surgery decreased each year (Risk ratio 0.94, 95% CI 0.91 to 0.97, p=0.0001) in the older patients (≥60 years) compared to younger patients (<60 years);
- risk for revision (due to loosening of components) attributable to year of surgery decreased each year (Risk ratio 0.91, 95% CI 0.87 to 0.96, p=0.0002) in the older patients (≥60 years) compared to younger patients (<60 years);
- there was no significant difference between the older (≥60 years) and younger patients (<60 years), for risk of revision due to infection attributable to year of surgery.

Symptoms (Pain, stiffness), Function, QoL

One case-series¹³⁵ (N=512) found that:

Younger age was a predictor of poor outcome (high pain score);
Age was a significant predictor of TKR outcome:
Younger patients were significantly associated with poor outcome (high pain score), pain at 5 years post-surgery (17% aged <60 years vs 7% aged 60–64, p<0.05; 13% aged 60–70; 7% aged >70);
Patients aged <60 years are more than twice as likely to report poor outcome scores (high pain at 5 years post-surgery) than those >60 years;
Patients who had unilateral TKA (first knee) and those who had staged unilateral TKA (second knee) were significantly more likely to have poor outcome scores (high pain at 5 years post-surgery) than those who had bilateral TKA at the same time (13%, 6% and 2% respectively, p<0.01);

One case-series²²² (N=454) found that for TKA patients, age was not a strong predictor of postoperative WOMAC pain or function.

One case-series²⁶⁶ (N=860) found that older age was a strong predictor of SF-36 physical functioning at 2 years post-surgery.

One case-series¹³⁷ (N=855) found that age was:

associated with post-operative SF-36 scores and WOMAC scores.
not a predictor of post-operative SF-36 physical function, bodily pain, vitality, social functioning, role emotional, mental health, role physical
a predictors of post-operative SF-36 general health
a predictor of post-operative WOMAC pain, and stiffness
not a predictor of post-operative WOMAC function

Hip osteoarthritis

Peri-operative complications / hospital stay

One case-series²²² (N=454) found that for THA patients there was a NS difference between younger and older patients for:

length of stay in the i) acute care setting; ii) rehabilitation facilities
in-hospital complications

Whilst the older age group were more likely to be transferred to rehabilitation facilities regardless of joint type replaced.

Long-term survival of prosthesis

One case-series²²¹ (N=36, 984) found that:

Older age was associated with increased RR of failure: In patients aged ≥80 years (RR 1.6, 95% CI 1.0 to 2.6) compared with patients aged 60–69 years at 0–30 days after primary THR.
Younger age was associated with increased RR of failure: In patients aged 10 to 49 years (RR 1.7, 95% CI 1.3 to 2.3) and patients aged 50 to 59 years (RR 1.3, 95% CI 1.0 to 1.6) compared with patients aged 60–69 years. Patients aged 70–79 years and ≥80 years were associated with a lower RR for failure (RR 0.9, 95% CI 0.7 to 1.0) and (RR 0.6, 95% CI 0.5 to 0.8) respectively at 6 months to 8.6 years after primary THR.

Symptoms (Pain, stiffness), Function, QoL

One case-series²²² (N=454) found that for THA patients, age was not a strong predictor of post operative WOMAC pain or function

One case-series³⁸⁶ (N=12,925) found by linear regression that patients were an average of 1.6 years older per category of reduced pre-operative walking capacity (p<0.01; effect size 0.4), indicating that age had a moderate effect on deterioration of pre-operative walking capacity.

Thumb osteoarthritis

Symptoms (pain, stiffness), function, QoL

One case-series¹⁰⁷ (N=36) found that age at operation was not a significant predictor of surgical outcome (DASH score - Disabilities of the arm, shoulder and hand).

11.1.5.2. Gender

Knee osteoarthritis

Peri-operative complications / hospital stay

One case-series⁴²³ (N=124) found that:

Male gender was a significant predictor of AEs;

Patients at low risk of AEs included those with fewer than 2 of the following risk factors ; age >70 years, male gender, 1 or more comorbid illnesses.

Long-term survival of prosthesis

One case-series¹⁷⁷ (N=35, 857) found that for TKA there was no significant risk of TKA revision due to any cause or component loosening associated with gender.

Men were significantly more likely than women to have TKA revision due to infection (risk ratio 1.64, 95% CI 1.23 to 2.18, p=0.0007).
The same case-series¹⁷⁷ (N=35, 857) found that for unicompartmental KA there was no significant risk of revision due to any cause or component loosening associated with gender.
Men were significantly more likely than women to have unicompartmental KA revision due to infection (risk ratio 1.88, 95% CI 1.13 to 3.14, p=0.0156).

Symptoms (pain, stiffness), function, QoL

One case-series¹³⁵ (N=512) found that gender was not associated with outcome of TKR (pain at 5 years post-surgery).

One case-series¹³⁷ (N=855) found that gender was:

Associated with post-operative SF-36 scores and WOMAC scores.
A predictor of post operative WOMAC stiffness
Not a predictor of post-operative:
- SF-36 physical function, bodily pain, role physical, vitality, role emotional, mental health
- WOMAC pain.
Whilst male gender was:
- Not a predictor of post-operative SF-36 general health;
- A predictor of post-operative SF-36 social functioning and WOMAC function;
And female gender was:
- Not a predictor of post-operative SF-36 social functioning;
- A predictor of post-operative SF-36 general health.,

Hip osteoarthritis

Long-term survival of prosthesis / hospital stay

One case-series²²¹ (N=36, 984) found that:

Male gender was associated with an increased RR of THR failure of any cause (RR 1.5, 95% CI 1.1 to 2.0) at 0–30 days (RR 1.2, 95% CI 1.0 to 1.4) at 6 months to 8.6 years after primary THR
There was no association between THR failure and gender or age at 31 days to 6 months after primary THR.

Symptoms (pain, stiffness), function, QoL

One cohort study³³³ (N=3144) found that:

There was no difference between men and women for post-operative outcome (WOMAC and SF-36) at 6 months and 12 months post-THR surgery.
Gender was not associated with post-operative WOMAC pain or physical function at 12 months post-THR surgery.

Thumb osteoarthritis

Long-term survival of prosthesis

One case-series⁶¹(N=71) found that women had a higher prosthesis survival rate than men (N=7, 85% and N=4, 36% respectively).

11.1.5.3. Weight/BMI

Knee osteoarthritis

Peri-operative complications / hospital stay

One case-series³⁰³ (N=124) found that body weight ≥180 lbs was not significantly associated with symptomatic pulmonary embolism.

One case-control study⁹ (N=79) found that overall rate of complications following TKR was significantly higher in the morbidly obese group compared to the non-obese group (32% and 0% respectively, p=0.001).

Long-term survival of prosthesis

One case-control study⁴²⁶ (N=656) found that:

There was NS difference between obese and non-obese patients for percentage of revisions (4.9% and 3.1% respectively);
Revision due to osteolysis was significantly higher in the obese group compared to the non-obese group (p=0.016);
Higher BMI was associated with an increase in incidence of focal osteolysis;
Survival analysis showed NS difference for revision of any component between obese and non-obese patients (98.1% and 99.9% survival rates respectively). This similarity was maintained until the 10th year post-operatively (97.2% and 95.5% respectively).

One case-control study⁹ (N=79) found that overall rate of TKR revisions and revisions plus pain (5-year survivorship) was significantly higher in the morbidly obese group compared to the non-obese group (p=0.01 and p=0.02 respectively)

Symptoms (pain, stiffness), function, QoL

One case-series¹⁰⁴ (N=101) found that improvement in post-operative QoL was significantly greater in the obese groups compared to the non-obese group.
Two case-control studies ⁴²⁶ ⁹ found that there was NS difference between obese and non-obese patients for KSS score at the most recent follow-up for function, absolute improvement and knee scores,

One case-series¹³⁷ (N=855) found that BMI was not associated with post-operative SF-36 scores and WOMAC scores.

Hip osteoarthritis

Peri-operative complications / hospital stay

One case-series³⁹⁴ (N=3309) found that:

Increasing BMI was significantly associated with length of stay in hospital (p<0.001)
Compare with the normal weight group, mean length of hospital stay increased 4.7% in the overweight group and 7.0% in the obese group (multivariate logistic regression)
There was NS association between increasing BMI and risk of systemic post-operative complications
In the obese group, there was a 58% risk (OR 1.58, 95% CI 1.06 to 2.35) of systemic post-operative complications compared to those of normal weight.

Symptoms (pain, stiffness), function, QoL

One case-series²¹⁷ (N=78) found that:

There was no correlation between pre-operative BMI and post-operative mobility, WOMAC pain, function or other complications;

11.1.5.4. Smoking

Hip osteoarthritis

Peri-operative complications / hospital stay

One case-series³⁹⁴ (N=3309) found that:

There was NS association between smoking status or tobacco preference and the mean length of stay (after adjusting for covariates of age, BMI and so on).
Smoking status was significantly increased the risk of systemic post-operative complications (p=0.013);
Previous and current smokers had increased risks of suffering from post-operative complications compared with non-smokers (multivariate logistic regression analysis): 43% (OR 1.32, 95% CI 1.04 to 1.97) and 56% (OR 1.56, 95% CI 1.14 to 2.14) respectively
There was NS association between post-operative complications and preference for different tobacco products
Number of pack years of tobacco smoking was significantly associated with increased risk of systemic post-operative complications (p=0.004)
The heaviest tobacco smoking group was associated with a 121% (OR 2.21, 95% CI 1.28 to 3.82) increased risk of systemic complications compared to non-smokers (multivariate logistic regression analysis)
There was NS difference between smoking for:
- 0–19.9 pack years and non-smokers for risk of systemic complications
- Status, preference of tobacco product or pack years and local complications.

11.1.5.5. Co-morbidities

Knee

Peri-operative complications / hospital stay

One case-series¹⁶⁴ (N=3048) found that cardiovascular comorbidities significantly influenced mortality rate after TKR (p<0.0001). Risk of mortality associated with comorbidities was 16 times higher than when comorbidities were absent (OR 15.9, 95% CI 3.4 to 143.5).

Symptoms (pain, stiffness), function, QoL

One case-series²⁶⁶ (N=860) found that a greater number of co-morbid conditions was a strong predictor of SF-36 physical functioning at 2 years post-surgery.

One case-series¹³⁷ (N=855) found that:

Low back pain and comorbidities were associated with post-operative SF-36 scores and WOMAC scores.
Low back pain and Charlson Index were not predictors of post-operative SF-36 physical function;
Low back pain and Charlson Index were predictors of post-operative SF-36 bodily pain;
Charlson index 1 and low back pain were not predictors of post-operative SF-36 general health;
Charlson Index ≥2 was a predictor of post-operative SF-36 general health;
Low back pain and Charlson Index were not predictors of post-operative SF-36 role physical;
Low back pain and Charlson Index were predictors of post-operative SF-36 vitality;
Low back pain was not a predictor of post-operative SF-36 social functioning;
Charlson index was a predictor of post-operative SF-36 social functioning;
Low back pain and Charlson Index ≥2 were not predictors of post-operative SF-36 role emotional;
Charlson Index 1 was a predictor of post-operative SF-36 role emotional;
Gender, age and Charlson Index were not predictors of post-operative SF-36 mental health;
Low back pain was a predictor of post-operative SF-36 mental health;
Charlson Index 1 was not a predictor of post-operative WOMAC Pain;
Low back pain and Charlson Index ≥2 were predictors of post-operative WOMAC pain;
Charlson Index 1 was not a predictor of post-operative WOMAC Function;
Low back pain and Charlson Index ≥2 were predictors of post-operative WOMAC function;
Charlson Index was not a predictor of post-operative WOMAC stiffness;
Low back pain and Charlson Index were predictors of post-operative WOMAC stiffness.

Hip osteoarthritis

Peri-operative complications / hospital stay

One case-series⁴²³ (N=124) found that:

Comorbid illnesses (1 or 2+ versus none) was a significant predictor of AEs.
Patients at low risk of AEs included those with fewer than 2 of the following risk factors: age >70 years, male gender, 1 or more comorbid illnesses.

Long-term survival of prosthesis

One case-series²²¹ (N=36, 984) found that:

A high co-morbidity index score was a strong predictor of THR failure compared with a low co-morbidity index score (RR 2.3, 95% CI 1.6 to 3.5) at 0–30 days and (RR 3.0, 95% CI 2.1 to 4.5) at 31 days to 6 months after primary THR.
A medium co-morbidity index score was associated with reduced RR of failure (RR 0.7, 95% CI 0.6 to 0.8) compared to a low co-morbidity score whereas a high co-morbidity index score was a strong predictor of THR failure compared with a low co-morbidity index score (RR 2.8, 95% CI 2.3 to 3.3) at 6 months to 8.6 years after primary THR.

Symptoms (pain, stiffness), function, QoL

One case-series³⁸⁶ (N=12,925) found that co-morbidities influenced the post-operative walking capacity: there was a consistent increase in the percentage of Charnley class-C patients with each decrease in category of pre-operative walking capacity at each of the follow-up years.

11.1.5.6. Structural features

Knee osteoarthritis

Symptoms (pain, stiffness), function, QoL

One cohort study⁵⁸ (N=146) found that in TKA patients pre-operative Charnley or modified Charnley Class C was not a predictor of post-operative WOMAC function.

One case-series¹⁶⁸ (N=68) found that preoperative medial femorotibial narrowing did not influence post-operative (valgus tibial osteotomy) functional outcome at the time of last follow-up or radiographic outcome at 1 year post-surgery;

Hip osteoarthritis

Symptoms (pain, stiffness), function, QoL

One cohort study⁵⁸ (N=146) found that in THA patients, pre-operative Charnley or modified Charnley Class C was not a predictor of post-operative WOMAC function.

One case-series³⁰⁰ (N=1015) found that:

Patients with a greater degree of pre-surgery cartilage space loss had significantly less hip pain at 6 months (p=0.0016) and 1 year (p=0.0028) post-THR surgery;
There was NS association between degree of cartilage space loss and hip pain at 3, 5 and 7 years post-THR surgery;
Patients with pre-surgery superior cartilage space loss (femoral head migration) had significantly less pain at 6 months post-THR surgery (p<0.05) compared to those with mainly global or medial hip cartilage space;
There was NS association between pre-surgery osteophyte formation and post-THR pain;
There was NS association between the pre-surgery degree of cartilage space loss, direction of cartilage space loss or osteophyte formation and post-operative Harris Hip Score at 1 month, 3 months, 5 years and 7 years post-THR surgery.

Shoulder osteoarthritis

Symptoms (Pain, stiffness), Function, QoL

One case-series²¹¹ (N=154) found that:

Patients with rotator cuff tear that were treated with total shoulder arthroplasty had better postoperative active external rotation that those treated with hemiarthroplasty;
Preoperative glenoid erosion significantly affected postoperative ROM for patients with hemiarthroplasty
Patients with moderate-severe glenoid erosion treated with total arthroplasty had significantly greater increase in postoperative active external rotation compared to hemiarthroplasty (p=0.0013);
There was NS difference between total and hemi- arthroplasty patients with glenoid erosion for postoperative active forward flexion;
There was NS difference between total and hemi- arthroplasty patients with or without glenoid erosion for postoperative American Shoulder and Elbow Surgeons scores;
Degree of glenoid erosion did not affect the outcome of shoulder arthroplasty in any of the patients;
For patients treated with total or hemi-arthroplasty, there was NS difference between Shoulders with or without preoperative posterior subluxation of the humeral head for:
Post-operative American Shoulder and Elbow Surgeons scores;
Post-operative pain;
Post-operative active external rotation;
There was NS difference between total or hemi-arthroplasty patients who were without pre-operative glenoid erosion or humeral head subluxation, for postoperative American Shoulder and Elbow Surgeons scores.

Thumb osteoarthritis

Symptoms (Pain, stiffness), Function, QoL

One case-series¹⁰⁷ (N=36) found that pre-operative web angle, hyperextension of the MCP and flexion of the MCP were all significant predictors (p<0.05) of surgical outcome (DASH score - Disabilities of the arm, shoulder and hand).

11.1.5.7. Symptoms, Function, QoL

Knee osteoarthritis

Symptoms (Pain, stiffness), Function, QoL

One case-series¹³⁵ (N=512) found that pre-operative pain scores as well as mobility on stairs was a predictors of poor outcome (high pain score).

One cohort study⁵⁸ (N=146) found that in TKA patients. pre-operative WOMAC function was:

significantly associated with post-operative function (p<0.001);
a significant predictor of higher post-operative WOMAC function (OR 1.15, 95% CI 1.04 to 1.28).

One case-series²⁶⁶ (N=860) found that:

Pre-operative WOMAC pain score was a strong determinant of post-operative WOMAC pain at 1 and 2 years post-surgery;
Pre-operative SF-36 score was a strong determinant of post-operative WOMAC pain at 1 and 2 years post-surgery;
Pre-operative WOMAC function score was a strong determinant of post-operative WOMAC function at 1 and 2 years post-surgery;
There was NS difference between men and women with respect to WOMAC function at 1 year and 2 years post-surgery;
Patients with pre-operative WOMAC function in the lowest quartile (<34) had considerable functional disability after TKA (mean scores 62.1 and 59.8 for 1 year and 2 years post-surgery);
Patients with pre-operative WOMAC function in the lowest quartile (<34) had considerable functional disability after TKA (mean scores 62.1 and 59.8 for 1 year and 2 years post-surgery);
Patients with pre-operative WOMAC function in the lowest quartile (<34) had the greatest improvement in WOMAC function after TKA compared to other groups: they were over 4 times more likely (OR 4.12, 95% CI 2.86 to 6.25) to have a score of ≤60 at 2 years post-surgery than patients with pre-oeprative WOMAC function score of >35.
Pre-operative SF-36 physical functioning score was a strong predictor of SF-36 physical functioning at 1 year and 2 years post-surgery
Older age and greater number of co-morbid conditions were also strong predictors of SF-36 physical functioning at 2 years post-surgery.

One case-series²³⁷ (N=812) found that:

There was NS difference between men and women for post-operative improvement in AKS score at 5 years post-TKR
Increased age (up to 70–73 age-group) was associated with an increase in post-operative improvement in AKS score at 5 years post-TKR
Older age (>73 years) was associated with a significant decrease (p<0.05) in post-operative improvement in AKS score at 5 years post-TKR – the 79–86 year age-group showed the least improvement
Patients with the worst pre-operative AKS scores had significantly greater improvement (p<0.001) in AKS score at 5 years post-TKR compared to those with higher pre-operative AKS scores

One case-series¹³⁷ (N=855) found that pre-operative SF-36 domains for mental health and:

physical function were predictors of post-operative SF-36 physical function;
bodily pain were predictors of post-operative SF-36 bodily pain;
general health were predictors of post-operative SF-36 general health;
role physical were predictors of post-operative SF-36 role physical;
vitality were predictors of post-operative SF-36 vitality;
social functioning were predictors of post-operative SF-36 social functioning;
role emotional were predictors of post-operative SF-36 role emotional;
pre-operative WOMAC pain were predictors of post-operative WOMAC pain;
pre-operative WOMAC function were predictors of post-operative WOMAC function;
pre-operative WOMAC stiffness were predictors of post-operative WOMAC stiffness.

Hip osteoarthritis

Symptoms (pain, stiffness), function, QoL

One cohort study⁵⁸ (N=146) found that in THA patients, pre-operative WOMAC function was:

significantly associated with post-operative function (p<0.005)
a significant predictor of higher post-operative WOMAC function (OR 1.44, 95% CI 1.07 to 1.92).

One cohort study³³³ (N=3144) found that pre-operative:

pain was significantly associated with post-operative pain at 12 months (p=0.011);
physical function was significantly associated with post-operative physical function at 12 months (p<0.006).

One case-series³⁸⁶ (N=12,925) found that:

There was NS difference between the proportion of pain-free patients in any of the pre-operative pain categories
There were significant differences (p<0.01) between the pre-operative walking capacity groups with respect to post-operative walking capacity >60 minutes.
Patients with the worst pre-operative walking capacity had the worst post-operative recovery of walking capacity
Patients with the highest pre-operative walking capacity had the best post-operative walking capacity
There were significant differences (p<0.01) between the pre-operative hip flexion groups with respect to post-operative hip flexion.
Patients with pre-operative flexion ≤70° had the worst post-operative recovery of motion (flexion)
Patients with excellent range of pre-operative flexion sustained a slight loss of flexion range post-surgery.
Patients with excellent pre-operative hip ROM (flexion) were an average of 3 years older (p<0.01) than those with the poorest pre-operative ROM.

Shoulder

Symptoms (pain, stiffness), function, QoL

One case-series²¹¹ (N=154) found that:

Severity of preoperative loss of passive external rotation was found to significantly affect the postoperative range of external motion (p=0.006):
Hemiarthroplasty: patients with preoperative external rotation of <10° had mean postoperative external rotation of 25°, compared to those with pre-operative ≥10° had mean 47° postoperatively;
Total arthroplasty: patients with preoperative external rotation of <10° had mean postoperative external rotation of 43°, compared to those with pre-operative ≥10° had mean 50° postoperatively.
Preoperative internal rotation contracture did not have an adverse effect on results of total shoulder arthroplasties;
The severity of preoperative loss of forward flexion had no effect on postoperative forward flexion after either hemi- or total- arthroplasty;
Presence of full thickness repairable rotator cuff tear (isolated to the supraspinatus tendon) did not affect post-operative American Shoulder and Elbow Surgeons scores for pain or function, decrease in pain or patient satisfaction.

Thumb osteoarthritis

Symptoms (pain, stiffness), function, QoL

One case-series¹⁰⁷ (N=36) found that range of motion was not a significant predictors of surgical outcome (DASH score - Disabilities of the arm, shoulder and hand).

11.1.5.8. Osteoarthritis Grade

Hip osteoarthritis

Symptoms (pain, stiffness), function, QoL

One cohort study³³³ (N=3144) found that:

Patients with severe pre-operative radiographic osteoarthritis did not differ from the moderate osteoarthritis group with respect to post-operative SF-36 and WOMAC scores at 6 months and 12 months post-THR surgery;
Pre-operative radiographic grade of osteoarthritis was not associated with post-operative WOMAC Pain or physical function at 12 months post-THR surgery.

One case-series⁴⁰⁷ (N=147) found that:

Pre-operative Hip Grade was not associated with post-operative Harris Hip score;
Post-operative UCLA activity scores were similar for all Pre-operative Hip Grades;
Pre-operative Hip Grade influenced the amount of post-operative pain:
Mild-moderate pain was significantly less frequent at latest follow-up in Grade A hips compared to Grade B and C combined (3% and 18% respectively, p=0.03);
Pre-operative lower grade hips showed greater post-operative improvement in ROM:
Improvement in flexion, extension, abduction and external rotation were significantly greater in Grade B and C hips combined compared to Grade A (all: p<0.04).

Thumb osteoarthritis

Symptoms (pain, stiffness), function, QoL

One case-series¹⁰⁷ (N=36) found that radiographic stage was not a significant predictor of surgical outcome (DASH score - Disabilities of the arm, shoulder and hand).

11.1.5.9. Other outcomes

Knee osteoarthritis

Symptoms (pain, stiffness), function, QoL

One case-series¹³⁷ (N=855) found that social support was:

associated with post-operative SF-36 scores and WOMAC scores.
not a predictor of post-operative SF-36 physical function, bodily pain, vitality, social functioning, WOMAC stiffness
a predictor of post-operative SF-36 general health, role physical, role emotional, mental health, WOMAC pain, WOMAC function,
hospital was not associated with post-operative SF-36 scores and WOMAC scores.

Peri-operative complications / hospital stay

One case-series³⁰³ (N=124) found that:

Pre-operative Hb level ≥14 g/L was significantly associated with the development of symptomatic pulmonary embolism (p=0.011);
Bilateral TKA was significantly associated with the development of symptomatic pulmonary embolism (p≤0.05).
Pre-operative Hb level ≥14 g/L was a predictor of pulmonary embolism (OR 2.4, 95% CI 1.2 to 4.6);
Bilateral TKA was a predictor of pulmonary embolism (OR 7.2, 95% CI 1.3 to 39.6).

Thumb osteoarthritis

Symptoms (Pain, stiffness), Function, QoL

One case-series¹⁰⁷ (N=36) found that surgical procedure and hand dominance were not significant predictors of surgical outcome (DASH score - Disabilities of the arm, shoulder and hand).

11.1.6. Health economic evidence

We looked at studies that conducted economic evaluations involving referral to joint surgery for patients with osteoarthritis. One paper from New Zealand investigating 153 patients on orthopaedic waiting lists was found.¹⁴⁴ The paper investigates the waiting times for patients, and the cost incurred by the patients, as well as considering the health status of patients at different time points before and after surgery. The paper found that the cost is significantly higher for patients who wait longer than 6 months for surgery compared to patients who wait less than 6 months. However it is interesting to note that this is from a societal perspective. Costs are significantly higher for personal and societal costs for the group that waits over 6 months, but for medical costs alone the cost is higher but not statistically significantly so. The paper also finds that the health of patients generally worsens over time up until their operation, after which health improves, suggesting that the longer a patient waits the more health losses they accrue as opposed to someone who is treated more quickly.

11.1.7. From evidence to recommendations

Although demand and frequency of joint replacement continues to rise there is very little evidence upon which to base decisions about who to refer. The most effective techniques for defining criteria to guide appropriate referral have been the development of expert guided consensus. The purpose of these criteria is to quantify the benefit /risk ratio in order to inform patients and referrers of the appropriateness of treatment. However each decision remains individual and ultimately it is the patient who must decide on their own risk / benefit calculation based upon the severity of their symptoms, their general health, their expectations of lifestyle and activity and the effectiveness of any non-surgical treatments. Referral for consideration of surgery should allow all patients who may benefit to have access to a health worker, usually the surgeon, who can inform that decision.

The use of orthopaedic scores and questionnaire based assessments has become widespread. These usually assess pain, functional impairment and sometimes radiographic damage. The commonest are the New Zealand score and the Oxford Hip or Knee score. Many (such as the Oxford tools) were designed to measure population based changes following surgery, and none have been validated for the assessment of appropriateness of referral.

Similarly the use of radiographic reports as a basis for referral decisions is unreliable. This is because radiographs appearances do not correlate well with symptoms, significant painful lesions may not be detectable on plain radiographs and the radiographs are often inadequately performed eg. non-weight bearing radiographs of the knee.

The restriction of referral for consideration of surgery based upon other health issues such as BMI age or co-morbidities has no basis in evidence. There are some groups of patients for whom the risks of post-operative complication may be slightly higher or the long term outcomes of joint replacement worse but there is no evidence supporting these as reasons to deny treatment. Indeed there is evidence to suggest these patients can have greater benefit than other groups.

11.1.8. Recommendations

35.: Clinicians with responsibility for referring a person with osteoarthritis for consideration of joint surgery should ensure that the person has been offered at least the core (non-surgical) treatment options (see recommendation 6 and Figure 3 in section 4.1.2). [2008]
36.: Base decisions on referral thresholds on discussions between patient representatives, referring clinicians and surgeons, rather than using scoring tools for prioritisation. [2008, amended 2014]
37.: Consider referral for joint surgery for people with osteoarthritis who experience joint symptoms (pain, stiffness and reduced function) that have a substantial impact on their quality of life and are refractory to non-surgical treatment. [2008, amended 2014]
38.: Refer for consideration of joint surgery before there is prolonged and established functional limitation and severe pain. [2008, amended 2014]
39.: Patient-specific factors (including age, sex, smoking, obesity and comorbidities) should not be barriers to referral for joint surgery. [2008, amended 2014]

Bookshelf ID: NBK333044

Contents

< Prev Next >

PubReader
Print View
Cite this Page
National Clinical Guideline Centre (UK). Osteoarthritis: Care and Management in Adults. London: National Institute for Health and Care Excellence (UK); 2014 Feb. (NICE Clinical Guidelines, No. 177.) 11, Referral for specialist services.
PDF version of this title (8.4M)

Other titles in this collection

National Institute for Health and Clinical Excellence: Guidance

Related NICE guidance and evidence

Recent Activity

Clear Turn Off Turn On

Referral for specialist services - Osteoarthritis
Referral for specialist services - Osteoarthritis

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on