Search results
A total of 8612 records were identified through electronic database searching, with 6452 records remained (2160 duplicates) for title and abstract screening (Fig. 1). Forty-seven records proceeded to full text screening (35 records were excluded with reasons) while an additional 42 records were identified through other sources (websites of individual internationally renowned arthritis societies and organisations, GIN library and citation search), of which 39 were excluded with reasons (Fig. 1). In total, 15 guidelines were included in this systematic review.
Guideline characteristics
Table 1 shows the characteristics and development processes of all the included guidelines. Guideline development groups were affiliated with either a professional organisation or a government department: American College of Rheumatology (ACR)3; European League Against Rheumatism (EULAR)34,35; Italian Society for Rheumatology (ISR)36; Turkish League Against Rheumatism (TLAR)37; Pan American League of Associations for Rheumatology (PANLAR)38; Rheumatology and Immunology Specialised Committee39; Malaysian Society of Rheumatology (MSR)40; Osteoarthritis Research Society International (OARSI)4; European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis and Musculoskeletal Diseases (ESCEO)41; American Academy of Orthopaedic Surgeons (AAOS)42,43; the Royal Australasian College of General Practitioners (RACGP)44; Department of Veterans Affairs and Department of Defence (VaDoD)45 and National Institute for Health and Care Excellence (NICE)46. Development of all guidelines involved a multidisciplinary team, most (n = 9) comprised of a working group of medical experts, literature review team, allied health and patient representatives3,4, 34–36,38,41,45,46. There were no patient representatives in the RACGP, AAOS, TLAR, MSR and the Chinese guidelines37,39,40,42−44. The target groups included mostly clinicians, health professionals and allied health managing patients with osteoarthritis3,4,34−46 while 4 guidelines also targeted other stakeholders (e.g. patients, policy makers and health insurance agencies)34–36, 40.
Table 1
Characteristics of the included guidelines (n = 15)
Author / Year / Country
|
Organisational affiliation
|
Funding body
|
Target group
|
Guideline development group
|
Guideline review / journal publication
|
Guideline update
|
Evidence base methods
|
LoE
|
SoR
|
AAOS
2021
US43
|
American Academy of Orthopaedic Surgeons (AAOS)
|
AAOS
|
Orthopaedic surgeons, other healthcare providers, medical practitioners
|
Multidisciplinary
|
Internal and external
N
|
5 years
|
SLR
GRADE
|
Y
|
Y
|
VADoD
2020
US45
|
Department of Veterans Affairs (VA) and Department of Defence (DoD)
|
VADoD
|
Primary care providers or specialists
|
Multidisciplinary
|
Internal and external
N
|
NR
|
SLR
GRADE
|
N
|
Y
|
Zhang
2020
China39
|
Rheumatology and Immunology Specialized Committee, Cross-Straits Medicine Exchange Association
|
Rheumatology and Immunology Expert Committee of the Cross-Strait medical and Health Exchange Association
|
Chinese clinicians,
specialists, professionals involved in management of OA
|
Multidisciplinary
|
Internal and external
Y
|
2022
|
SLR
GRADE
RIGHT checklist
|
Y
|
Y
|
Kolasinski
2019
US3
|
American College of Rheumatology (ACR)
|
ACR and the Arthritis Foundation
|
Patients and clinicians
|
Multidisciplinary
|
External
Y
|
NR
|
SLR
GRADE
|
Y
|
Y
|
Bannuru
2019
US4
|
Osteoarthritis research society international (OARSI)
|
OARSI
|
Clinicians
|
Multidisciplinary
|
External
Y
|
NR
|
SLR
GRADE
|
Y
|
Y
|
Bruyere
2019
Belgium41
|
European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis and Musculoskeletal Diseases (ESCEO)
|
ESCEO
|
Clinicians
|
Multidisciplinary
|
Internal and external
Y
|
NR
|
SLR
GRADE
|
N
|
Y
|
Ariani
2019
Italy36
|
Italian Society for Rheumatology (ISR)
|
ISR
|
Physicians, health professionals, patients and policy makers
|
Multidisciplinary
|
External
Y
|
Stated (no planned date)
|
SLR
AGREEII
|
Y
|
N
|
RACGP
2018
Australia44
|
Royal Australasian College of General Practitioners (RACGP)
|
Funded in part by Medibank Better Health Foundation
|
General practitioners, health professionals
|
Multidisciplinary
|
External (approved by Chief Executive Officer of NHMRC)
N
|
5 years
|
SLR
GRADE
|
Y
|
Y
|
Kloppenburg
2018
The Netherlnds35
|
European League Against Rheumatism
(EULAR)
|
EULAR
|
All health professionals, patients and relevant stakeholders (e.g. policy makers, health insurance companies)
|
Multidisciplinary
|
Internal
Y
|
NR
|
SLR
AGREEII
|
Y
|
Y
|
AAOS
2017
US42
|
American Academy of Orthopaedic Surgeons (AAOS)
|
AAOS
|
Clinicians, surgeons, specialists, allied health
|
Multidisciplinary
|
Internal and External
N
|
5 years
|
SLR
GRADE
|
Y
|
Y
|
Tuncer
2017
Turkey37
|
Turkish League Against Rheumatism (TLAR)
|
NR
|
Clinicians
|
Multidisciplinary
|
Internal
Y
|
NR
|
SLR
Oxman-Guyatt index and Jadad Scale
|
Y
|
Y
|
Rillo
2016
Venezuela38
|
Pan American League of Associations for Rheumatology (PANLAR)
|
PANLAR
|
NR
|
Multidisciplinary
|
Internal and External
Y
|
NR
|
SLR
Jadad scale
|
Y
|
Y
|
NICE
2014
UK46
|
National Institute for Health and Care Excellence (NICE)
|
NICE
|
Clinicians, patients
|
Multidisciplinary
|
External
N
|
Stated (No planned date)
|
SLR
GRADE
|
Y
|
N
|
Fernandes
2013
Norway34
|
European League Against Rheumatism
(EULAR)
|
EULAR
|
Clinicians, healthcare providers, researchers in OA, policy makers
|
Multidisciplinary
|
Internal
Y
|
NR
|
SLR
EULAR Standard operating procedure
|
Y
|
Y*
|
MOH
2013
Malaysia40
|
Malaysian Society of Rheumatology (MSR)
|
Ministry of Health Malaysia (MOH) and MSR
|
Healthcare professionals, relevant stakeholders in all healthcare setting
|
Multidisciplinary
|
External
N
|
2017
|
SLR
Scottish Intercollegiate Guidelines Network
|
Y
|
Y
|
* Presented as Level of agreement |
In all the included guidelines, the evidence to support recommendations was derived from a systematic literature review (SLR)3,4,34−46, with detailed methodology outlined, except for the PANLAR guideline (Appendix eTable1). The method used to grade the quality/certainty of the evidence differed among guidelines: majority (8 of 15) of the guidelines used Grading of Recommendations Assessment, Development and Evaluation (GRADE)3,4,39,41,43−46; others are detailed in Table 1 and Appendix eTable 1. Thirteen guidelines described the strength of their recommendations3,4,35−45 using different criteria, as described in Appendix eTable1. The 2 guidelines (EULAR and NICE) that did not provide strength of recommendations (SoR) had graded the level of evidence34,46. In 8 of 13 guidelines, the SoR was concordant with the quality of evidence from the SLR3,4,39,41−45. The EULAR guidelines (knee/hip25 and hand35) provided level of agreement among all task force members34,35 with additional grading of their recommendations35. Economic considerations were taken into account in the NICE guideline46.
All guidelines were peer-reviewed, either internally from experts within the affiliated organisation group (n = 9)34,35,37–39,41−43,45 and/or externally by experts within the relevant field (n = 12)3,4,36,38−46. Nine guidelines were also subjected to a peer-review process required for journal publication3,4, 34–39,41.
The site of osteoarthritis in the included guidelines varies (Tables 2 to 4): knee osteoarthritis (n = 13)3,4,34,36–41,43−46, followed by hip osteoarthritis (n = 11)3,4,34,36,38–40,42,44–46 and hand osteoarthritis (n = 4)3,35,36,38. The OARSI provided guideline for polyarticular osteoarthritis4. The NICE and osteoarthritis guideline in China did not specify the type of osteoarthritis, but encompassed knee, hip and hand osteoarthritis39,46.
Table 4
Summary of weight loss recommendation in osteoarthritis (other than knee and hip or unspecified) guidelines
|
Zhang
202039
|
ACR
20193
|
OARSI
20194
|
ISR
201936
|
EULAR
201835
|
PANLAR
201638
|
NICE
201446
|
Site of osteoarthritis
|
All types of OA
|
Hand
|
Polyarticular
|
Hand
|
Hand
|
Hand
|
All types of OA
|
Is weight loss recommended?
|
Y
|
N
|
Y
|
N
|
N
|
N
|
Y
|
LoE
|
Level A
|
|
Polyarticular: Level 1B
|
|
|
|
NR
|
SoR
|
Strong
|
|
Polyarticular OA: Conditional
|
|
|
|
NR
|
Target group
|
|
|
no comorbid conditions
|
|
|
|
overweight or obese who have associated functional limitations
|
Recommendation on magnitude of weight loss
|
NR
|
|
NR
|
|
|
|
NR
|
Weight loss strategies / other comments
|
Evidence for weight loss were mainly derived from knee osteoarthritis.
|
Weight management only recommended for knee and/or hip osteoarthritis
|
Dietary ± exercise in certain subgroup (no comorbid conditions; widespread pain and/or depression).
|
Weight loss recommended for patients with hip and knee osteoarthritis who are overweight.
|
|
|
NR
|
Abbreviation: |
LoE: level of evidence |
N: no |
NR: not reported |
OA: osteoarthritis |
SoR: strength of recommendation |
Y: yes |
Methodological quality
The mean scores for the AGREE-II domains were 35.5 to 92.5 (Table 5). Six guidelines (AAOS knee43, ACR3, OARSI4, RACGP44, EULAR hand35, EULAR knee/hip34) had mean domain scores of > 80%. Overall guideline assessment scores ranged from 3.00 to 6.50 out of 7 maximum possible score [median(%) 6.0, (interquartile range 4.6, 6.0)]. Of the guidelines (ISR36 and MSR40) that scored low (< 50% mean domain scores), shortcomings included limited or no descriptions of input from guideline end users or patients; criteria for selecting evidence, strengths and limitations of evidence, and methods for formulating recommendations; external reviews before publication; plans for updating; barriers to implementation, resource implications, and how to implement guideline recommendations; and measures taken to ensure editorial independence.
Table 5
Guideline assessment according to the AGREE-II instrument
Author/ Guideline organisation or society / year
|
Domain scores (%)
|
Mean overall quality (maximum possible score = 7)
|
Scope and purpose
|
Stakeholder involvement
|
Rigour of development
|
Clarity and presentation
|
Applicability
|
Editorial independence
|
Mean domain scores (%)
|
AAOS 202143 (knee)
|
100.0
|
88.1
|
96.4
|
100.0
|
82.1
|
82.1
|
91.5
|
6.0
|
VADoD 202045
|
95.2
|
61.9
|
93.8
|
95.2
|
73.2
|
53.6
|
78.8
|
6.0
|
Zhang 202039
|
90.5
|
81.0
|
87.5
|
88.1
|
30.4
|
82.1
|
76.6
|
6.0
|
ACR 20193
|
97.6
|
92.9
|
94.6
|
100.0
|
37.5
|
75.0
|
82.9
|
6.0
|
OARSI 20194
|
96.1
|
74.3
|
100.0
|
95.3
|
68.9
|
81.7
|
86.1
|
6.0
|
ESCEO 201941
|
90.5
|
69.0
|
76.8
|
100.0
|
50.0
|
85.7
|
78.7
|
6.0
|
ISR 201936
|
30.7
|
78.0
|
54.8
|
89.0
|
33.0
|
12.8
|
49.7
|
3.5
|
RACGP 201844
|
100.0
|
98.0
|
84.0
|
79.9
|
87.2
|
82.3
|
88.6
|
6.2
|
EULAR 201835 (hand)
|
90.0
|
88.9
|
100.0
|
97.6
|
85.3
|
93.2
|
92.5
|
5.6
|
AAOS 201742 (hip)
|
56.0
|
78.5
|
65.2
|
84.0
|
78.9
|
99.0
|
77.1
|
4.6
|
TLAR 201737
|
58.0
|
78.2
|
88.3
|
28.3
|
36.2
|
32.2
|
53.5
|
3.0
|
PANLAR 201638
|
81.0
|
69.0
|
58.9
|
92.6
|
41.1
|
85.7
|
71.4
|
5.0
|
NICE 201446
|
86.1
|
44.2
|
99.0
|
100.0
|
70.6
|
50.1
|
75.0
|
5.8
|
EULAR 201334 (knee/hip)
|
94.4
|
90.6
|
89.4
|
53.1
|
100.0
|
98.0
|
87.9
|
6.5
|
MSR 201340
|
20.0
|
18.8
|
12.0
|
67.1
|
30.2
|
65.0
|
35.5
|
4.0
|
Median score (%, IQR)
|
90.5 (58.0,96.1)
|
78.2 (69.0,88.9)
|
88.3 (65.2,96.4)
|
92.6 (79.9,100.0)
|
68.9 (36.2,82.1)
|
82.1 (53.6,85.7)
|
78.7 (71.4,87.9)
|
6.0 (4.6,6.0)_
|
Abbreviation: |
AGREE II: Appraisal of Guidelines Research and Evaluation II |
IQR: interquartile range |
Weight management recommendations
Thirteen guidelines incorporated obesity management as part of weight management recommendations3,4,34,36–41,43−46. There were no weight management recommendations in the EULAR hand osteoarthritis35 and AAOS hip osteoarthritis guidelines42. Where weight management was included, it was one of the key recommendations in 93,4,34,37,40,41,44−46 of the 123,4,34,36,37,39–41,43−46 guidelines for knee osteoarthritis and 83,4,34,38,40,44−46 of the 103,4,34,36,38–40,44−46 guidelines for hip osteoarthritis.
Most knee osteoarthritis guidelines (12 of 13)3,4,34,36,37,39–41,43−46 had recommendations for management of obesity (Table 2). All guidelines recommended “weight loss” while 4 guidelines4,37,39,44 used the term “weight management” instead of “weight loss”. Of these 4 guidelines, 2 guidelines recommended controlling body weight for all patients37,39, regardless of obesity status. Nine of 12 knee osteoarthritis guidelines had moderate to strong recommendations for weight loss3,4,34,37, 39–41,43,44, but the level of evidence behind these recommendations varied (Appendix eTable1). Notably, in 3 of 12 knee osteoarthritis guidelines, the SoR for weight loss was discordant to the level of evidence: ACR3 had strong recommendation on moderate level of evidence; ESCEO41 had strong recommendation while level of evidence not reported; RACGP44 had strong recommendation on very low level of evidence. Although the level of evidence behind the weight loss recommendation was drawn from randomised clinical trials, there were several issues that resulted in the quality of evidence being rated moderate (ACR3) and very low (RACGP44): serious risk of bias from single-blind or unblinded study design; high attrition rates; wide confidence interval and short study period3,47. The strength of recommendation was justified by the general view that weight loss has low risk of harms, such that the overall benefits outweigh the risks3,47.
Of the 10 hip osteoarthritis guidelines had weight loss recommendation for weight management3,4,34,36,38–40,44−46 (Table 3), 7 guidelines3,34,36,38,40,45,46 recommended weight loss for hip osteoarthritis while 3 guidelines4,39,44 used the term “weight management”. While the strength for weight loss recommendation was strong in 4 of 9 guidelines (ACR3, RACGP44, EULAR34 and PANLAR38), the level of evidence behind these recommendations varied (Appendix eTable 1). Discordance were seen in ACR3 (strong recommendation on moderate level of evidence) and RACGP44 (strong recommendation on very low level of evidence).
None of the hand osteoarthritis guidelines recommended weight management3,35,36,38 (Table 4). For polyarticular osteoarthritis guideline, weight management was conditionally recommended4.
Target group for weight loss
Eleven of 12 recommendations for weight loss in knee osteoarthritis specifically targeted people who were overweight or obese3,34,36,37,39–41,43−46 (Table 2). Table 3 shows 10 of 11 guidelines that recommended weight loss for hip osteoarthritis targeted people who are overweight or obese, with the OARSI4 guideline specifically targeted those with BMI of ≥ 30 kg/m2. While the NICE46 guideline recommended weight loss for those who are overweight or obese with associated functional limitations, the OARSI4 guideline specifically targeted weight loss to those with no comorbid conditions, with gastrointestinal or cardiovascular conditions and with widespread pain and/or depression4 (Table 4).
Weight loss strategies suggested by guidelines
Nine of 12 knee osteoarthritis guidelines provided a general weight loss strategy3,4,34,36,37,43−46 (Table 2), with 8 of the 9 guidelines supported a general combination approach of exercise and/or dietary weight loss3,4,34,36,37,43−45. NICE46 provided reference to its own obesity guideline (NICE guideline for obesity48 on evidence of the most effective weight loss strategies) for strategies to lose weight. The EULAR34 guideline further described examples of strategies that were recognised to effect successful weight loss and maintenance, such as increase physical activity, follow a structured meal plan, limit portion size, nutritional education etc.
For hip osteoarthritis, 7 of 10 guidelines described a general, non-specific weight loss strategy that comprised of a combination of dietary and/or concomitant exercise3,4,34,36,44−46 (Table 3). Conversely, OARSI4 guideline recommended against dietary weight loss for individuals with hip osteoarthritis of any comorbidity but acknowledged that it may be recommended as part of a healthy lifestyle regimen to those with BMI ≥ 30 kg/m24.
The OARSI4 guideline for polyarticular osteoarthritis recommended weight loss using a combination of dietary weight management with or without an exercise component for those without comorbid conditions but recommended against dietary weight management for individuals with frailty4. The NICE46 guideline specifically referred to the NICE obesity guideline48 to provide recommendation for individuals who are overweight or obese with associated functional limitations.
Except for the EULAR34 guideline, all other guidelines have not provided details on strategies to effective dietary or concomitant exercise interventions for weight loss, specifically no details regarding type, duration, frequency or intensity of the recommended approach. There were no guidelines that mentioned the role of pharmacological or surgical weight loss interventions for osteoarthritis except for EULAR knee/hip34 guideline (Tables 2 and 3) that acknowledged the role of bariatric surgery as part of comprehensive weight management in people with knee or hip osteoarthritis who are morbidly obese.
Magnitude of weight loss
Two3,44 of 12 knee and 23,44 of 10 hip osteoarthritis guidelines specified the magnitude of weight loss required for weight management: ACR guideline recommended ≥ 5% of body weight3; RACGP guideline recommended a minimum weight loss target of 5-7.5% for those with BMI ≥ 25 kg/m244. The ACR guideline acknowledged a dose-response relationship in the degree of weight loss, such that clinically important benefits continue to increase with weight loss of 5–10%, 10–20% and > 20% of body weight3.