Professional Documents
Culture Documents
Abstract
David S. Jevsevar, MD, MBA Treatment of Osteoarthritis of the Knee: Evidence-Based Guideline,
2nd Edition, is based on a systematic review of the current
scientific and clinical research. This guideline contains 15
recommendations, replaces the 2008 AAOS clinical practice
guideline, and was reevaluated earlier than the 5-year
recommendation of the National Guideline Clearinghouse because
of methodologic concerns regarding the evidence used in the first
guideline. The current guideline does not support the use of
viscosupplementation for the treatment of osteoarthritis of the knee.
In addition, the work group highlighted the need for better research
in the treatment of knee osteoarthritis.
ticipants. This was done to limit the mendations 3A, 6, 9, and 12) and plementation were less likely to be
“small study” effect of lower-powered two recommendations to “we cannot published than were studies with
clinical trials. It also helped to mitigate suggest” (Recommendations 5 and positive results.
against publication bias in the evidence 11), which implies evidence only of a Current published studies, despite
review. Additionally, a follow-up pe- lack of efficacy.1 a clear publication bias toward posi-
riod of at least 4 weeks was required. The 2013 OA of the knee CPG tive results, do not show a clinically
Thus, studies showing potential effi- contains one significant recommen- effective response for HA injections
cacy at 2 weeks status post-intervention dation change from the 2008 CPG. based on MCII. Some peer reviewers
were not included in the update. The current CPG could not support were critical of this finding, espe-
More than 10,000 separate pieces the use of viscosupplementation for cially in light of the important clini-
of literature were reviewed during the treatment of OA of the knee. cal practice implications. Many high-
the evidence analysis phase. The This Strong recommendation (Rec- lighted prior systematic reviews
AAOS uses a “best-evidence synthe- ommendation 9) differs from the In- supported the use of HA.3,5 We re-
sis” form of evidence analysis, mean- conclusive recommendation in the viewed these published systematic re-
ing that, although all studies that previous guideline, which was views and found that they suffer sev-
meet the inclusion criteria are exam- largely the motivating factor for the eral faults. Most of them do not
ined, only the highest levels of avail- expedited update of the CPG. The address the issues of publication
able evidence are used in the meta- recommendation in the second edi- bias, between-study heterogeneity,
analysis and network meta-analysis. tion reads, “We cannot recommend and clinical significance in determin-
When completed, the second edi- using hyaluronic acid (HA) for pa- ing final recommendations. Addi-
tion of the osteoarthritis (OA) of the tients with symptomatic osteoarthri- tionally, several reviewers noted in-
knee CPG was subjected to the most tis of the knee.”1 The work group herent faults in using the MCII to
extensive peer review yet for an understands the potential impact determine clinical significance. The
AAOS CPG. Sixteen peer reviewers, that this recommendation could have AAOS CPG process has used MCII
representing multiple specialty soci- on clinical practice. The evidence did to elucidate clinical significance since
eties, submitted formal peer reviews. not support the efficacy of viscosup- the inception of the guidelines; it rep-
Each reviewer meticulously dissected plementation. Although statistically resents the best validated measure of
the final recommendations of the significant outcomes were seen in MCII when trying to determine
document, and important changes studies using higher molecular whether a treatment truly has effi-
were made to the final document weight HA preparations, these were cacy rather than providing just slight
based on the work group’s consider- not clinically significant, based on a improvements that register as statis-
ation of the well-informed and in- lack of minimum clinically important tically significant.7,10-13
sightful comments from the peer re- improvement (MCII). Of note, 14 Two other effect-size tools, the pa-
viewers. Reviewers correctly noted high and moderate research-quality tient acceptable symptomatic state,
that the recommendations did not in- articles were analyzed using this met- which is an absolute score beyond
clude “harm/risk” analysis but ric in determining clinically signifi- which patients are satisfied, and the
rather only evidence of the presence cant differences. The AAOS believes Initiative on Methods, Measurement,
or absence of effect. The original that MCII is the best way to measure and Pain Assessment in Clinical Trials
recommendations “recommended such differences.7 Overall, the litera- score, which denotes a specific percent-
against” several of the treatments, ture on viscosupplementation has a age of patients who are satisfied, both
which could have been interpreted as significant degree of publication have methodologic limitations that
implying “harm” or “risk.” The bias, as highlighted by other system- make them less appropriate for the
work group agreed and changed the atic reviews,3,5,8,9 and the analysis in- AAOS CPG analysis process. Addi-
language of four recommendations dicates that studies with results that tional high methodologic-quality stud-
to “we cannot recommend” (Recom- did not support the use of viscosup- ies on the effects of high molecular
Treatment of Osteoarthritis of the Knee: Evidence-Based Guideline, 2nd Edition, Work Group: David S. Jevsevar, MD, MBA (Chair),
Gregory Alexander Brown, MD, PhD (Co-chair), Dina L. Jones, PT, PhD, Elizabeth G. Matzkin, MD, Paul Manner, MD, FRCSC,
Peeka Mooar, MD, John T. Schousboe, MD, PhD, Steven Stovitz, MD, Michael Goldberg, MD (Chair, Guidelines Oversight
Committee), James O. Sanders, MD, (Vice-chair, Guidelines Oversight Committee), and Kevin Bozic, MD, MBA (Chair, Council on
Research and Quality). Staff of the American Academy of Orthopaedic Surgeons: William R. Martin III, MD, Deborah S. Cummins,
PhD, Patrick Donnelly, MA, and Anne Woznica, MLS.
weight preparations on OA of the knee underway and should further define ues, preferences, and rights. For
with subgroup analysis are awaited. clinical scenarios for patients with treatment procedures to provide ben-
Other recommendations changed OA of the knee. efit, mutual collaboration with
based on the strength of their sup- The work group highlighted the need shared decision-making between pa-
porting evidence only. Some review- for better research in the management tient and physician/allied healthcare
ers expressed concern over changing of knee OA, in addition to improved provider is essential.
the intra-articular corticosteroid in- methodologies that differentiate re- A Strong recommendation means
jection recommendation to a lower sponders from nonresponders. Evi- that the quality of the supporting ev-
grade. The evidence supported this dence, whether strong or inconclusive, idence is high. A Moderate recom-
change, and the use of intra-articular is never sufficient to make important mendation means that the benefits
corticosteroid injections does remain clinical decisions. “Individual values exceed the potential harm (or that
supported in clinical practice. The and preferences must balance this ev- the potential harm clearly exceeds
recommendation regarding arthro- idence to achieve optimal shared the benefits in the case of a negative
scopic meniscectomy increased in decision-making and highlight that the recommendation), but the quality/
strength from a Consensus to an In- practice of evidence-based medicine is applicability of the supporting evi-
conclusive recommendation, which not a “one size fits all” approach.14 dence is not as strong. A Consensus
is now supported by evidence and is It is again important to note that recommendation means that expert
no longer based solely on expert evidence-based practice incorporates opinion supports the guideline rec-
opinion. The current Inconclusive three components: scientific evi- ommendation even though there is
recommendation does help the dence, the clinician’s experience, and no available empirical evidence that
AAOS support the use of this proce- the patient’s values. No single com- meets the inclusion criteria of the
dure in our patients with OA of the ponent of patient care can stand guideline’s systematic review. An In-
knee. alone. conclusive recommendation means
The second edition of the OA of that there is a lack of compelling evi-
the knee CPG addresses concerns dence that has resulted in an unclear
raised regarding methodologic flaws Recommendations balance between benefits and poten-
associated with the evidence base of tial harm.
the first edition. The AAOS CPG This Summary of Recommendations
process benefitted from the extensive of the AAOS Treatment of Osteoar- Recommendation 1
involvement of the peer reviewers thritis of the Knee: Evidence-Based
We recommend that patients with
and specialty societies and will con- Guideline, 2nd Edition, contains a
symptomatic OA of the knee partici-
tinue to do so. The process improves list of the evidence-based treatment
pate in self-management programs,
with the thoughtful criticism of our recommendations and includes only
strengthening, low-impact aerobic
guidelines and the evidence synthesis less invasive alternatives to knee re-
exercises, and neuromuscular educa-
process. This CPG, as with all AAOS placement. Discussion of how each
tion and engage in physical activity
CPGs, is not intended as a tool for recommendation was developed and
consistent with national guidelines.
coverage determinations. The AAOS the complete evidence report are
Strength of recommendation:
also remains committed to ensuring contained in the full guideline at
Strong.
that the guidelines are interpreted www.aaos.org/guidelines. Readers
Implication: Practitioners should
and used properly and will advocate are urged to consult the full guideline
follow a Strong recommendation un-
vigorously on behalf of patients and for the comprehensive evaluation of
less a clear and compelling rationale
members. the available scientific studies. The
for an alternative approach is pres-
Although a CPG delineates recommendations were established
ent.
whether a procedure, intervention, using methods of evidence-based
or diagnostic test “works,” the medicine that rigorously control for
bias, enhance transparency, and pro- Recommendation 2
AAOS also will follow this CPG with
an Appropriate Use Criteria (AUC). mote reproducibility. We suggest weight loss for patients
The AUC further defines “in which This Summary of Recommenda- with symptomatic OA of the knee
patients” and “when” an interven- tions is not intended to stand alone. and a body mass index ≥25.
tion, procedure, or diagnostic test is Medical care should be based on evi- Strength of recommendation:
appropriate. Work on the accompa- dence, a physician’s expert judgment, Moderate.
nying AUC for this CPG is currently and the patient’s circumstances, val- Implication: Practitioners should
bridement in patients with a primary symptomatic medial compartment 8. Rutjes AW, Jüni P, da Costa BR, Trelle S,
Nüesch E, Reichenbach S:
diagnosis of symptomatic OA of the OA of the knee. Viscosupplementation for osteoarthritis
knee. Strength of Recommendation: of the knee: A systematic review and
Consensus. meta-analysis. Ann Intern Med 2012;
Strength of recommendation:
157(3):180-191.
Strong. A harms analysis on this rec- Implication: Practitioners should
9. Wang CT, Lin J, Chang CJ, Lin YT, Hou
ommendation was not performed. be flexible in deciding whether to SM: Therapeutic effects of hyaluronic
Implication: Practitioners should follow a recommendation classified acid on osteoarthritis of the knee: A
meta-analysis of randomized controlled
follow a Strong recommendation un- as Consensus, although they may
trials. J Bone Joint Surg Am 2004;86(3):
less a clear and compelling rationale give it preference over alternatives. 538-545.
for an alternative approach is pres- Patient preference should have a sub- 10. Angst F, Aeschlimann A, Michel BA,
ent. stantial influencing role. Stucki G: Minimal clinically important
rehabilitation effects in patients with with osteoarthritis of the lower important difference, low disease activity
osteoarthritis of the lower extremities. extremities. Arthritis Rheum 2001;45(4): state, and patient acceptable symptom
J Rheumatol 2002;29(1):131-138. 384-391. state: Methodological issues.
J Rheumatol 2005;32(10):2025-2029.
11. Angst F, Aeschlimann A, Stucki G: 12. Guyatt GH, Juniper EF, Walter SD,
Smallest detectable and minimal Griffith LE, Goldstein RS: Interpreting 14. Straus SE, Glasziou P, Richardson WS,
clinically important differences of treatment effects in randomised trials. Haynes RB: Evidence-Based Medicine:
rehabilitation intervention with their BMJ 1998;316(7132):690-693. How to Practice and Teach It, ed 4.
implications for required sample sizes Edinburgh, UK, Churchill Livingstone
using WOMAC and SF-36 quality of life 13. Tubach F, Wells GA, Ravaud P, Elsevier, 2011.
measurement instruments in patients Dougados M: Minimal clinically