THE DIAGNOSIS OF PERIPROSTHETIC JOINT
INFECTIONS OF THE HIP AND KNEE GUIDELINE AND EVIDENCE REPORT
Adopted by the American Academy of Orthopaedic Surgeons Board of Directors June
18, 2010
Disclaimer This Clinical Practice Guideline was developed by an AAOS physician
volunteer Work Group based on a systematic review of the current scientific and
clinical information and accepted approaches to treatment and/or diagnosis.
This Clinical Practice Guideline is not intended to be a fixed protocol, as
some patients may require more or less treatment or different means of
diagnosis. Clinical patients may not necessarily be the same as those found in
a clinical trial. Patient care and treatment should always be based on a
clinicians independent medical judgment, given the individual patients
clinical circumstances. Disclosure Requirement In accordance with AAOS policy,
all individuals whose names appear as authors or contributors to Clinical Practice
Guideline filed a disclosure statement as part of the submission process. All
panel members provided full disclosure of potential conflicts of interest prior
to voting on the recommendations contained within this Clinical Practice
Guidelines. Funding Source This Clinical Practice Guideline was funded
exclusively by the American
Academy of Orthopaedic
Surgeons who received no funding from outside commercial sources to support the
development of this document. FDA Clearance Some drugs or medical devices referenced
or described in this Clinical Practice Guideline may not have been cleared by
the Food and Drug Administration (FDA) or may have been cleared for a specific
use only. The FDA has stated that it is the responsibility of the physician to
determine the FDA clearance status of each drug or device he or she wishes to
use in clinical practice. Copyright All rights reserved. No part of this
Clinical Practice Guideline may be reproduced, stored in aretrieval system, or
transmitted, in any form, or by any means, electronic, mechanical,
photocopying, recording, or otherwise, without prior written permission from
the AAOS. Published 2010 by the American Academy of Orthopaedic Surgeons 6300
North River Road Rosemont, IL 60018 First Edition Copyright 2010 by the American
Academy of Orthopaedic Surgeons
ii
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Summary of Recommendations
The following is a summary of the recommendations in the AAOS clinical
practice guideline, The Diagnosis of Periprosthetic Joint Infections of the Hip
and Knee. This summary does not contain rationales that explain how and why
these recommendations were developed nor does it contain the evidence
supporting these recommendations. All readers of this summary are strongly
encouraged to consult the full guideline and evidence report for this
information. We are confident that those who read the full guideline and
evidence report will note that the recommendations were developed using
systematic evidence-based processes designed to combat bias, enhance
transparency, and promote reproducibility. This summary of recommendations is
not intended to stand alone. Clinical decisions should be made in light of all
circumstances presented by the patient. Procedures applicable to the individual
patient rely on mutual communication between patient, physician, and other
healthcare practitioners. 1. In the absence of reliable evidence about risk
stratification of patients with a potential periprosthetic joint infection, it
is the opinion of the work group that testing strategies be planned according
to whether there is a higher or lower probability that a patient has a hip or
knee periprosthetic infection. Strength of Recommendation: Consensus Note:
Please see page 17 of this document for a definition of higher and lower
probability. 2. We recommend erythrocytesedimentation rate and C-reactive
protein testing for patients assessed for periprosthetic joint infection.
Strength of Recommendation: Strong 3. We recommend joint aspiration of patients
being assessed for periprosthetic knee infections who
have abnormal erythrocyte sedimentation rate AND/OR Creactive protein results.
We recommend that the aspirated fluid be sent for microbiologic culture,
synovial fluid white blood cell count and differential. Strength of
Recommendation: Strong
iii
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
4. We recommend a selective approach to aspiration of the hip based on the
patients probability of periprosthetic joint infection and the results of the
erythrocyte sedimentation rate (ESR) AND C-reactive protein (CRP). We recommend
that the aspirated fluid be sent for microbiologic culture, synovial fluid
white blood cell count and differential. Selection of Patients for Hip
Aspiration
Probability of Infection Higher Lower Lower Lower Higher or Lower ESR and CRP
Results + + or + − + + or + − ++ +− −− Planned
Reoperation Status Planned or not planned Planned Not planned Not planned
Planned or not planned Recommended Test Aspiration Aspiration or Frozen Section
Aspiration Please see Recommendation 6 No further testing
Key for ESR and CRP results + + = ESR and CRP test results are abnormal + −
= either ESR or CRP test result is abnormal − − = ESR and CRP test
results are normal
Strength of Recommendation: Strong 5. We suggest a repeat hip aspiration when
there is a discrepancy between the probability of periprosthetic joint
infection and the initial aspiration culture result. Strength of
Recommendation: Moderate 6. In the absence of reliable evidence, it is the opinion
of the work group that patients judged to be at lower probability for
periprostheic hip infection and without planned reoperation who
have abnormal erythrocyte sedimentation rates ORabnormal C-reactive protein
levels be re-evaluated within three months. We are unable to recommend specific
diagnostic tests at the time of this follow-up. Strength of Recommendation:
Consensus 7. In the absence of reliable evidence, it is the opinion of the work
group that a repeat knee aspiration be performed when there is a discrepancy
between the probability of periprosthetic joint infection and the initial
aspiration culture result. Strength of Recommendation: Consensus 8. We suggest
patients be off of antibiotics for a minimum of 2 weeks prior to obtaining
intra-articular culture. iv
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Strength of Recommendation: Moderate 9. Nuclear imaging (Labeled leukocyte
imaging combined with bone or bone marrow imaging, FDG-PET imaging, Gallium
imaging, or labeled leukocyte imaging) is an option in patients in whom diagnosis of periprosthetic joint infection has not been
established and are not scheduled for reoperation. Strength of
Recommendation: Weak 10. We are unable to recommend for or against computed
tomography (CT) or magnetic resonance imaging (MRI) as a diagnostic test for
periprosthetic joint infection. Strength of Recommendation: Inconclusive 11. We
recommend against the use of intraoperative Gram stain to rule out
periprosthetic joint infection. Strength of Recommendation: Strong 12. We
recommend the use of frozen sections of peri-implant tissues in patients who
are undergoing reoperation for whom the diagnosis of periprosthetic joint
infection has not been established or excluded. Strength of Recommendation:
Strong 13. We recommend that multiple cultures be obtained at the time of
reoperation in patients being assessed for periprosthetic joint infection.
Strength of Recommendation: Strong 14. We recommend against initiating
antibiotic treatment in patients with suspected periprosthetic joint infection
until after cultures from thejoint have been obtained. Strength of
Recommendation: Strong 15. We suggest that prophylactic preoperative
antibiotics not be withheld in patients at lower probability for periprosthetic
joint infection and those with an established diagnosis of periprosthetic joint
infection who are undergoing reoperation. Strength of Recommendation: Moderate
v
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Work Group
Craig Della Valle MD, Chair Rush University Medical Center 1611 W Harrison St #
300 Chicago, IL 60612-4861 Javad Parvizi, MD, Vice-Chair Rothman Institute 925
Chestnut St - 5th Fl Philadelphia, PA 19107 Thomas W Bauer, MD PhD Cleveland
Clinic Foundation Department of Pathology 9500 Euclid Ave Desk L25 Cleveland,
OH 44195 Paul E DiCesare, MD UC Davis Medical Center Department of Orthopaedic
Surgery 4860 Y St Ste 3800 Sacramento, CA 95817 Richard Parker Evans, MD
University of Arkansas for Medical Sciences Department of Orthopedics 4301 W
Markham, #531 Little Rock, AR 72205 John Segreti, MD Rush University Medical
Center 600 S Paulina St. Ste 143 Chicago, Il 60612 Mark Spangehl, MD Mayo
Clinic 5777 East Mayo Blvd Phoenix, AZ 85054 Guidelines and Technology
Oversight Chair: William C. Watters III MD 6624 Fannin #2600 Houston, TX 77030
Evidence Based Practice Committee Chair: Michael Keith, MD 2500 Metro Health
Drive Cleveland, OH 44109-1900 AAOS Staff: Charles M. Turkelson, PhD Director
of Research and Scientific Affairs 6300 N River Road Rosemont, IL 60018 Janet
L. Wies MPH AAOS Clinical Practice Guideline Mgr Patrick Sluka MPH AAOS
Research Analyst Kristin Hitchcock, MSI AAOS Medical Librarian Special
Acknowledgements Sara Anderson MPH Kevin Boyer Laura Raymond, MA
vi
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Peer Review
Participation in the AAOS peer review process does not constitute an
endorsement of this guideline by the participating organization. Thefollowing
organizations participated in peer review of this clinical practice guideline
and gave explicit consent to be listed in this document: American Association
of Hip and Knee Surgeons European Bone and Joint Infection Society Knee Society
Musculoskeletal Infection Society Society of Nuclear Medicine Participation in
the AAOS peer review process does not constitute an endorsement of this
guideline by the participating organization.
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
I. INTRODUCTION
OVERVIEW
This clinical practice guideline is based on a systematic review of published
studies on the diagnosis of periprosthetic joint infections of the hip and
knee. In addition to providing practice recommendations, this guideline also
highlights gaps in the literature and areas that require future research. This
guideline is intended to be used by all appropriately trained surgeons and
allqualified physicians evaluating patients for periprosthetic joint infections
of the hip and knee. It is also intended to serve as an information resource
for decision makers and developers of practice guidelines and recommendations.
GOALS AND RATIONALE
The purpose of this clinical practice guideline is to
help improve treatment based on the current best evidence. Current
evidence-based practice (EBP) standards demand that physicians use the best
available evidence in their clinical decision making. This clinical practice
guideline was developed following systematic review of the available literature
regarding the diagnosis of periprosthetic infections of the hip and the knee.
The systematic review detailed herein was conducted between October 2008 and
September 2009 and demonstrates where there is good evidence, where evidence is
lacking, and what topics future research must target in order to improve the
diagnosis of periprosthetic joint infections of the hip and knee. AAOS staff
and the physician work group systematically reviewed the available literature
and subsequently wrote the following recommendations based on a rigorous,
standardized process. Musculoskeletal care is provided in many different
settings by many different providers. We created this guideline as an
educational tool to guide qualified physicians through a series of diagnostic
decisions in an effort to improve the quality and efficiency of care. This
guideline should not be construed as including all proper methods of care or
excluding methods of care reasonably directed to obtaining the same results.
The ultimate judgment regarding any specific procedure or treatment must be
made in light of all circumstances presented by the patient and the needs and
resources particular to the locality or institution. Further, the scope of the
guideline does not include information on laboratory techniques or tissue
samplingtechniques beyond the evidence presented to support Recommendations 13
and 14. It is assumed that the well qualified physician will use his/her best
judgement, considering the individual patient circumstances as well as the
available resources, when addressing these issues.
INTENDED USERS
This guideline is intended to be used by orthopaedic
surgeons and all qualified physicians managing the diagnosis of periprosthetic
joint infections of the hip and knee. Typically, orthopaedic surgeons will have
completed medical training, a qualified residency in orthopaedic surgery, and
some may have completed additional sub-specialty training. It is also intended
to serve as an information resource for professional healthcare practitioners
and developers of practice guidelines and recommendations.
1
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
PATIENT POPULATION
This document addresses the diagnosis of periprosthetic joint infections in
patients who have undergone arthroplasty of the hip or knee This
guideline is not intended for patients with superficial infections.
ETIOLOGY
Periprosthetic joint infection can be caused by entry of organisms into the
wound during surgery, hematogenous spread, recurrence of sepsis in a previously
infected joint, or contiguous spread of infection from a local source.22
INCIDENCE
The incidence of periprosthetic joint infection after primary hip or knee
arthroplasty is over 2% among the Medicare population.56, 76 The incidence of
periprosthetic infection is higher after revision arthroplasty than after
primary arthroplasty in both joints.64, 82
BURDEN OF DISEASE
Periprosthetic joint infection requires significant resources to diagnose and
treat. Infection is a common cause of revision arthroplasty, with 15% of
revision total hip arthroplasties and 25% of revision total knee arthroplasties
being due to infection.12, 13
RISK FACTORS
Risk factorsidentified and supported by the evidence include prior infection of
the joint (knee), superficial surgical site infection (hip and knee), obesity
(hip), extended operative time (>2.5 hours, hip and knee) and
immunosuppression (knee). The work group also discusses additional risk factors
based on consensus in Recommendation 1.
EMOTIONAL AND PHYSICAL IMPACT
Periprosthetic joint infection is a serious complication of total joint
replacement resulting in significant morbidity, including pain, loss of
function, and potential removal of the prosthesis.
POTENTIAL BENEFITS AND HARMS
There is the possibility of harm resulting from diagnostic tests, including the
possibility of the introduction of bacteria into the joint during an aspiration
and patient pain and/or discomfort while undergoing the procedure. Therefore, discussion of available diagnostic procedures applicable to the
individual patient rely on mutual communication between the patient and
physician, weighing the potential risks and benefits for that patient.
2
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
II. METHODS
This section describes the methods used to prepare this guideline and
systematic review, including search strategies used to identify literature,
criteria for selecting eligible articles, determining the strength of the
evidence, data extraction, methods of statistical analysis, and the review and
approval of the guideline. The methods used to perform this systematic review
were employed to minimize bias in the selection and analysis of the available
evidence.17, 69 This is vital to the development of reliable, transparent, and
accurate clinical recommendations for diagnosing periprosthetic joint
infections of the hip and knee. The AAOS Diagnosis of Periprosthetic Joint
Infections of the Hip and Knee physician work group prepared this guideline and
systematic review with the assistance of the AAOS ClinicalPractice Guidelines
Unit in the Department of Research and Scientific Affairs at the AAOS (Appendix
I). To develop this guideline, the work group held an introductory meeting to
develop the scope of the guideline on October 17 and 18, 2008. Upon completion
of the systematic review, the work group met again on September 12 and 13, 2009
to write and vote on the final recommendations and rationales for each
recommendation. The resulting draft guidelines were then peer-reviewed,
subsequently sent for public commentary, and then sequentially approved by the
AAOS Evidence Based Practice Committee, AAOS Guidelines and Technology
Oversight Committee, AAOS Council on Research, Quality Assessment, and Technology,
and the AAOS Board of Directors (see Appendix II for a description of the AAOS
bodies involved in the approval process)
FORMULATING PRELIMINARY RECOMMENDATIONS
The work group began work on this guideline by constructing a set of
preliminary recommendations. These recommendations specify [what] should be
done in [whom], [when], [where], and [how often or how long]. They function as
questions for the systematic review, not as final recommendations or
conclusions. Preliminary recommendations are almost always modified on the
basis of the results of the systematic review. Once established, these a priori
preliminary recommendations cannot be modified until the final work group
meeting, they must be addressed by the systematic review, and the relevant review
results must be presented in the final guideline.
STUDY SELECTION CRITERIA
We developed a priori article inclusion criteria for
our review. These criteria are our rules of evidence and articles that do not
meet them are, for the purposes of this guideline, not evidence. To be included
in our systematic reviews (and hence, in this guideline) an article had to be a
report of a study that: Addressed the value of tests
for diagnosis ofperiprosthetic joint infection in the hip and knee.
3
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Was written in English and published in or after 1970.
For MRI, CT and PET only studies published after 2000 were included. For other
nuclear imaging modalities, only studies published after 1975 were considered. Was
not an abstract, unpublished study report, letter, case report, historical
article, editorial, tutorial, traditional review or commentary. Was published in the peer-reviewed literature. Was not a cadaveric, animal, or an in vitro study. Included 25 or more patients per study arm. Reported sufficient data to construct a 2 x 2 table. Was of
the highest level of available evidence (best available evidence), assuming
that there were two or more studies of that higher level. Hip and knee studies
were considered separately. For example, if there were two Level II studies of
the hip and two Level II studies of the knee that addressed the recommendation
Level III and IV studies were not included. Was not evaluating a test in a
cohort where the infected prosthesis had already been removed (no more than 10%
of cohort or data from these patients reported separately).
INCLUSION OF STUDIES WITH MIXED PATIENT POPULATIONS The work group specified a
priori to the literature search that data would be stratified by joint but that
mixed studies could be accepted and reported as such. BEST AVAILABLE EVIDENCE When examining primary studies, we analyzed the best
available evidence regardless of study design. We first considered the
randomized controlled trials identified by the search strategy. In the absence
of two or more RCTs, we sequentially searched for prospective controlled
trials, prospective comparative studies, retrospective comparative studies, and
prospective case-series studies. Only studies of the highest level of available
evidence were included,assuming that there were 2 or
more studies of that higher level. For example, if there were two Level II
studies that addressed the recommendation, Level III and IV studies were not
included.
LITERATURE SEARCHES
We attempted to make our searches for articles
comprehensive. Using comprehensive literature searches ensures that the
evidence we considered for this guideline is not biased for (or against) any
particular point of view. We searched for articles published from January 1970
to August 10, 2009. Strategies for searching electronic databases were
constructed by a Medical Librarian to identify 4
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
relevant clinical studies. We searched four electronic databases; PubMed,
EMBASE, CINAHL, and The Cochrane Central Register of Controlled Trials. All
searches of electronic databases were supplemented with manual screening of the
bibliographies of all retrieved publications. We also searched the bibliographies
of recent systematic reviews and other review articles for potentially relevant
citations. Finally, work group members provided a list of potentially relevant
studies that were not identified by our searches. All articles identified were
subject to the study selection criteria listed above. The study attrition
diagram in Appendix III provides details about the inclusion and exclusion of
the studies considered for this guideline. The search strategies used to
identify these studies are provided in Appendix IV.
DATA EXTRACTION
Data elements extracted from studies were defined in consultation with the
physician work group. One analyst completed data extraction for all studies.
The elements extracted are shown in Appendix V. Evidence tables were
constructed to summarize the best evidence pertaining to each preliminary
recommendation. Disagreements about the accuracy of extracted data were
resolved by discussion and consulting thephysician work group. The work group
specified a priori to the literature search that data would be stratified by
joint but that mixed studies could be accepted and reported as such. When
studies did not separate the data by joint, we could not report them
separately. If a study with mixed joints reported the data for each joint we
reported them as such. If a study reported mixed joints but had fewer than 25
patients per joint, only the mixed data was reported.
JUDGING THE QUALITY OF EVIDENCE
Determining the quality of the included evidence is
vitally important when preparing any evidence-based work product. Doing so
conveys the amount of confidence one can have in any studys results. One has
more confidence in high quality evidence than in low quality evidence.
Assigning a level of evidence on the basis of study design plus other quality
characteristics ties the levels of evidence we report more closely to quality
than levels of evidence based only on study design. Because we tie quality to
levels of evidence, we are able to characterize the confidence one can have in
their results. Accordingly, we characterize the confidence one can have in
Level I evidence as high, the confidence one can have in Level II and III
evidence as moderate, and the confidence one can have in Level IV and V
evidence as low. Similarly, throughout the guideline we refer to Level I
evidence as reliable, Level II and III evidence as moderately reliable, and
Level IV and V evidence as not reliable. DIAGNOSTIC STUDIES We used the Quality
Assessment of Diagnostic Accuracy Studies (QUADAS) instrument to identify potential
bias and assess variability and the quality of reporting in studies 5
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
reporting the effectiveness of diagnostic techniques.109 Studies without any
indication of bias are categorized as Level I studies. Studies with a known
bias are downgraded toLevel IV or Level III (for a bias for which we have or
have not found published evidence that this bias affects diagnostic results,
respectively). Studies with more than one bias are downgraded to Level V. Those
studies that do not sufficiently report their methods for a potential bias are
downgraded to Level II since we are unable to determine if the bias did or did
not bias the results of the study (see Appendix VI). Although the definition of
a reference standard varies in the periprosthetic joint infection literature,
intraoperative cultures have been used most often as the reference standard in
interpreting the results of a diagnostic test, as indicated in Table 1.
Accordingly, when possible we also used intraoperative cultures as a reference
standard when computing measures of test performance. Table
1. Reference Standard Used in Included Diagnostic Studies
Reference Standard (Infection defined as positive results on following test(s))
Intraoperative cultures Intraoperative cultures and histology Histology At
least 2 of intraoperative cultures, purulence, and histology Intraoperative
cultures or histology Open wound or sinus communicating with the joint OR
systemic infection with pain in the hip and purulent fluid within the joint OR
positive result on at least 3 tests (ESR, CRP, joint aspiration, intraoperative
frozen section, and intraoperative culture) Histology or purulence or sinus
tract communicating with the prosthesis Intraoperative cultures or purulence
Aspiration or intraoperative cultures At least 2 of intraoperative cultures,
purulence, and histology; or 2 positive cultures Histology and gross operative
findings Intraoperative cultures or histology or purulence Intraoperative
cultures or histology or deep abscess Cultures or purulence or histology or
sinus tract communicating with the prosthesis Correlation between
intraoperative cultures and histology; the appearance ofthe tissue
intraoperatively; and the clinical course Intraoperative cultures and gross
sepsis At least 3 of CRP, ESR, aspiration culture, intraoperative purulence,
intraoperative culture Number of Studies 18 7 4 4 3 3
2 2 1 1 1 1 1 1 1 1 1
6
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Reference Standard (Infection defined as positive results on following test(s))
Abscess or sinus tract communicating with the joint space OR aspiration culture
OR ≥2 intraoperative cultures OR 1 culture and purulence or histology OR
purulence and histology
Number of Studies 1
PROGNOSTIC STUDIES In studies investigating the effect of a characteristic on
the outcome of disease, we assessed quality using a two-step process. Any study
that investigated a prospectively enrolled cohort of patients and utilized
regression analysis was initially categorized as a Level I study. A study that
used regression analysis in a retrospectively enrolled cohort of patients was
categorized as a Level II study. A case-control study that used regression
analysis was categorized as a Level III study. We next assessed the outcome
(dependent variable) for each prognostic factor (independent variable) using a
quality questionnaire and, when quality standards were not met, we downgraded
the level of evidence by one level (Appendix VI).
DEFINING THE STRENGTH OF THE RECOMMENDATIONS
Judging the quality of evidence is only a stepping
stone towards arriving at the strength of a guideline recommendation. Unlike
Levels of Evidence (which apply only to a given result at a given follow-up
time in a given study) strength of recommendation takes into account the
quality, quantity, and applicability of the available evidence. Strength also
takes into account the trade-off between the benefits and harms of a treatment
or diagnostic procedure, and the magnitude of a treatments effect. Strength of
recommendation expresses thedegree of confidence one can have in a
recommendation. As such, the strength expresses how possible it is that a
recommendation will be overturned by future evidence. It is very difficult for
future evidence to overturn a recommendation that is based on many high quality
randomized controlled trials that show a large effect. It is much more likely
that future evidence will overturn recommendations derived from a few small
case series. Consequently, recommendations based on the former kind of evidence
are given a high strength of recommendation and recommendations based on the
latter kind of evidence are given a low strength. To develop the strength of a
recommendation, AAOS staff first assigned a preliminary strength for each
recommendation that took only the quality and quantity of the available
evidence into account (see Table 2). Work group members then modified the
preliminary strength using the Form for Assigning Strength of Recommendation
(Interventions) shown in Appendix VII.
7
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 2. Strength of recommendation descriptions Overall Quality of
Evidence Good
Strength Strong
Moderate
Fair
Weak
Poor
Inconclusive
None or Conflicting No Evidence
Consensus
Description of Evidence Level I evidence from more than one study with
consistent findings for recommending for or against the intervention or
diagnostic. Level II or III evidence from more than one study with consistent
findings, or Level I evidence from a single study for recommending for or
against the intervention or diagnostic. Level IV or V
evidence from more than one study with consistent findings, or Level II or III
evidence from a single study for recommending for against the intervention or diagnostic.
The evidence is insufficient or conflicting and does not allow a recommendation
for or against the intervention or diagnostic. There is nosupporting evidence.
In the absence of reliable evidence, the work group is making a recommendation
based on their clinical opinion considering the known harms and benefits
associated with the treatment.
Each recommendation was written using language that accounts for the final
strength of the recommendation. This language, and the corresponding strength,
is shown in Table 3. Table 3. AAOS Guideline Language
Guideline Language Strength of Recommendation Strong Moderate Weak Inconclusive
We recommend We suggest option We are unable to recommend for or against In the
absence of reliable evidence, it is the Consensus opinion of this work group
CONSENSUS DEVELOPMENT
The recommendations and their strength were voted on using a structured voting
technique known as the nominal group technique.71 We present details of this
technique in Appendix VIII. Voting on guideline recommendations was conducted
using a secret ballot and workgroup members were blinded to the responses of
other members. If disagreement between workgroup members was significant, there
was further discussion
8
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
to see whether the disagreement(s) could be resolved. Up to three rounds
of voting were held to attempt to resolve disagreements. If disagreements were
not resolved following three voting rounds, no recommendation was adopted. Lack
of agreement is a reason that the strength for some recommendations are labeled as Inconclusive.
STATISTICAL METHODS
Likelihood ratios, sensitivity, specificity and 95% confidence intervals were
used to determine the accuracy of diagnostic modalities based on two by two
diagnostic contingency tables extracted from the included studies. When possible, prognostic factors were analyzed according to
sensitivity and specificity as well.80 Likelihood ratios (LR) indicate the
magnitude of the change in probability of disease due toa given test result.
For example, a positive likelihood ratio of 10 indicates that a positive test
result is 10 times more common in patients with disease than in patients
without disease. Likelihood ratios are interpreted according to previously
published values, as seen in Table 4.42 Likelihood ratios, sensitivity,
specificity and 95% confidence intervals were calculated in STATA 10.0
(StataCorp LP, College Station,
Texas). For meta-analysis of
diagnostic tests, we used the bivariate random effects model.38, 39, 89
Meta-analyses were conducted only if there were at least four studies for a
given diagnostic test. When a meta-analysis indicated between-study
heterogeneity (I2 >50%) and thresholds for a positive test result varied
between studies, we included the threshold as a covariate in the model to
determine whether the heterogeneity could be explained by the variation in
thresholds. Table 4. Interpreting Likelihood Ratios
Positive Likelihood Ratio >10 5-10 2-5 1-2 Negative Likelihood Ratio 0.5mg/dL)
CRP (>0.5mg/dL) CRP (>0.5mg/dL) CRP (>1 mg/dL) CRP (>2 mg/dL) CRP
(>2 mg/dL) CRP (>2 mg/dL) CRP (>1.0 mg/dL)
Level of Evidence I I I I I I I I I I I I I I I I
a a a a a a a a a a a a
a a a a
a a a a a a ? a a a a a a a a a
a a a a a a a a a a a a
a a a a
a a a a a a a a a a a a
a a a a
a a a a a a a a a a a a
a a a a
a a a a a a a a a a a a a a a a
a a a a a a a a a a a a
a a a a
a a a a a a a a a a a a
a a a a
a a a a ? ? a
a a a a a a a a ?
a a a a a a a a a a a a a a a a
a ? ? a a a ?
? a ? ? ? a ? ? a
a a a a a a a a a a a a
a a a a
a a a a a a a a a a a a
a a a a
52
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Withdrawals explained
Spectrum bias avoided
a a a a a a a a a a a a
a a a a
Table 23. Radiography and Serology Quality (Continued)
Disease progression bias avoided Uninterpretable/Intermediate test
result(s)reported Partial verification bias avoided Diagnostic review bias
avoided Appropriate reference standard Index test execution described Reference
standard execution described Clinical review bias avoided Differential
verification bias avoided Selection criteria described Incorporation bias
avoided Test review bias avoided
a = Yes a = No ? = Unclear
Author Greidanus Fink Bottner Bottner Della Valle Greidanus Greidanus Greidanus
Greidanus Schinsky Cyteval Cyteval Cyteval
N 151 145 78 78 94 151 151 151 151 201 65 65 65
Index Test CRP (>1.35 mg/dL) CRP (>1.35 mg/dL) CRP (>1.5 mg/dL) CRP
(>3.2 mg/dL) CRP (>1 mg/dL) ESR and CRP (22.5/1.35) ESR and CRP (30/1.0)
ESR or CRP (22.5/1.35) ESR or CRP (30/1.0) ESR or CRP (30/1.0) radiograph -
asymmetric position of femoral head radiograph - bone abnormalities (focal or
nonfocal lucency, periostitis) radiograph - focal lucency
Level of Evidence I I I I I I I I I I I I I
a a a a a a a a a a a a
a
a a ? ? a a a a a a a a a
a a a a a a a a a a a a
a
a a a a a a a a a a a a
a
a a a a a a a a a a a a
a
a a a a a a a a a a a a
a
a a a a a a a a a a a a
a
a a a a a a a a a a a a
a
? a a a a ? ? ? ?
a a a a
a a a a a a a a a a a a
a
a ? ? ? ? a a
a a ? a a a
a a a a a a a a a a
a a a a a a
a a a a a a a
53
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Withdrawals explained
Spectrum bias avoided
a a a a a a a a a a a a
a
Table 23. Radiography and Serology Quality (Continued)
Disease progression bias avoided Uninterpretable/Intermediate test result(s)
reported Partial verification bias avoided Diagnostic review bias avoided
Appropriate reference standard Index test execution described Reference
standard execution described Clinical review bias avoided Differential
verification bias avoided Selection criteria described Incorporation bias
avoided Test review bias avoided
a = Yes a = No? = Unclear
Author Cyteval Cyteval Bernay Barrack Barrack
N 65 65 31 69 69
Index Test radiograph - nonfocal lucency radiograph - periostitis radiograph
radiograph: complete radiolucent line adjacent to tibial component radiograph:
gross loosening radiograph: Knee Society Grade 3 classification for tibial
component radiolucency in addition to Grade 1 under femoral component
radiograph: periostitis WBC (>11.0*10^9/L) WBC (>9500/mm^3) WBC
(>9500/mm^3)
Level of Evidence I I II III III a a a a a
a a a a a
a a ? a a
a a a a a
a a a a a
a a a a a
a a a a a
a a a a a
a a a a a
a a a a a
a a ? a a
? ? a a a
a a a a a
Barrack
69
III
a
a
a
a
a
a
a
a
a
a
a
a
a
Barrack Spangehl Savarino Savarino
69 202 26 26
III I I I
a a a a
a a a a
a a a a
a a a a
a a a a
a a a a
a a a a
a a a a
a a a a
a a a a
a ? a
a a a a
a a a a
54
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Withdrawals explained
Spectrum bias avoided
a a a a a
a
a a a a
Table 23. Radiography and Serology Quality (Continued)
Disease progression bias avoided Uninterpretable/Intermediate test result(s)
reported Partial verification bias avoided Diagnostic review bias avoided
Appropriate reference standard Index test execution described Reference
standard execution described Clinical review bias avoided Differential
verification bias avoided Selection criteria described Incorporation bias
avoided Test review bias avoided
a = Yes a = No ? = Unclear
Author Savarino Pill Spangehl Bottner Di Cesare Bernard Trampuz
N 26 92 202 78 58 228 296
Index Test WBC (>9500/mm^3) WBC (>11000/μL) WBC differential
(>75% neutrophils) WBC (>6200/μL) WBC (>11.0*10^9) Polynuclear
neutrophil count (≥6000 cells/ml) WBC (>10x10^9/L)
Level of Evidence I I I I II II II
a a a a a ? ?a a a ? a a
a
a a a a a a a
a a a a a a a
a a a a a a a
a a a a a a a
a a a a a a a
a a a a a a a
a a a a a a a
a a a a a a a
? ? ? ? ? a
a a a a a a
a
a a a a a a a
55
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Withdrawals explained
Spectrum bias avoided
a a a a a a a
STUDY RESULTS Figure 5. ESR Meta-analysis Results - Likelihood Ratios
StudyId
StudyId
Savarino - 15mm/hr Greidanus - 22.5mm/hr Schinsky - 30mm/hr Kamme - 30mm/hr
Della Valle - 30mm/hr Bottner - 32mm/hr
Savarino - 15mm/hr Greidanus - 22.5mm/hr Schinsky - 30mm/hr Kamme - 30mm/hr
Della Valle - 30mm/hr Bottner - 32mm/hr
COMBINED 1 2 5 10
COMBINED .1 .2 .5
DLR NEGATIVE
1
DLR POSITIVE
INTERPRETING THE FOREST PLOTS Throughout the guideline we use descriptive
diagrams or forest plots to present data from studies comparing a diagnostic
test to a reference standard. The positive and negative likelihood ratios are
the effect measures used to depict study results. The horizontal line running
through each point represents the 95% confidence interval for that point. In
the graphs above, the solid vertical line represents no effect where the likelihood
ratio is equal to one. The dashed line represents the value of a likelihood
ratio that indicates a large and conclusive change in the probability of
disease. For example, in the figures above, the summary estimate (indicated by
the diamond) indicates a positive likelihood ratio close to 3 and a negative
likelihood ratio of between 0.2 and 0.1. This result is statistically
significant because the 95% Confidence Interval does not cross the no effect
line. Also please note that summary measures (diamond) are not reported in the
plot if high heterogeneity (>50%) is present.
56
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Figure 6. CRP Results - Likelihood Ratios
StudyId StudyId
Savarino- 0.5mg/dL Della Valle - 1 mg/dL Schinsky - 1 mg/dL Greidanus - 1 mg/dL
Fink - 1.35 mg/dL Greidanus - 1.35 mg/dL Bottner - 1.5 mg/dL Savarino - 2 mg/dL
Bottner - 3.2 mg/dL
Savarino - 0.5mg/dL Della Valle - 1 mg/dL Schinsky - 1 mg/dL Greidanus - 1
mg/dL Fink - 1.35 mg/dL Greidanus - 1.35 mg/dL Bottner - 1.5 mg/dL Savarino - 2
mg/dL Bottner - 3.2 mg/dL
1 2 5 10
DLR POSITIVE
.1.2
.5
1
DLR NEGATIVE
Figure 7. ESR and CRP - Likelihood Ratios
StudyId
StudyId
Greidanus both positive (22.5/1.35)
Greidanus both positive (22.5/1.35)
Greidanus both positive (30/1.0)
Greidanus both positive (30/1.0)
Schinsky - both positive (30/1.0)
Schinsky - both positive (30/1.0)
Greidanus 1 or both positive (22.5/1.35)
Greidanus 1 or both positive (22.5/1.35)
Greidanus 1 or both positive (30/1.0)
Greidanus 1 or both positive (30/1.0)
Schinsky 1 or both positive (30/1.0)
Schinsky 1 or both positive (30/1.0)
1
2
5
10
.1.2
.5
1
DLR POSITIVE
DLR NEGATIVE
57
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Figure 8. Radiography Results - Likelihood Ratios
StudyId StudyId
Barrack - periostitis Barrack - radiolucency grade Barrack - gross loosening
Barrack - radiolucent line Bernay - plain radiograph Cyteval - periostitis
Cyteval - nonfocal lucency Cyteval - focal lucency Cyteval - bone abnormalities
Cyteval - femoral head asymmetry
Barrack - periostitis Barrack - radiolucency grade Barrack - gross loosening
Barrack - radiolucent line Bernay - plain radiograph Cyteval - periostitis
Cyteval - nonfocal lucency Cyteval - focal lucency Cyteval - bone abnormalities
Cyteval - femoral head asymmetry
1 2
5 10
.1 .2
.5
1
DLR POSITIVE
DLR NEGATIVE
Figure 9. White Blood Cell Count Results - Likelihood Ratios
StudyId StudyId
Bernard Trampuz Di Cesare Bottner Spangehl - %neutrophils Pill Savarino
Spangehl - WBC count
Bernard Trampuz Di Cesare Bottner Spangehl - % neutrophils Pill Savarino
Spangehl - WBC count
1 2
5 10
.1 .2
0.5
1
DLR POSITIVE
DLR NEGATIVE
58
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 24. Erythrocyte Sedimentation Rate Results
Level of Evidence I I I Author N Test ESR (15mm/hr) ESR (15mm/hr) ESR (15mm/hr)
Reference Standard Intraoperative Cultures Histology Intraoperative cultures
and histology at least 2 of: 1)a positive intraoperative culture (on solid
media) 2)gross purulence 3)final histopathological result consistent with
infection (average of >10 PMN in the 5 most cellular high power fields)
Intraoperative Cultures Joint Positive Likelihood Ratio (95% CI) 0.94 (0.48,
1.82) 1.02 (0.53, 1.97) Negative Likelihood Ratio (95% CI) 1.09 (0.43, 2.8)
0.97 (0.39, 2.4) Sensitivity (95% CI) 0.56 (0.3, 0.8) 0.58 (0.28, 0.85) 0.6
(0.26, 0.88) Specificity TP FN FP (95% CI) 0.4 (0.12, 0.74) 0.43 (0.18, 0.71)
0.44 (0.2, 0.7) 9 7 6 7 5 4 6 8 9 TN
Savarino Savarino Savarino
26 26 26
Hip Hip Hip
4 6 7
1.07 0.91 (0.55, 2.08) (0.36, 2.34)
I
Schinsky 201
ESR (30 mm/hr)
Hip
1.58 0.09 (1.37, 1.82) (0.02, 0.37)
0.96 (0.87, 1)
0.39 (0.31, 0.47)
53
2
89
57
I
Savarino
26
ESR (50 mm/hr)
Hip
3.75 (0.53, 26.73)
0.69 (0.45, 1.07)
0.38 (0.15, 0.65)
0.9 (0.55, 1)
6
10
1
9
59
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 24. Erythrocyte Sedimentation Rate Results (Continued)
Level of Evidence I I I I I Author N Test ESR (50 mm/hr) ESR (50 mm/hr) ESR
(30mm/hr) ESR (22.5 mm/hr) ESR (30 mm/hr) Reference Standard Histology
Intraoperative cultures and histology Intraoperative Cultures Cultures
Intraoperative or Aspiration Cultures Intraoperative or Aspiration Joint
Positive Likelihood Ratio (95% CI) 7 (0.97, 50.27) 9.6 (1.35,68.43) Negative
Likelihood Ratio (95% CI) 0.54 (0.3, 0.97) 0.43 (0.2, 0.92) Sensitivity (95%
CI) 0.5 (0.21, 0.79) 0.6 (0.26, 0.88) 0.89 (0.75, 0.97) 0.93 (0.82, 0.99) 0.82
(0.68, 0.92) Specificity TP FN FP (95% CI) 0.93 (0.66, 1) 0.94 (0.7, 1) 0.72
(0.51, 0.88) 0.83 (0.74, 0.9) 0.88 (0.8, 0.93) 6 6 34 42 37 6 4 4 3 8 1 1 7* 18
13 TN
Savarino Savarino Kamme
26 26 63
Hip Hip Hip Knee Knee
13 15 18 88 93
3.2 0.15 (1.69, 6.05) (0.06, 0.38) 5.5 0.08 (3.58, 8.43) (0.03, 0.24) 6.7
(3.96, 11.36) 0.2 (0.11, 0.38)
Greidanus 151 Greidanus 151
60
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 24. Erythrocyte Sedimentation Rate Results (Continued)
Level of Evidence Author N Test Reference Standard at least 2 of 3 positive
intraoperative cultures on solid media or if 2 of following: 1)at least 1
positive culture 2)final histopathology consistent with infection 3)gross
purulence seen at time of revision Intraoperative cultures and histology Joint
Positive Likelihood Ratio (95% CI) Negative Likelihood Ratio (95% CI)
Sensitivity (95% CI) Specificity TP FN FP (95% CI) TN
I
Della Valle
94
ESR (30 mm/hr)
Knee
2.66 (1.8, 3.92)
0.15 (0.06, 0.38)
0.9 (0.77, 0.97)
0.66 (0.52, 0.78)
37
4
18
35
Mixed (50 7.69 0.81 0.89 0.21 I Bottner 78 17 4 6 hip, (3.51, (0.58, (0.78,
(0.09, 0.52) 28 16.86) 0.95) 0.96) knee) *Five false positives were from
patients with rheumatoid arthritis, one malignancy, and one with elevated ESR
prior to primary operation ESR (32 mm/hr)
51
61
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 25. C-Reactive Protein Results
Level of Evidence I I I Author N Test CRP (0.5mg/dL) CRP (0.5mg/dL) CRP
(0.5mg/dL) Reference Standard Intraoperative Cultures Histology Intraoperative
cultures and histology at least 2 of: 1)a positive intraoperative culture (on
solid media) 2)gross purulence 3)finalhistopathological result consistent with
infection (average of >10 PMN in the 5 most cellular high power fields)
Intraoperative Cultures Positive Likelihood Joint Ratio (95% CI) Hip Hip Hip
1.25 (0.4, 3.9) 0.93 (0.32, 2.71) 0.8 (0.26, 2.5) Negative Likelihood Ratio
(95% CI) 0.89 (0.51, 1.56) 1.04 (0.59, 1.81) 1.12 (0.64, 1.95) Sensitivity (95%
CI) 0.38 (0.15, 0.65) 0.33 (0.1, 0.65) 0.3 (0.07, 0.65) Specificity TP FN FP
(95% CI) 0.7 (0.35, 0.93) 0.64 (0.35, 0.87) 0.63 (0.35, 0.85) 6 4 3 10 8 7 3 5
6 TN
Savarino Savarino Savarino
26 26 26
7 9 10
I
Schinsky
201
CRP (1 mg/dL)
Hip
3.29 (2.53, 4.28)
0.08 (0.03, 0.23)
0.95 (0.85, 0.99)
0.71 (0.63, 0.78)
52
3
42
104
I
Savarino
26
CRP (2 mg/dL)
Hip
0.94 (0.19, 4.67)
1.02 (0.69, 1.5)
0.19 (0.04, 0.46)
0.8 (0.44, 0.97)
3
13
2
8
62
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 25. C-Reactive Protein Results (Continued)
Level of Evidence I I I I I Author N Test CRP (2 mg/dL) CRP (2 mg/dL) CRP (1.0
mg/dL) CRP (1.35 mg/dL) CRP (1.35 mg/dL) Reference Standard Histology
Intraoperative cultures and histology Cultures Intraoperative or Aspiration
Cultures Intraoperative or Aspiration Intraoperative cultures and histology
Joint Positive Likelihood Ratio (95% CI) 4.67 (0.6, 36.29) 2.4 (0.48, 11.95)
5.5 (3.58, 8.43) 6.9 (4.2, 11.33) 3.81 (2.46, 5.9) Negative Likelihood Ratio
(95% CI) 0.72 (0.47, 1.1) 0.8 (0.51, 1.25) 0.08 (0.03, 0.24) 0.1 (0.04, 0.26)
0.34 (0.2, 0.57) Sensitivity (95% CI) 0.33 (0.1, 0.65) 0.3 (0.07, 0.65) 0.93
(0.82, 0.99) 0.91 (0.79, 0.98) 0.73 (0.56, 0.85) Specificity (95% CI) 0.93
(0.66, 1) 0.88 (0.62, 0.98) 0.83 (0.74, 0.9) 0.87 (0.79, 0.93) 0.81 (0.72,
0.88) TP FN FP TN
Savarino Savarino
26 26
Hip Hip Knee Knee Knee
4 3 42 41 29
8 7 3 4 11
1 2 18 14 20
13 14 88 92 85
Greidanus 151 Greidanus 151 Fink 145
63
AAOS ClinicalPractice Guidelines Unit v0.1 03.25.10
Table 25. C-Reactive Protein Results (Continued)
Level of Evidence Author N Test Reference Standard at least 2 of 3 positive
intraoperative cultures on solid media or if 2 of following: 1)at least 1
positive culture 2)final histopathology consistent with infection 3)gross
purulence seen at time of revision Intraoperative cultures and histology
Intraoperative cultures and histology Joint Positive Likelihood Ratio (95% CI)
Negative Likelihood Ratio (95% CI) Sensitivity (95% CI) Specificity (95% CI) TP
FN FP TN
I
Della Valle
94
CRP (1 mg/dL)
Knee
3.88 (2.41, 6.25)
0.06 (0.02, 0.25)
0.95 (0.83, 0.99)
0.75 (0.62, 0.86)
39
2
13
40
I
Bottner
78
CRP (1.5 mg/dL) CRP (3.2 mg/dL)
I
Bottner
78
Mixed (50 hip, 28 knee) Mixed (50 hip, 28 knee)
10.86 (4.67, 25.22) 27.14 (6.93, 106)
0.05 (0.01, 0.35) 0.05 (0.01, 0.33)
0.95 (0.76, 1) 0.95 (0.76, 1)
0.91 (0.81, 0.97) 0.96 (0.88, 1)
20
1
5
52
20
1
2
55
64
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 26. ESR and CRP Results
Level of Evidence Author N Test ESR and CRP positive if both positive
(22.5/1.35) ESR and CRP positive if both positive (30/1.0) ESR and CRP
positive if both positive (30/1.0) ESR and CRP positive if one positive
(22.5/1.35) Reference Standard Joint Positive Likelihood Ratio (95% CI) 13.5
(6.53, 27.7) Negative Likelihood Ratio (95% CI) 0.12 (0.05, 0.27) Sensitivity
(95% CI) Specificity (95% CI) TP FN FP TN
I
Greidanus 151
Cultures Intraoperative or Aspiration
Knee
0.89 (0.76, 0.96)
0.93 (0.87, 0.97)
40
5
7
99
I
Greidanus 151
Cultures Intraoperative or Aspiration At least 2 of: 1)positive
intraoperative culture (on solid media) 2)gross purulence 3)final
histopathology Cultures Intraoperative or Aspiration
Knee
12.1 (5.83,25.2)
0.21 (0.12, 0.38)
0.8 (0.65, 0.90)
0.93 (0.87, 0.97)
36
9
7
99
I
Schinsky
201
Hip
4.34 (3.11, 6.04)
0.14 (0.06, 0.26)
0.89 (0.78, 0.96)
0.79 (0.72, 0.86)
49
6
30
116
I
Greidanus 151
Knee
4.22 (2.95, 6.03)
0.06 (0.01, 0.22)
0.96 (0.85, 0.99)
0.77 (0.68, 0.85)
43
2
24
82
65
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 26. ESR and CRP Results (Continued)
Level of Evidence Author N Test ESR and CRP positive if one positive (30/1.0)
ESR and CRP positive if one positive (30/1.0) Reference Standard Joint
Positive Likelihood Ratio (95% CI) 4.22 (2.95, 6.03) Negative Likelihood Ratio
(95% CI) 0.06 (0.01, 0.22) Sensitivity (95% CI) Specificity (95% CI) TP FN FP
TN
I
Greidanus 151
Cultures Intraoperative or Aspiration At least 2 of: 1)positive
intraoperative culture (on solid media) 2)gross purulence 3)final
histopathology
Knee
0.96 (0.84, 0.99)
0.77 (0.68, 0.85)
43
2
24
82
I
Schinsky
201
Hip
1.74 (1.51, 2.0)
0
1 (0.94, 1)
0.43 (0.34, 0.51)
55
0
84
62
66
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 27. Radiography Results
Level of Evidence Author N Test Radiograph asymmetric position of femoral head
Radiograph bone abnormalities (focal or nonfocal lucency, periostitis)
Radiograph focal lucency Radiograph nonfocal lucency Radiograph periostitis
Reference Standard Joint Positive Likelihood Ratio (95% CI) 0.52 (0.14, 1.95)
Negative Likelihood Ratio (95% CI) 1.23 (0.9, 1.68) Sensitivity (95% CI) 0.17
(0.02, 0.48) Specificity (95% CI) TP FN FP TN
I
Cyteval 65
Intraoperative Cultures (≥2)
Hip
0.68 (0.54, 0.8)
2
10
17
36
I
Cyteval 65
Intraoperative Cultures (≥2)
Hip
1.05 (0.72, 1.51)
0.88 (0.3, 2.57)
0.75 (0.43, 0.95)
0.28 (0.17, 0.42)
9
3
3815
I
Cyteval 65
Intraoperative Cultures (≥2) Intraoperative Cultures (≥2) Intraoperative
Cultures (≥2) Pathological and Gross Operative Findings
Hip
1.2 (0.4, 3.66) 0.98 (0.53, 1.83) 20.77 (1.06, 407) 2.4 (1.22, 4.74)
0.95 (0.66, 1.35) 1.02 (0.54, 1.91) 0.82 (0.62, 1.06) 0.3 (0.08, 1.07)
0.25 (0.05, 0.57) 0.5 (0.21, 0.79) 0.17 (0.02, 0.48) 0.8 (0.44, 0.97)
0.79 (0.66, 0.89) 0.49 (0.35, 0.63) 1 (0.93, 1) 0.67 (0.43, 0.85)
3
9
11
42
I
Cyteval 65
Hip
6
6
27
26
I
Cyteval 65
Hip
2
10
0
53
II
Bernay
31
Radiograph
Hip
8
2
7
14
67
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 27. Radiography Results (Continued)
Level of Evidence Author N Test Reference Standard Intraoperative cultures on
solid media, aspiration culture confirmed by intraoperative culture in liquid
media, or intraoperative histology Intraoperative cultures on solid media,
aspiration culture confirmed by intraoperative culture in liquid media, or
intraoperative histology Joint Positive Likelihood Ratio (95% CI) Negative
Likelihood Ratio (95% CI) Sensitivity (95% CI) Specificity (95% CI) TP FN FP TN
III
Barrack 69
Radiograph: complete radiolucent line adjacent to tibial component
Knee
1.96 (0.59, 6.56)
0.89 (0.7, 1.13)
0.2 (0.06, 0.44)
0.9 (0.78, 0.97)
4
16
5
44
III
Barrack 69
Radiograph: gross loosening
Knee
0.92 (0.27, 3.11)
1.02 (0.81, 1.27)
0.15 (0.03, 0.38)
0.84 (0.7, 0.93)
3
17
8
41
68
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 27. Radiography Results (Continued)
Level of Evidence Author N Test Radiograph: Knee Society Grade 3 classification
for radiolucency for the tibial component in addition to Grade 1 under femoral
component Reference Standard Intraoperative cultures on solid media, aspiration
cultureconfirmed by intraoperative culture in liquid media, or intraoperative
histology Intraoperative cultures on solid media, aspiration culture confirmed
by intraoperative culture in liquid media, or intraoperative histology Joint
Positive Likelihood Ratio (95% CI) Negative Likelihood Ratio (95% CI)
Sensitivity (95% CI) Specificity (95% CI) TP FN FP TN
III
Barrack 69
Knee
2.99 (1.47, 6.1)
0.55 (0.33, 0.91)
0.55 (0.32, 0.77)
0.82 (0.68, 0.91)
11
9
9
40
III
Barrack 69
Radiograph: periostitis
Knee
1.04 (0.73, 1.47)
0.92 (0.42, 2.01)
0.7 (0.46, 0.88)
0.33 (0.2, 0.48)
14
6
33
16
69
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 28. White Blood Cell Count Results
Level of Evidence Author N Test Reference Standard at least 1 of: 1)open wound
of sinus in communication with the joint 2)systemic infection with pain in the
joint and purulent fluid within the joint 3)positive result on at least 3
investigations(ESR>30, CRP>10, preoperative aspiration with at least 1
positive culture, frozen section with >5PMN/HPF, intraoperative culture
(>1/3 of cultures positive) Intraoperative cultures Histology Intraoperative
cultures and histology Positive Likelihood Joint Ratio (95% CI) Negative
Likelihood Ratio (95% CI) Sensitivity (95% CI) Specificity TP FN FP (95% CI) TN
I
Spangehl 202
WBC (11.0x10^9/L
Hip
5.57 (1.99, 15.56)
0.83 (0.7, 0.98)
0.2 (0.08, 0.37)
0.96 (0.92, 0.99)
7
28
6
161
I I I
Savarino Savarino Savarino
26 26 26
WBC (9500/mm^3) WBC (9500/mm^3) WBC (9500/mm^3)
Hip Hip Hip
1.94 (0.09, 43.5) 3.46 (0.15, 77.86) 4.64 (0.21, 103.9)
0.96 (0.79, 1.16) 0.92 (0.74, 1.14) 0.89 (0.69, 1.14)
0.06 (0, 0.3) 0.08 (0, 0.38) 0.1 (0, 0.45)
1 (0.69, 1) 1 (0.77, 1) 1 (0.79, 1)
1 1 1
15 11 9
0 0 0
10 14 16
70
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 28. White Blood Cell Count Results (Continued)
Level of Evidence Author N Test Reference Standard at least 1 of 3 criteria:
1)an open wound or sinus communicating with the joint 2) systemic infection
with pain in the hip and purulent fluid within the joint 3)positive result on
at least 3 tests (ESR, CRP, joint aspiration, intraoperative frozen section,
and intraoperative culture) Joint Positive Likelihood Ratio (95% CI) Negative
Likelihood Ratio (95% CI) Sensitivity Specificity (95% CI) (95% CI) TP FN FP TN
I
Pill
92
WBC (11000/μL)
Hip
8.45 (1.77, 40.46)
0.78 (0.62, 1)
0.24 (0.08, 0.47)
0.97 (0.9, 1)
5
16
2
69
71
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 28. White Blood Cell Count Results (Continued)
Level of Evidence Author N Test Reference Standard at least 1 of: 1)open wound
of sinus in communication with the joint 2)systemic infection with pain in the
joint and purulent fluid within the joint 3)positive result on at least 3
investigations(ESR>30, CRP>10, preoperative aspiration with at least 1
positive culture, frozen section with >5PMN/HPF, intraoperative culture
(>1/3 of cultures positive) Intraoperative cultures and histology
Intraoperative cultures and histology Joint Positive Likelihood Ratio (95% CI)
Negative Likelihood Ratio (95% CI) Sensitivity Specificity (95% CI) (95% CI) TP
FN FP TN
I
Spangehl 202
WBC differential (75% neutrophils)
Hip
2.01 (0.96, 4.22)
0.87 (0.72, 1.05)
0.23 (0.1, 0.4)
0.89 (0.83, 0.93)
8
27
19
148
I
Bottner
78
WBC (6200/μL) WBC (11.0x10^9)
Mixed (50 hip, 28 knee) Mixed
1.77 (1.17, 2.68) 4.29 (2.38, 7.73)
0.48 (0.24, 0.97) 0.08 (0.01, 0.51)
0.71 (0.48, 0.89) 0.94 (0.71, 1)
0.6 (0.46, 0.72) 0.78 (0.62, 0.89)
15
6
23
34
II
Di Cesare
58
16
1
9
32
72
AAOS Clinical Practice Guidelines Unitv0.1 03.25.10
Table 28. White Blood Cell Count Results (Continued)
Level of Evidence Author N Test Reference Standard at least 1 of: 1)visible
purulence of synovial fluid or area surrounding the prosthesis 2)acute
inflammation on histopathologic exam of permanent periprosthetic tissue
sections (as determined by the clinical pathologist) 3)a sinus tract
communicating with the prosthesis Intraoperative cultures Joint Positive
Likelihood Ratio (95% CI) Negative Likelihood Ratio (95% CI) Sensitivity
Specificity (95% CI) (95% CI) TP FN FP TN
II
Trampuz
296
WBC count (10x10^9/L)
Mixed (207 knee, 124 hip)
3.11 (1.51, 6.4)
0.87 (0.78, 0.97)
0.18 (0.1, 0.29)
0.94 (0.9, 0.97)
13
59
13
211
II
Bernard
228
Polynuclear neutrophil count (6000 cells/ml)
Mixed (167 hip, 63 knee)
2.84 (1.17, 6.92)
0.57 (0.44, 0.73)
0.54 (0.47, 0.61)
0.81 (0.58, 0.95)
112
95
4
17
73
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
RECOMMENDATION 3
We recommend joint aspiration of patients being assessed for periprosthetic
knee infections that have abnormal erythrocyte sedimentation rate AND/OR
C-reactive protein results. We recommend that the aspirated fluid be sent for
microbiologic culture, synovial fluid white blood cell count and differential.
Strength of Recommendation: Strong Rationale Our systematic review of the
literature suggests that ESR/CRP testing is valuable for screening (ruling out)
of periprosthetic infection with extremely high sensitivity (>90%). These
tests are not, however, specific for diagnosis of periprosthetic infection and
may be elevated with any type of infection or inflammation. Hence, for patients
with abnormal ESR/CRP who are being investigated for periprosthetic infection
of the knee, the most appropriate next test is aspiration of the knee joint. We
recommend that the fluid obtained from the joint besent for analysis of
synovial fluid white blood cell count, percentage of neutrophils, and also
culture for aerobic and anerobic organisms. Studies suggest either synovial
fluid white blood cell count over 1700 cells/μl (range, 1100-3000) or
neutrophil percentage greater than 65% (range 64%80%) is highly suggestive of
chronic periprosthetic infection.21, 31, 103 However, the threshold for cell
count and neutrophil percentage indicative of acute periprosthetic joint infection
(within six weeks of index arthroplasty) is yet to be determined and the values
and ranges reported above may not be applicable when diagnosing acute
periprosthetic infections. Supporting Evidence Two studies with reliable data
addressed the diagnostic efficacy of aspiration cultures among patients being
assessed for periprosthetic knee infections.21, 31, 103 Both studies indicated
that this test is a good rule in test but only moderately good at ruling out
infection (positive likelihood ratio: 14.2-15.2, negative LR: 0.21-0.29; see
Table 32). Three studies with reliable data addressed synovial fluid white
blood cell count and differential among patients being assessed for
periprosthetic knee infections.21, 33 The studies indicated that both of these
tests are good rule in and rule out tests (synovial fluid white blood cell
count: LR+: 7.6-35.6, LR-: 0.01-0.11; differential: LR+: 6.5-48, LR-:
0.03-0.05; see Table 33, Table 34).
74
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
SUMMARY OF EVIDENCE Table 29. Knee Aspiration Summary of Evidence
Test Number of Studies Positive Likelihood Ratio (95% CI) 14.2 15.2 Negative
Likelihood Ratio (95% CI) 0.21 0.29 Sensitivity (95% CI) Specificity (95% CI)
Aspiration Cultures (range) Synovial fluid WBC Count (range) Synovial fluid
Neutrophil Percentage (range)
2
0.73 0.80
0.94 0.95
3
7.6 35.6
0.01 0.11
0.91 1.0
0.88 0.98
36.5 48.0
0.03 0.05
0.95 0.98
0.85 0.98
75
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
EXCLUDED ARTICLES Table 30. Excluded Articles - Recommendation 3
Author Bach, et al. 2002 Barrack, et al. 1997 Duff, et al. 1996 Kersey, et al.
2000 Mason, et al. 2003 Parvizi, et al. 2006 Parvizi, et al. 2008 Trampuz, et
al. 2006 Trampuz, et al. 2007 Van den Bekerom, et al. 2006 Virolainen, et al.
2002 Title Total knee arthroplasty infection: significance of delayed
aspiration The Coventry Award. The value of preoperative aspiration before
total knee revision Aspiration of the knee joint before revision arthroplasty
White blood cell counts and differential in synovial fluid of aseptically
failed total knee arthroplasty The value of white
blood cell counts before revision total knee arthroplasty Periprosthetic
infection: What are the diagnostic challenges? Diagnosis of infected total
knee: findings of a multicenter database Sonication of explanted prosthetic
components in bags for diagnosis of prosthetic joint infection is associated
with risk of contamination Sonication of removed hip and knee prostheses for
diagnosis of infection The value of pre-operative aspiration in the diagnosis
of an infected prosthetic knee: a retrospective study and review of literature
The reliability of diagnosis of infection during revision arthroplasties Reason
for Exclusion Does not address recommendation Not best available evidence Not
best available evidence Does not address recommendation Not best available
evidence Not best available evidence Not best available evidence Not best
available evidence Not best available evidence Not best available evidence Not
best available evidence
76
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
STUDY QUALITY Table 31. Knee Aspiration - Quality
Disease progression bias avoided Uninterpretable/Intermediate test result(s)
reported Partialverification bias avoided Diagnostic review bias avoided
Appropriate reference standard Index test execution described Reference
standard execution described Clinical review bias avoided Differential
verification bias avoided Selection criteria described Incorporation bias
avoided Test review bias avoided
a = Yes a = No ? = Unclear
Author Della Valle Fink Della Valle Ghanem Trampuz Della Valle Ghanem Trampuz
N 94 145 94 429 133 94 429 133
Index Test Aspiration culture Aspiration culture Synovial fluid WBC count
(>3000/μL) Synovial fluid WBC count (>1100/μL) Synovial fluid
WBC count (>1700/μL) Synovial fluid WBC differential (>65% PMN)
Synovial fluid % neutrophil (>64%) Synovial fluid % neutrophil (>65%)
Level of Evidence I I I I I I I I
a a a a ? a a ?
a a a a a a a a
a a a a a a a a
a a a a a a a a
a a a a a a a a
a a a a a a a a
a a a a a a a a
a a a a a a a a
a a a a a a a a
a a a a a a a a
? ? ? ? ? ? ? ?
a a a a a a
a a
a a a a a a a a
77
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Withdrawals explained
Spectrum bias avoided
a a a a a a a
a
STUDY RESULTS Table 32. Aspiration Culture - Knee
Level of Evidence Author N Test Reference Standard At least 2 of 3 positive
intraoperative cultures on solid media or 2 of following: 1)at least 1 positive
culture 2)final histopathology consistent with infection 3)gross purulence at
revision Intraoperative Cultures and Histology Joint Positive Likelihood Ratio
(95% CI) Negative Likelihood Ratio (95% CI) Sensitivity (95% CI) Specificity
(95% CI) TP FP FN TN
I
Della Valle
94
Aspiration culture on solid media
Knee
14.22 (4.69, 43.12)
0.21 (0.11, 0.39)
0.8 (0.65, 0.91)
0.94 (0.84, 0.99)
3
3
8
50
I
Fink
145
Aspiration culture
Knee
15.23 (6.34, 36.58)
0.29 (0.17, 0.48)
0.73 (0.56, 0.85)
0.95 (0.89, 0.98)
295
11
100
78
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 33. Synovial Fluid White Blood Cell Count
Level of Evidence Author N Test Reference Standard At least 2 of 3 positive
intraoperative cultures on solid media or 2 of following: 1)at least 1 positive
culture 2)final histopathology consistent with infection 3)gross purulence at
revision Joint Positive Likelihood Ratio (95% CI) Negative Likelihood Ratio
(95% CI) Sensitivity (95% CI) Specificity (95% CI) TP FP FN TN
I
Della Valle
94
WBC (3.0x10^3/μL)
Knee
35.57 (7.34, 172.4)
0.01 (0, 0.19)
1 (0.91, 1)
0.98 (0.9, 1)
41
0
1
52
79
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 33. Synovial Fluid White Blood Cell Count (Continued)
Level of Evidence Author N Test Reference Standard at least 1 of 3 criteria:
1)presence of an abscess or sinus tract communicating with the joint space
2)positive culture of aspirate on solid medium 3)≥2 positive
intraoperative cultures of the same organism, or one positive culture on solid
medium and the presence of gross intracapsular purulence or abnormal
histological findings; when cultures were negative, infection was present if
had both grossly purulent fluid and an abnormal frozen section (Mirra et al.
criteria) Joint Positive Likelihood Ratio (95% CI) Negative Likelihood Ratio
(95% CI) Sensitivity (95% CI) Specificity (95% CI) TP FP FN TN
I
Ghanem
WBC 429 (1.1x10^3 /μL)
Knee
7.59 (5.47, 10.55)
0.11 (0.07, 0.17)
0.91 (0.85, 0.95)
0.88 (0.84, 0.92)
146
32
15
236
80
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 33. Synovial Fluid White Blood Cell Count (Continued)
Level of Evidence Author N Test Reference Standard Joint Positive Likelihood
Ratio (95% CI) Negative Likelihood Ratio (95% CI) Sensitivity (95% CI) Specificity
(95% CI) TP FP FN TN
I
at least 1 of thefollowing criteria: growth of the same microorganism in at
least 2 cultures of synovial fluid or periprosthetic tissue; visible synovial
fluid purulence at the time of 7.76 WBC 0.07 0.94 0.88 arthrocentesis or during
(4.54, 32 12 2 87 Knee Trampuz 133 (1.7x10^3 (0.02, 0.26) (0.8, 0.99) (0.8,
0.94) surgery; acute 13.28) /μL) inflammation on histopathologic
examination of permanent periprosthetic tissue sections; or presence of a sinus
tract communicating with the prosthesis 49, 66 Note: Authors of other articles
investigating synovial fluid cell counts who expressed counts in milliliters
(ml) were contacted; they confirmed that the units reported for the cell count
in the published report were incorrect. The actual units to indicate
periprosthetic infection were microliters 90 (΅l) and fall within the limits
cited above. Although the units to express cell count differs
between published reports, the American
College of 90
Rheumatology (ACR) recommends that cell count be expressed as cells/μl.
81
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 34. Synovial Fluid Neutrophil Percentage
Level of Evidence Author N Test Reference Standard At least 2 of 3 positive
intraoperative cultures on solid media or 2 of following: 1)at least 1 positive
culture 2)final histopathology consistent with infection 3)gross purulence at
revision Joint Positive Likelihood Ratio (95% CI) Negative Likelihood Ratio
(95% CI) Sensitivity (95% CI) Specificity (95% CI) TP FP FN TN
I
Della Valle
94
>65% PMN
Knee
6.46 (3.41, 12.26)
0.03 (0, 0.2)
0.98 (0.87, 1)
0.85 (0.72, 0.93)
40
8
1
45
82
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 34. Synovial Fluid Neutrophil Percentage (Continued)
Level of Evidence Author N Test Reference Standard at least 1 of 3 criteria:
1)presence of an abscess or sinus tract communicating with the joint space
2)positiveculture of aspirate on solid medium 3)≥2 positive
intraoperative cultures of the same organism, or one positive culture on solid
medium and the presence of gross intracapsular purulence or abnormal
histological findings; when cultures were negative, infection was present if
had both grossly purulent fluid and an abnormal frozen section (Mirra et al.
criteria) Joint Positive Likelihood Ratio (95% CI) Negative Likelihood Ratio
(95% CI) Sensitivity (95% CI) Specificity (95% CI) TP FP FN TN
I
Ghanem
429
> 64% Neutrophils
Knee
18.19 (10.91, 30.33)
0.05 (0.03, 0.1)
0.95 (0.9, 0.98)
0.95 (0.91, 0.97)
153
14
8
254
83
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 34. Synovial Fluid Neutrophil Percentage (Continued)
Level of Evidence Author N Test Reference Standard at least 1 of the following
criteria: growth of the same microorganism in at least 2 cultures of synovial
fluid or periprosthetic tissue; visible synovial fluid purulence at the time of
arthrocentesis or during surgery; acute inflammation on histopathologic
examination of permanent periprosthetic tissue sections; or presence of a sinus
tract communicating with the prosthesis Joint Positive Likelihood Ratio (95%
CI) Negative Likelihood Ratio (95% CI) Sensitivity (95% CI) Specificity (95%
CI) TP FP FN TN
I
Trampuz 133
> 65% Neutrophils
Knee
48.04 (12.17, 189.65)
0.03 (0, 0.21)
0.97 (0.85, 1)
0.98 (0.93, 1)
33
2
1
97
84
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
RECOMMENDATION 4
We recommend a selective approach to aspiration of the hip based on the
patients probability of periprosthetic joint infection and the results of the
erythrocyte sedimentation rate (ESR) AND C-reactive protein (CRP). We recommend
that the aspirated fluid be sent for microbiologic culture, synovial fluid
white blood cell count and differential. Selection ofPatients for Hip
Aspiration
Probability of Infection Higher Lower Lower Lower Higher or Lower ESR and CRP
Results + + or + − + + or + − ++ +− −− Planned
Reoperation Status Planned or not planned Planned Not planned Not planned
Planned or not planned Recommended Test Aspiration Aspiration or Frozen Section
Aspiration Please see Recommendation 6 No further testing
Key for ESR and CRP results + + = ESR and CRP test results are abnormal + −
= either ESR or CRP test result is abnormal − − = ESR and CRP test
results are normal
Strength of Recommendation: Strong Rationale AAOS conducted a systematic review
that identified six Level I hip,4, 27, 57, 65, 68, 111 and one hip and knee
study35 on the diagnostic performance of hip aspiration and culture. (Please
see Recommendation 3 for data supporting knee patients.) These studies did not
stratify patients as to whether they were at higher or lower probability for
infection, and all patients underwent re-operation. The indications for
patients having an aspiration varied between studies, with aspiration often a
routine part of preoperative investigations. Our meta-analysis indicated that
hip aspiration for culture has a moderate to large ability to rule in
infection but a small to moderate ability to rule out infection (positive
likelihood ratio 9.8, negative LR 0.33). Based on the meta-analysis, hip
aspiration is a useful test to diagnose periprosthetic hip infection.
Therefore, we recommend hip aspiration in all higher probability or lower
probability patients undergoing reoperation of the hip with abnormal ESR
and/or CRP results. 4, 27, 35, 57, 65, 68, 111 Given the potential problems
with instituting treatment and missing the diagnosis of infection, higher
probability patients with an abnormal ESR AND/OR CRP without planned
reoperation should also receive hip aspiration. There is no reliable evidence,
however, on the diagnosticperformance of hip aspirations in patients who are
not to undergo reoperation. Possible harms include the possibility of false
positive results, the possibility of the introduction of bacteria into the
joint during the procedure and patient pain and/or discomfort while undergoing
the procedure.4 There 85
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
are also concerns about the cost of the procedure.4 Therefore, universal hip
aspiration does not seem indicated. Aspiration is indicated for lower
probability hip patients without planned reoperation only if both ESR AND CRP
levels are abnormal. Lower probability hip arthroplasty patients without
planned reoperation who have an abnormal ESR OR CRP are addressed in
Recommendation 6. We do not recommend that higher or lower probability
patients with normal ESR and CRP have hip aspiration prior to planned reoperation. Supporting Evidence Six studies of hip
patients4, 27, 57, 65, 68, 111 and one study of mixed hip and knee (80% hip patients, all with reliable data, addressed the
diagnostic efficacy of hip aspiration cultures. Our meta-analysis of these
studies indicated that this test is a good rule in test, but not as good at
ruling out infection (LR+: 9.8, LR-: 0.33; see Figure 10). The indications for
patients having an aspiration varied between studies, with aspiration often a
routine part of preoperative investigations. Two studies reported that patients
were off antibiotics for at least two weeks prior to aspiration.4, 35 Three
studies reported injecting saline for re-aspiration,35, 68, 111 while three
studies reported not doing so.4, 27, 57 Two studies of hip patients with
reliable data addressed synovial fluid white blood cell count and
differential.96, 99 The data from these studies indicated that synovial fluid
white blood cell count is a good rule in and possibly a good rule out test
(LR+: 12-55, LR-: 0.18-0.65; see Figure 11).The higher threshold used in one study99
may account for its higher negative likelihood ratio (0.65). Neutrophil
percentage (both studies used an 80% threshold) was a moderately good rule in
and rule out test (LR+: 4.8-5.9, LR-: 0.130.22; see Figure 12). None of the
results presented includes data based on the use of diagnostic arthrography. SUMMARY OF EVIDENCE Table 35. Aspiration Summary of Evidence
Test Number of Studies Positive Likelihood Ratio (95% CI) Negative Likelihood
Ratio (95% CI) 0.33 (0.18, 0.60) 0.18 0.65 Sensitivity (95% CI) 0.69 (0.50,
0.84) 0.36 0.84 Specificity (95% CI) 0.93 (0.89, 0.95) 0.93 0.99
Aspiration 9.8 Cultures 7 (5.4, 17.9) (I2=0%) Synovial fluid 2 12.2 55.4 WBC
Count (range) Synovial fluid Neutrophil 2 4.8 5.9 Percentage (range) *Range
presented when fewer than four studies
0.13 0.22
0.82 0.89
0.83 0.85
86
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
EXCLUDED ARTICLES Table 36. Excluded Articles - Recommendation 4
Author Ali, et al. 2006 Bernard, et al. 2004 Buchholz, et al. 1981 Cheung, et
al. 1997 Chryssikos, et al. 2008 Dussault, et al. 1977 Fehring, et al. 1996
Gelman, 1976 Gould, et al. 1990 Guercio, et al. 1990 Itasaka, et al. 2001
Johnson, et al. 1988 Title Accuracy of joint aspiration for the preoperative
diagnosis of infection in total hip arthroplasty Value of preoperative
investigations in diagnosing prosthetic joint infection: retrospective cohort
study and literature review Management of deep infection of total hip
replacement The role of aspiration and contrast-enhanced arthrography in
evaluating the uncemented hip arthroplasty FDG-PET imaging can diagnose
periprosthetic infection of the hip Radiologic diagnosis of loosening and
infection in hip prostheses Aspiration as a guide to sepsis in revision total
hip arthroplasty Arthrography in total hip prosthesis complications Role of
routine percutaneouship aspirations prior to prosthesis revision Arthrography
of the prosthesetized painful hip: the importance of imaging and functional
testing Diagnosis of infection after total hip arthroplasty Detection of occult
infection following total joint arthroplasty using sequential technetium99m HDP
bone scintigraphy and indium-111 WBC imaging Reason for Exclusion Not best
available evidence Not best available evidence Not best available evidence Not
best available evidence Not best available evidence Insufficient Data Not best
available evidence Case report/commentary Not best available evidence Does not
address recommendation Not best available evidence Not best available evidence
87
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 36. Excluded Articles (Continued
Author Kraemer, et al. 1993 Levitsky, et al. 1991 Lieberman, et al. 1993 Lyons,
et al. 1985 Magnuson, et al. 1988 Maus, et al. 1987 McLaughlin, et al. 1977 Muller,
et al. 2008 Murray, et al. 1975 O'Neill, et al. 1984 Phillips, et al. 1983
Pons, et al. 1999 Title Bone scan, gallium scan, and hip aspiration in the
diagnosis of infected total hip arthroplasty Evaluation of the painful
prosthetic joint. Relative value of bone scan, sedimentation rate, and joint
aspiration Evaluation of painful hip arthroplasties. Are technetium bone scans
necessary? Evaluation of radiographic findings in painful hip arthroplasties
In-111-labeled leukocyte scintigraphy in suspected orthopedic prosthesis
infection: comparison with other imaging modalities Arthrographic study of
painful total hip arthroplasty: refined criteria Evaluation of the painful hip
by aspiration and arthrography Diagnosis of periprosthetic infection following
total hip arthroplasty - evaluation of the diagnostic values of pre- and
intraoperative parameters and the associated strategy to preoperatively select
patients with a high probability of joint infectionArthrography for the
assessment of pain after total hip replacement. A comparison of arthrographic
findings in patients with and without pain Failed total hip replacement:
assessment by plain radiographs, arthrograms, and aspiration of the hip joint
Efficacy of preoperative hip aspiration performed in the radiology department
Infected total hip arthroplasty--the value of intraoperative histology Reason
for Exclusion Not best available evidence Not best available evidence Not best
available evidence Not best available evidence Not best available evidence Insufficient
Data Insufficient Data Not best available evidence 4200/μl) I
Table 37. Aspiration Quality (Continued
Synovial fluid WBC count (>50000/μL) Synovial fluid WBC differential
(>80% PMN) Synovial fluid % neutrophils (>80%) I Level of Evidence I I a a a a a a a a a a a a
Spectrum bias avoided Selection criteria described Appropriate reference
standard
91
v0.1 03.25.10
a
a
a a a a a a a ? a a a
a a a a ? a
a ? a a a
Disease progression bias avoided
a a a a a a ? a
a a
a a a ? a a
? a a a
Partial verification bias avoided Differential verification bias avoided
Incorporation bias avoided Index test execution described Reference standard
execution described Test review bias avoided Diagnostic review bias avoided
Clinical review bias avoided Uninterpretable/Intermediate test result(s)
reported Withdrawals explained
STUDY RESULTS Figure 10. Hip Aspiration Cultures Results - Meta-Analysis
StudyId
StudyId
Glithero Barrack Malhotra Eisler Mulcahy Williams Lachiewicz
Glithero Barrack Malhotra Eisler Mulcahy Williams Lachiewicz
COMBINED 1 2 5 10
COMBINED .1 .2 .5
DLR NEGATIVE
1
DLR POSITIVE
92
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Figure 11. Synovial Fluid WBC Count Results - Likelihood Ratios (Hip and Knee)
StudyId StudyId
Ghanem (knee)- 1100/uL
Ghanem(knee) - 1100/uL
Trampuz (knee) - 1700/uL
Trampuz (knee) - 1700/uL
Della Valle (knee)- 3000/uL
Della Valle (knee) - 3000/uL
Schinsky (hip) - 4200/uL
Schinsky (hip) - 4200/uL
Spangehl (hip)- 50000/uL
Spangehl (hip) - 50000/uL
1 2 510
DLR POSITIVE
.1.2
.5
1
DLR NEGATIVE
Figure 12. Neutrophil Percentage Results Likelihood Ratios (Hip and Knee)
StudyId
StudyId
Della Valle (knee)
Della Valle (knee)
Trampuz (knee)
Trampuz (knee)
Ghanem (knee)
Ghanem (knee)
Schinsky (hip)
Schinsky (hip)
Spangehl (hip)
Spangehl (hip)
1
2
5
10
.1 .2
.5
DLR NEGATIVE
1
DLR POSITIVE
93
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 38. Aspiration Culture - Hip
Level of Evidence Author N Test Reference Standard Correlation between
intraoperative cultures and histology; the appearance of the tissue
intraoperatively; and the clinical course Correlation between intraoperative
cultures and histology; the appearance of the tissue intraoperatively; and the
clinical course Correlation between intraoperative cultures and histology; the
appearance of the tissue intraoperatively; and the clinical course Joint
Positive Likelihood Ratio (95% CI) Negative Likelihood Ratio (95% CI)
Sensitivity (95% CI) Specificity (95% CI) TP FP FN TN
I
Barrack 291
Aspiration culture
Hip
5.11 (2.81, 9.3)
0.45 (0.21, 0.97)
0.6 0.88 (0.26, 0.88) (0.84, 0.92)
6
33
4
248
I
Barrack 260
Aspiration (initial)
Hip
4 (1.42, 11.23)
0.57 (0.21, 1.52)
0.5 0.88 (0.07, 0.93) (0.83, 0.91)
2
33
2
224
I
Barrack
31
Aspiration (repeat)
Hip
16.67 (2.25, 123.39)
0.35 (0.11, 1.08)
0.67 (0.22, 0.96)
0.96 (0.8, 1)
4
1
2
24
94
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 38. Aspiration Culture (Continued)
Level of Evidence I AuthorN Test Aspiration culture Aspiration culture
Aspiration culture Aspiration culture Aspiration culture Aspiration culture
Reference Standard Intraoperative Cultures Intraoperative Cultures
Intraoperative Cultures or Gross Purulence Histology Intraoperative Cultures
and Histology Intraoperative Cultures Joint Positive Likelihood Ratio (95% CI)
2.16 (0.12, 39.05) 31.32 (4.51, 217) 27.47 (10.34, 73.01) 4.74 (1.29, 17.42)
7.56 (3.08, 18.58) 12.47 (7.28, 21.37) Negative Likelihood Ratio (95% CI) 0.94
(0.7, 1.27) 0.11 (0.03, 0.4) 0.15 (0.06, 0.38) 0.61 (0.34, 1.11) 0.34 (0.17,
0.71) 0.21 (0.13, 0.34) Sensitivity (95% CI) 0 (0, 0.6) 0.89 (0.67, 0.99) 0.85
(0.66, 0.96) 0.44 (0.14, 0.79) 0.69 (0.41, 0.89) 0.8 (0.69, 0.89) Specificity
(95% CI) 0.96 (0.87, 1) 0.97 (0.85, 1) 0.97 (0.92, 0.99) 0.91 (0.75, 0.98) 0.91
(0.8, 0.97) 0.94 (0.89, 0.97) TP FP FN TN
Eisler
57
Hip Mixed (43 hip, 11 knee) Hip
0
2
4
51
I
Glithero
54
17
1
2
34
I
Lachiewicz 156
23
4
4
125
I
Malhotra
41
Hip
4
3
5
29
I
Mulcahy
71
Hip
11
5
5
50
I
Williams
273
Hip
57
13
14
189
95
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 39. Synovial Fluid White Blood Cell Count
Level of Evidence Author N Test Reference Standard At least 2 of: 1)a positive
intraoperative culture (on solid media) 2)gross purulence 3)final
histopathology At least 2 of: 1)a positive intraoperative culture (on solid
media) 2)gross purulence 3)final histopathology At least 2 of: 1)a positive
intraoperative culture (on solid media) 2)gross purulence 3)final
histopathology Joint Positive Likelihood Ratio (95% CI) 12.21 (6.64, 22.46)
Negative Likelihood Ratio (95% CI) Sensitivity (95% CI) Specificity (95% CI) TP
FP FN TN
I
Schinsky 201
WBC (4.2x10^3/μl)
Hip
0.18 (0.1, 0.32)
0.84 (0.71, 0.92)
0.93 (0.88,0.97)
46
10
9
136
I
Schinsky
79
WBC (3.0x10^3/μl), among patients with elevated ESR and CRP WBC
(9.0x10^3/μl), among patients with elevated ESR and CRP
Hip
8.98 (3.06, 26.4)
0.11 (0.05, 0.26)
0.90 (0.78, 0.97)
0.90 (0.73, 0.98)
44
3
5
27
I
Schinsky
79
Hip
8.16 (2.77, 24.1)
0.20 (0.11, 0.37)
0.82 (0.68, 0.91)
0.90 (0.73, 0.98)
40
3
9
27
96
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 39. Synovial Fluid White Blood Cell Count (Continued)
Level of Evidence Author N Test WBC (3.0x10^3/μl), among patients with
elevated ESR or CRP, not both WBC (9.0x10^3/μl), among patients with
elevated ESR or CRP, not both Reference Standard At least 2 of: 1)a positive
intraoperative culture (on solid media) 2)gross purulence 3)final
histopathology At least 2 of: 1)a positive intraoperative culture (on solid
media) 2)gross purulence 3)final histopathology Positive Likelihood Joint Ratio
(95% CI) Negative Likelihood Ratio (95% CI) 0.19 (0.03, 1.15) Sensitivity (95%
CI) Specificity TP FP FN (95% CI) TN
I
Schinsky
60
Hip
6.43 (2.95, 14)
0.83 (0.36, 1)
0.87 (0.75, 0.95)
5
7
1
47
I
Schinsky
60
Hip
86.4 (5.3, 1402)
0.22 (0.05, 0.89)
0.83 (0.36, 1)
1 (0.93, 1)
5
0
1
54
97
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 39. Synovial Fluid White Blood Cell Count (Continued)
Level of Evidence Author N Test Reference Standard at least 1 of: 1)open wound
of sinus in communication with the joint 2)systemic infection with pain in the
joint and purulent fluid within the joint 3)positive result on at least 3
investigations(ESR, CRP, preoperative aspiration, frozen section, intraoperative
cultures Positive Likelihood Joint Ratio (95% CI) Negative Likelihood Ratio
(95% CI) Sensitivity (95% CI) Specificity TP FP FN (95% CI) TN
ISpangehl 183
WBC (5.0x10^4/μL)
Hip
55.36 (7.37, 415.6)
0.65 (0.49, 0.85)
0.36 (0.19, 0.56)
0.99 (0.96, 1)
10
1
18
154
98
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 40. Synovial Fluid Neutrophil Percentage
Level of Evidence Author N Test Reference Standard At least 2 of: 1)positive
intraoperative culture (on solid media) 2)gross purulence 3)final
histopathology At least 2 of: 1)positive intraoperative culture (on solid
media) 2)gross purulence 3)final histopathology Joint Positive Likelihood Ratio
(95% CI) Negative Likelihood Ratio (95% CI) Sensitivity (95% CI) Specificity
(95% CI) TP FP FN TN
I
Schinsky 201
>80% PMN
Hip
4.78 (3.27, 6.97)
0.22 (0.12, 0.39)
0.82 (0.69, 0.91)
0.83 (0.76, 0.89)
45
25
10
121
I
Schinsky
79
>80% PMN, among patients with elevated ESR and CRP
Hip
8.78 (2.98, 25.8)
0.14 (0.06, 0.29)
0.88 (0.75, 0.95)
0.90 (0.73, 0.98)
43
3
6
27
99
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 40. Synovial Fluid Neutrophil Percentage (Continued)
Level of Evidence Author N Test Reference Standard at least 1 of: 1)open wound
of sinus in communication with the joint 2)systemic infection with pain in the
joint and purulent fluid within the joint 3)positive result on at least 3
investigations (ESR, CRP, preoperative aspiration, frozen section,
intraoperative cultures Joint Positive Likelihood Ratio (95% CI) Negative
Likelihood Ratio (95% CI) Sensitivity (95% CI) Specificity (95% CI) TP FP FN TN
I
Spangehl 181
>80% Neutrophils
Hip
5.94 (3.99, 8.84)
0.13 (0.04, 0.37)
0.89 (0.72, 0.98)
0.85 (0.78, 0.9)
25
23
3
130
100
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
RECOMMENDATION 5
We suggest a repeat hip aspiration when there is a discrepancy between the
probability of periprosthetic joint infectionand the initial aspiration culture
result. Strength of Recommendation: Moderate Rationale A
repeat hip aspiration is suggested when there is a discrepancy between the
clinical probability of infection and the result of the initial aspiration
culture. One Level I study addressed this recommendation. This study examined
performing a repeat hip aspiration in 28 patients because their initial
aspiration result conflicted with the clinical suspicion for periprosthetic
infection.4 This report suggests that repeat aspiration of the hip be
considered when 1) the clinical probability of infection is low but the initial
aspiration culture result is positive, or 2) if the clinical probability of
infection is high and the initial aspiration culture result is negative (see
Supporting Evidence below, for details). The results of additional tests will
raise or lower the probability of periprosthetic infection. The study on which
this recommendation is based predated the more routine use and acceptance of
synovial fluid white blood cell count and differential. Thus, depending on the
results of the synovial fluid white blood cell count and differential, as well
as the ESR and CRP, the diagnosis or exclusion of periprosthetic infection may
be apparent and repeat aspiration may not be necessary. Supporting Evidence One
study4 with reliable data performed repeat aspiration in 28 patients because
the initial aspiration results contradicted clinical suspicions: 25 patients
had a positive initial culture but no clinical or radiographic indication of
infection while 3 patients had either a negative culture or a culture positive
for S. epidermidis only but infection was clinically suspected. 24 of the 25
with a positive initial culture had a negative repeat aspiration; the other had
a false positive repeat aspiration. The three patients with suspected infection
each had two repeat aspirations. The second aspiration
wasnegative in two and positive in one, while the third aspiration was positive
in all three. Repeat aspiration was attempted but no fluid was obtained
in five additional hips. The results are presented in Table 43.
101
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
EXCLUDED ARTICLES Table 41. Excluded Articles - Recommendation 5
Author Fehring, et al. 1996 Glithero, et al. 1993 Lachiewicz, et al. 1996
Somme, et al. 2003 Spangehl, et al. 1999 Taylor, et al. 1995 Title Aspiration
as a guide to sepsis in revision total hip arthroplasty White cell scans and
infected joint replacements. Failure to detect chronic
infection Aspiration of the hip joint before revision total hip arthroplasty.
Clinical and laboratory factors influencing attainment of a positive culture
Contribution of routine joint aspiration to the diagnosis of infection before
hip revision surgery Prospective analysis of preoperative and intraoperative
investigations for the diagnosis of infection at the sites of two hundred and
two revision total hip arthroplasties Fine needle aspiration in infected hip
replacements Reason for Exclusion 5 PMNs/HPF) Frozen Section (≥10 PMN/HPF
in 5 fields) Frozen Section (≥5 PMN/HPF in 5 fields) Reference Standard
Joint Positive Likelihood Ratio (95% CI) Negative Likelihood Ratio (95% CI)
Sensitivity (95% CI) Specificity TP FP FN (95% CI) TN
I
Fehring
97
Intraoperative Cultures
Mixed
1.74 (0.43, 7.03)
0.91 (0.69, 1.22)
0.18 (0.02, 0.52)
0.90 (0.81, 0.95)
2
9*
9
77
I
Ko
40
Intraoperative Cultures
Mixed (34 hip, 6 knee)
20.67 (2.85, 150.1)
0.34 (0.14, 0.87)
0.67 (0.3, 0.93)
0.97 (0.83, 1)
6
1
3
30
I
Lonner
175
Intraoperative Cultures
Mixed (142 hip, 33 knee)
65.7 (16.4, 264)
0.16 (0.06, 0.45)
0.84 (0.60, 0.97)
0.99 (0.95, 1)
16
2
3
154
I
Lonner
175
IntraoperativeCultures
Mixed (142 hip, 33 knee)
18.8 (8.87, 39.7)
0.17 (0.06, 0.47)
0.84 (0.60, 0.97)
0.96 (0.91, 0.98)
16
7
3
149
°Three of seven positive cultures considered possible contaminants due to
negative clinical and intraoperative findings *Study noted 6 cases with
negative cultures but high clinical suspicion of infection; 3 cases with
indeterminate results not included
170
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
RECOMMENDATION 13
We recommend that multiple cultures be obtained at the time of reoperation in
patients being assessed for periprosthetic joint infection. Strength of
Recommendation: Strong Rationale Four Level I studies compared the effects of
using one as opposed to more than one positive culture as a threshold for
diagnosing infection, with histology of peri-implant tissue as the gold
standard.2, 94, 104, 105 Two of these studies obtained at least three cultures
per patient and two of these studies obtained at least two cultures per
patient. Compared to histology, more than one positive culture result had a
higher positive likelihood ratio for diagnosing infection than a single
positive culture result. Assessing infection using multiple cultures increases
the chances of identifying infection. Cultures are easily performed during the
procedure and provide reliable results as indicated by the postive likelihood
ratio. Obtaining more than one culture decreases the likelihood of false negative
result and may assist the clinician in clarifying a result that may be deemed a
false positive based on the results of other tests. Supporting Evidence Four
studies presented reliable data on the effects of using one or more than one
positive culture as a threshold for infection.2, 94, 104, 105 In these studies, authors sampled from the most suspicious
areas or areas that appear to be most infected or inflamed. In these
studies, cultureswere compared to histologic findings or a combination of
histologic findings, visible purulence, and a sinus tract communicating with
the prosthesis. There was between-study heterogeneity in each category. Samples
were taken for aerobic and anaerobic culture in each study. One study with
moderately reliable data compared swab vs. tissue
and another study with less reliable data compared fluid vs. tissue.78 In each
study, the differences between the two techniques were not statistically
significant. SUMMARY OF EVIDENCE Table 71. Summary of
Evidence
Test Intraoperative Cultures (≥1 positive culture) range Intraoperative
Cultures (≥2 Number of Studies 4 4 Positive Likelihood Ratio (95% CI)
2.87 8.1 11.5 76.6 Negative Likelihood Ratio (95% CI) 0.09 0.29 0.25
0.47 Sensitivity (95% CI) Specificity (95% CI)
0.73 0.94 0.54 0.77
0.67 0.91 0.93 0.99
171
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
positive cultures) range *Range presented because meta-analysis indicated
heterogeneity
172
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
EXCLUDED ARTICLES Table 72. Excluded Articles - Recommendation 13
Author Baker, et al. 1994 Bare, et al. 2006 Barrack, et al. 2007 Barrack, et
al. 1993 Berend, et al. 2007 Cune, et al. 2008 Duff, et al. 1996 Dupont, 1986
Fehring, et al. 1996 Feldman, et al. 1995 Fitzgerald, et al. 1973 Ince, et al.
2004 Title Use of Sentinel blood culture system for analysis of specimens from
potentially infected prosthetic joints Preoperative evaluations in revision
total knee arthroplasty The fate of the unexpected positive intraoperative
cultures after revision total knee arthroplasty The value of aspiration of the
hip joint before revision total hip arthroplasty Unexpected positive
intraoperative cultures and gram stain in revision total hip arthroplasty for
presumed aseptic failure A Superficial Swab Culture is Useful forMicrobiologic
Diagnosis in Acute Prosthetic Joint Infections Aspiration of the knee joint
before revision arthroplasty Significance of operative cultures in total hip
arthroplasty Aspiration as a guide to sepsis in revision total hip arthroplasty
The role of intraoperative frozen sections in revision total joint arthroplasty
Bacterial colonization of wounds and sepsis in total hip arthroplasty Is
'aseptic' loosening of the prosthetic cup after total hip replacement due to
nonculturable bacterial pathogens in patients with low-grade infection? Reason
for Exclusion Does not address recommendation Not best available evidence Not
best available evidence Not best available evidence Not best available evidence
Not best available evidence Not best available evidence Insufficient Data Not
best available evidence Not best available evidence Does not address
recommendation 30, (>1/3 of CRP>10, preoperative samples aspiration with
at least positive) 1 positive culture, frozen section with >5PMN/HPF,
intraoperative culture (>1/3 of cultures positive)
Hip
115 (16.1, 829)
0.24 (0.1, 0.56)
0.76 (0.5, 0.93)
0.99 (0.96, 1)
13
1
4
150
186
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 75. Tissue vs. Swab vs. Fluid (Continued)
Level of Evidence
Author
N
Test
Reference Standard
Negative Positive Likelihood Likelihood Sensitivity Specificity TP FP FN Joint
(95% CI) (95% CI) Ratio Ratio (95% CI) (95% CI)
TN
III
At least 1 of: 1)open wound of sinus in communication with the joint 2)systemic
infection with pain in the joint and purulent fluid within the joint
Intraoperative 3)positive result on at Tissue least 3 Cultures Spangehl 180
investigations(ESR>30, (>1/3 of CRP>10, preoperative sample aspiration
with at least positive) 1 positive culture, frozen section with >5PMN/HPF,
intraoperative culture (>1/3 of cultures positive)
Hip
30.6 (12.8,73.1)
0.06 (0.01, 0.39)
0.94 (0.73, 1)
0.97 (0.93, 0.99)
17
5
1
157
187
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
Table 75. Tissue vs. Swab vs. Fluid (Continued)
Level of Evidence
Author
N
Test
Reference Standard
Negative Positive Likelihood Likelihood Sensitivity Specificity TP FP FN Joint
(95% CI) (95% CI) Ratio Ratio (95% CI) (95% CI)
TN
IV
Parvizi
70
Intraoperative Tissue Culture
IV
Parvizi
70
Intraoperative Fluid Culture
At least 3 of: 1)CRP >1mg/dL 2)ESR >30mm/hr 3)positive joint aspiration
culture 4)purulent intraoperative tissue appearance 5)positive intraoperative
culture At least 3 of: 1)CRP >1mg/dL 2)ESR >30mm/hr 3)positive joint
aspiration culture 4)purulent intraoperative tissue appearance 5)positive
intraoperative culture
Knee
53.6 (3.41, 841)
0.17 (0.08, 0.33)
0.85 (0.69, 0.94)
1 (0.89, 1)
33
0
6
31
Knee
27.8 (4.03, 191)
0.11 (0.04, 0.27)
0.9 (0.76, 0.97)
0.97 (0.83, 1)
35
1
4
30
188
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
RECOMMENDATION 14
We recommend against initiating antibiotic treatment in patients with suspected
periprosthetic joint infection until after cultures from the joint have been
obtained. Strength of Recommendation: Strong Rationale One Level I study
addressed whether administration of antibiotic therapy prior to obtaining
cultures from the joint affected the sensitivity of the cultures in diagnosing
periprosthetic infections.105 The study found a false negative rate of 55% in
patients receiving antibiotics within the previous 14 days compared to 23% in
patients not receiving antibiotics during the same time period. The difference
was statistically significant. Hence, there is a concern that antibiotics can
interfere with isolation of the infecting organism leading to confusion
regarding the diagnosisor inability to use organism specific antibiotics
subsequently when infection is confirmed. Thus, administration of oral or
intravenous antibiotics to patients with suspected diagnosis of periprosthetic
joint infection is discouraged, until aspiration of the joint is performed or
samples for culture are obtained. The finding of only one Level I study
supporting this recommendation would be evaluated as moderate strength.
However, because of the severity of the potential harm to the patient in
getting a false negative culture result, the strength of the recommendation was
elevated to strong. Supporting Evidence One included study with reliable data
addressed whether antibiotic therapy decreases the sensitivity of
intraoperative cultures in diagnosing periprosthetic infection.105 The group of
patients who had received antimicrobial therapy within 14 days of surgery had a
statistically significantly higher rate of false-negative culture results than
those who had not. In this study, a positive result was defined as growth from
at least two specimens.
189
AAOS Clinical Practice Guidelines Unit v0.1 03.25.10
EXCLUDED ARTICLES Table 76. Excluded Articles -
Recommendation14
Author Barrack, et al. 1997 Datz, et al. 1986 Spangehl, et al. 1999 Spangehl,
et al. 1999 Trampuz, et al. 2006 Title The Coventry Award. The value of
preoperative aspiration before total knee revision Effect of antibiotic therapy
on the sensitivity of indium-111-labeled leukocyte scans Prospective analysis
of preoperative and intraoperative investigations for the diagnosis of
infection at the sites of two hundred and two revision total hip arthroplasties
The role of intraoperative gram stain in the diagnosis of infection during
revision total hip arthroplasty Sonication of explanted prosthetic components
in bags for diagnosis of prosthetic joint infection is associated with risk of
contamination Reason for Exclusion
