Artritis septica estudio

1. n Feature Article
S
eptic arthritis is a devastating con-
dition that requires a timely diag-
nosis to minimize articular damage
and the associated high rate of morbid-
ity and mortality. The pathogenesis typi-
cally involves 1 of 3 mechanisms: hema-
togenous seeding, direct inoculation, or
spread through local infection. A delay in
diagnosis increases the risk of long-term
sequelae including posttraumatic arthritis,
osteomyelitis, osteonecrosis, sepsis, and
death.1-5
The mortality rate following sep-
tic arthritis is 7% to 15%, thus an urgent
diagnosis is paramount in the evaluation
of an acutely irritated joint.2,4,6-9
Arthrocentesis is a useful diagnostic
modality in differentiating between an in-
flammatory arthropathy, benign process,
and infectious process.2,7,8,10-25
Although a
positive fluid culture is important in mak-
ing an accurate diagnosis, there are high
false-negative rates, relatively poor accu-
JULY/AUGUST 2016 | Volume 39 • Number 4
Predictors of Septic Arthritis in the Adult
Population
Robert Borzio, MD; Neil Mulchandani, MD; Robert Pivec, MD; Bhaveen H. Kapadia, MD;
Dante Leven, DO; Steven F. Harwin, MD; William P. Urban, MD
The authors are from the Department of
Orthopaedic Surgery (RB, NM, RP, BHK, DL,
WPU), SUNY Downstate Medical Center, Brook-
lyn, and the Department of Orthopaedic Surgery
(SFH), Beth Israel Medical Center, New York,
New York.
Drs Borzio, Mulchandani, Leven, and Urban
have no relevant financial relationships to dis-
close. Dr Pivec receives personal fees from DJ
Orthopaedics. Dr Kapadia is a paid consultant
for and is on the speaker’s bureau of Sage Prod-
ucts, Inc. Dr Harwin is a paid consultant for and
is on the speaker’s bureau of Stryker and Con-
vatec and holds stock in Stryker.
Correspondence should be addressed to:
Bhaveen H. Kapadia, MD, Department of Ortho-
paedic Surgery, SUNY Downstate Medical Center,
450 Clarkson Ave, Box 30, Brooklyn, NY 11203
(bhaveen.k@gmail.com).
Received: March 4, 2015; Accepted: Decem-
ber 28, 2015.
doi: 10.3928/01477447-20160606-05
Septic arthritis is a devastating condition; well-established criteria for diagno-
sis exist in the pediatric population, but not for adults. This study evaluated
patient factors and laboratory parameters that may be associated with the di-
agnosis of septic arthritis in adults. A total of 458 knee aspirates for suspected
septic arthritis were evaluated with serum and synovial leukocyte counts and
differentials as well as Kocher criteria for pediatric septic arthritis. Twenty-
two patients (4.8%) had septic arthritis confirmed by a positive synovial fluid
culture. Erythrocyte sedimentation rate (ESR) and serum white blood cell
(WBC) counts were not statistically different between the 2 groups, with 64%
of septic arthritis patients having a normal serum WBC count and 77% being
afebrile. Mean synovial fluid WBC count was 26,758 cells/µL and 70,581
cells/µL in the nonseptic and septic groups, respectively. The likelihood ratio
for a synovial fluid WBC count greater than 65,000 cells/µL was 2.8 (95%
confidence interval, 1.2-6.7). Evaluation receiver operating characteristic
curves using synovial WBC counts resulted in a significant area under the
curve of 0.66 (P=.02). To achieve 90% specificity, a WBC cutoff of 64,000
cells/µL was required with a corresponding sensitivity of 40%. There was
no significant difference in the synovial cell differential of 80% vs 90% in
diagnosing infection. Synovial fluid WBC count greater than 64,000 cells/µL
yielded the optimal combination of sensitivity and specificity. Polymorpho-
nuclear leukocytes, ESR, serum WBC count, fever, and weight-bearing status
were not significant predictors of septic arthritis. This study demonstrates the
limited utility of Kocher criteria in the adult population and the importance of
synovial leukocyte counts. [Orthopedics. 2016; 39(4):e657-e663.]
abstract
e657

2. Copyright © SLACK Incorporated
n Feature Article
racy, and modest sensitivity and specificity
in using the current synovial fluid thresh-
olds.26
Several laboratory parameters such
as white blood cell (WBC) count, eryth-
rocyte sedimentation rate (ESR), and C-
reactive protein are commonly used, but
the current literature is inconsistent with
regard to acceptable thresholds for opti-
mal sensitivity and specificity (Table 1).
In addition, it is unclear whether combin-
ing these factors to rule out infection (eg,
Kocher criteria) can be applied to the adult
population. Synovial fluid analysis is also
considerably variable with regard to WBC
count thresholds and polymorphonuclear
(PMN) percentiles in diagnosing septic
arthritis.6,27,28
The objective of the current study was
to analyze patient factors and laboratory
parameters that may be associated with
a diagnosis of septic arthritis in the adult
population. Kocher was able to establish
a set of factors to assess the pediatric
population, and although not intended to
be applied to adults, those factors, along
with others, have been used to assist in
diagnosing septic arthritis in adults, of-
ten with little literature to justify their
use. The current authors hypothesized
that specific patient baseline factors and
laboratory parameters may be predictors
of septic arthritis and improve diagnostic
accuracy.
Materials and Methods
Patient records of all adults from 2
institutions between June 2005 and June
2011 were evaluated to identify patients
who were diagnosed with suspected sep-
tic arthritis. A total of 458 patients were
identified using the inpatient database on
the basis of Current Procedural Terminol-
ogy (CPT) code 20610 and International
Classification of Diseases, Ninth Revision
(ICD-9) codes 711.46 and 711.41 (arthro-
centesis and septic arthritis of the knee
and shoulder, respectively). A total of 231
patients were identified from hospital A,
a 620-bed level 1 trauma center, and 236
patients were identified from hospital B, a
state-funded, 320-bed institution. Serum
and synovial leukocyte counts and WBC
differentials (neutrophils, lymphocytes,
monocytes, eosinophils, and basophils)
were compared between a group of pa-
tients who had culture-positive septic ar-
thritis (n=22) and those who had negative
cultures (n=436). Patients with incomplete
clinical or imaging data and atypical pa-
tients, as well as those with periprosthetic
infections, postoperative septic arthritis,
and associated proximal femoral osteo-
myelitis were excluded (n=188). All charts
were reviewed to confirm that no patients
were given antibiotics prior to aspiration.
In addition, all synovial fluid samples were
tested for aerobic, anaerobic, mycobacte-
rium, and fungal pathogens.
Data collected included patient age,
sex, date of presentation, disease history
(duration history, previous health care
visit, fever, and weight-bearing status),
comorbidities (diabetes, hypertension,
HIV infection, gout, alcohol abuse, hepa-
titis, malignancy, kidney disease, rheuma-
toid arthritis, and osteoarthritis), clinical
findings (body temperature), laboratory
parameters (ESR, hematocrit, platelet
count, serum WBC count, and neutro-
phil, lymphocyte, monocyte, band neu-
Table 1
Literature Review
Synovial Fluid WBC Count/
Study (Year) LOE Sensitivity Specificity + LR - LR
25-50 × 109
/L
Carpenter et al29
(2011) 2 73% 77% 3.2 0.35
Margaretten et al6
(2007) 2 77% 73% 2.9 0.32
Soderquist et al32
(1998) 3 73% 58% 1.7 0.47
Shmerling et al26
(1990) 4 63% 83% 3.6 0.45
Krey and Bailen33
(1979) 3 88% 71% 3.1 0.17
50-100 × 109
/L
Carpenter et al29
(2011) 2 90% 56% 4.7 0.52
Li et al20
(2007) 3 50% 88% 4 0.57
Margaretten et al6
(2007) 2 62% 92% 7.7 0.42
Li et al21
(2004) 3 62% NR NR NR
Soderquist et al32
(1998) 3 58% 74% 2.2 0.57
Kortekangas et al30
(1992) 3 31% 75% 1.3 0.92
Shmerling et al26
(1990) 4 50% 97% 15 0.52
Krey and Bailen33
(1979) 3 70% 92% 7.7 0.42
>100 × 109
/L
Carpenter et al29
(2011) 2 19% 99% 13.2 0.83
Margaretten et al6
(2007) 2 29% 99% 28 0.71
Li et al21
(2004) 3 31% NR NR NR
Soderquist et al32
(1998) 3 30% 93% 4.7 0.75
Kortekangas et al30
(1992) 3 6% 100% ∞ 0.94
Shmerling et al26
(1990) 4 13% 100% 31 0.84
Krey and Bailen33
(1979) 3 40% 99% 42 0.61
Abbreviations: LOE, level of evidence; LR, likelihood ratio; NR, not reported; WBC, white
blood cell.
e658

3. JULY/AUGUST 2016 | Volume 39 • Number 4
n Feature Article
trophil, atypical lymphocyte, eosinophil,
and basophil percentage), synovial fluid
findings (cultured organism, cell count,
Gram stain, and segmented neutrophil,
lymphocyte, monocyte, and red blood cell
percentage) for those patients who were
taken to the operating room.
The sensitivity and specificity of the
values for synovial leukocytes were de-
termined by generating receiver operat-
ing characteristic (ROC) curves. Posi-
tive and negative likelihood ratios (LR)
also were calculated for history of fever,
documented fever, weight-bearing status,
synovial fluid leukocyte cutoff levels, and
segmental leukocyte differentials. The ef-
fect of comorbidities, using the Charlson
Comorbidity Index (CCI) on the level of
synovial leukocytes in patients with septic
arthritis, was analyzed using regression
analysis. Application of Kocher criteria
(WBC count >12,000 cells/µL, ESR >40
mm/h, temperature >38.5°C, and history
of non-weight bearing [on the affected
extremity]) was used for comparing the 2
groups of patients.
Positive and negative predictive values
were calculated for each patient. The pro-
portion of true positives was calculated
for each patient along with positive and
negative LR. A P value of less than .05
was used as the cutoff to determine statis-
tical significance. Levene’s test for equal-
ity of variances and t test for equality of
means were used for synovial fluid WBC
count and ESR. Statistical analysis was
performed using SPSS version 22.0 soft-
ware (IBM, Armonk, New York).
Results
Within the combined cohort, a total of
22 patients had septic arthritis confirmed
by a positive synovial fluid culture. Of
these, 15 patients (68%) had a positive
Gram stain, with the most common organ-
ism being Staphylococcus aureus. Bac-
teria identified from the septic arthritis
group are listed in Table 2.
Patients in the nonseptic arthritis group
were older than those in the septic arthritis
group (61.6 vs 51.2 years, P=.012). A his-
tory of fever, temperature, number of co-
morbidities, or weight-bearing status were
not significantly different between the 2
groups (P>.05). The ESR, serum WBC
count, and serum neutrophil percentage
also were not significantly different be-
tween the 2 groups (P>.05) (Table 3). The
platelet count was higher and hematocrit
was lower in the septic arthritis group than
in the nonseptic arthritis group (P=.017
and P=.018, respectively).
The mean synovial fluid WBC count
was 26,758 cells/µL for the nonseptic ar-
thritis group and 70,581 cells/µL for the
septic arthritis group (P=.001) (Figure 1).
The percentage of PMN cells in the syno-
vial fluid was slightly higher in the septic
arthritis group; however, this difference
was not statistically significant (84.4% in
the septic arthritis group vs 79.3% in the
nonseptic arthritis group, P=.349). Syno-
vial fluid lymphocyte, monocyte, and red
blood cell counts were not significantly
different in the 2 groups (P>.05).
In the group of patients with culture-
positive septic arthritis and a synovial flu-
id WBC count greater than 64,000 cells/
µL, the PMN cell percentage was slightly
higher than in patients with a synovial flu-
id WBC count less than 64,000 cells/µL;
this difference was not statistically signifi-
cant (93.8% vs 71%, respectively; P>.05).
Gram stain also was more likely to be
positive (10 of 12 vs 3 of 6, respectively;
P>.05). The serum WBC count tended to
be higher in patients with synovial fluid
WBC count greater than 64,000 cells/µL
compared with patients with synovial flu-
id WBC count less than 64,000 cells/µL;
however, this difference was not statisti-
cally significant (11.93 vs 8.631 cells/µL,
Table 2
Isolated Organisms
Organism
No. of
Cases
Staphylococcus aureus 5
MRSA 3
MSSA 2
Acinetobacter lwoffii 2
Klebsiella oxytoca 2
Staphylococcus epidermidis 1
Flavimonas oryzihabitans 1
Staphylococcus capitis
ureolyticus
1
Staphylococcus hominis 1
Streptococcus viridans 1
Escherichia coli 1
S aureus/E coli 1
Bacteria (unspecified) 1
Abbreviations: MRSA, methicillin-
resistant Staphylococcus aureus; MSSA,
methicillin-susceptible Staphylococcus
aureus.
Table 3
Laboratory Values
Mean (Range)
Patient Factor Septic Arthritis Nonseptic Arthritis P
ESR, mm/h 94 (18-140) 82 (1-140) .2582
WBC count (serum), cells/µL 10.1 (0.35-22.7) 11.1 (1.9-36.3) .6149
WBC count (synovial fluid),
cells/µL
70,581 (350-220,000) 26,758 (1-307,000) .7288
Neutrophil 70.1% (4%-90%) 74.9% (5%-91%) .2984
Lymphocyte 17.7% (2%-72%) 15.9% (0.8%-52%) .4331
Temperature, °F 99.1 (97.5-102.8) 99.2 (96.3-104) .9686
Abbreviations: ESR, erythrocyte sedimentation rate; WBC, white blood cell.
e659

4. Copyright © SLACK Incorporated
n Feature Article
respectively; P>.05). Neutrophil percent-
age (75.7% vs 59.7%; P>.05) and PMN
cell percentage in synovial fluid (71% vs
93.8%; P>.05) also were not statistically
significant for patients with synovial fluid
WBC count greater than 64,000 cells/µL
and patients with synovial fluid WBC
count less than 64,000 cells/µL, respec-
tively.
Receiver operating characteristic curve
showed the area under the curve (AUC)
was 0.66 (P=.02) (Figure 2). With WBC
counts of 64,000 cells/µL in synovial fluid,
a specificity of 90% and sensitivity of 40%
was obtained. The LR of septic arthritis
for a synovial fluid of 40,000 cells/µL was
1.88 (95% confidence interval [CI], 0.90-
3.93), the LR for 50,000 cells/µL was 2.02
(95% CI, 0.94-5.11), and the LR for 65,000
cells/µL was 2.8 (95% CI, 1.2-6.7).A weak
correlation (R2
=0.11) was noted between
the synovial cell count and the CCI (Fig-
ure 3). With application of the Kocher cri-
teria, based on 4 positive factors, the posi-
tive predictive value (PPV) was 12.5%, the
negative predictive value (NPV) was 91%,
and the LR was 5.02 (95% CI, 1.11-22.79).
Addition of a fifth factor (synovial leuko-
cyte count) resulted in a PPV of 0% and
NPV of 90.8% (Figure 4).
Addition of a synovial fluid cutoff to
the classic Kocher criteria demonstrated
a sensitivity of 0%, specificity of 98.6%,
NPV of 90.8%, and PPV of 0%. Frequen-
cy distribution of the number of criteria in
both true positive (septic) and true nega-
tive (nonseptic) patients demonstrated an
increasing number of patients with 2 or
more Kocher criteria in the setting of true
septic arthritis (Figure 4).
Discussion
This investigation assessed the utility
of serum values, patient comorbidities, and
synovial fluid values in predicting sep-
tic arthritis in the adult population. Acute
nongonococcal septic arthritis is often a
diagnostic challenge given the heterogene-
ity of clinical presentations, confounding
laboratory parameters, and the inaccuracy
of microbiologic tests.6
In this study, 22 of
458 patients (4.8%) who underwent arthro-
centesis had culture-positive septic arthri-
tis. In addition, other than older age (61.6
vs 51.2 years; P=.012), patient baseline
characteristics, history of fever, and serum
laboratory parameters (ESR, serum WBC
count, and differentials) were not reliable
predictors of septic arthritis (P>.05). Sy-
novial fluid WBC count was a significant
predictor of septic arthritis.
The mean synovial WBC count was
26,758 cells/µL in the nonseptic group
and 70,581 cells/µL in the septic arthritis
group; the synovial WBC count threshold
that showed the highest combined sensi-
tivity and specificity was 64,000 cells/
µL. Evaluation of the ROC curve using
synovial WBC count resulted in a sig-
nificant AUC of 0.66 (P=.02). To achieve
90% specificity, a WBC count of 64,000
cells/µL or greater was needed with a cor-
responding sensitivity of 40%. The LR of
detecting septic arthritis for a synovial flu-
Figure 2: Receiver operating characteristic curve of sensitivity and specificity of synovial white blood cell
count for diagnosis of native joint septic arthritis with an area under the curve of 0.66 (P=.02).
Figure 1: Graph showing synovial white blood cell (WBC) count relative distribution (% of total group
within specified synovial WBC count range) in the septic and nonseptic arthritis groups.
e660

5. JULY/AUGUST 2016 | Volume 39 • Number 4
n Feature Article
id WBC count greater than 64,000 cells/
µL was 2.8 (95% CI, 1.2-6.7). Although
this may represent a low sensitivity and
increase the chances of a false-negative
results, a value above this cutoff value
should be seen as indicative of a true septic
arthritis, whereas values below this cutoff
may represent a more mixed picture (Fig-
ure 1). Thus, the authors believe that if an
aspirate is obtained with a PMN cell count
above the cutoff value, no further studies
are necessary, and irrigation and debride-
ment should not be delayed. In contrast,
values below the cutoff value should place
increased impetus on the patient’s clinical
picture and the possible need for further
imaging or serologic studies to rule out
other causes of acute joint pain and thus
avoid unnecessary surgical morbidity in a
patient who may not have septic arthritis.
Several studies have shown sensitivity
of the synovial WBC count greater than
50,000 cells/µL ranges from 49% to 63%,
with a specificity of 88% to 92%.4,29,30
A
single-institution retrospective study by
Khan et al4
of 56 patients found that using
a lower synovial WBC count threshold
of 50,000 cells/µL yielded a sensitivity
of 58.9%. This study similarly observed
a wide range of synovial WBC counts
(range, 1620-230,000 cells/µL) in patients
with culture-positive septic arthritis. It
is possible that some of this variability,
particularly low synovial cell counts, can
be explained by aspirates early in the in-
fectious process, immunocompromised
patients, or partially treated infections.22
However, a high WBC count should not
be exclusively seen as being caused by
bacterial infection because inflammatory
processes such as rheumatoid arthritis,
gout, and pseudogout also can have mark-
edly elevated values.14
This variability
highlights the importance of further stud-
ies and particularly to improve microbio-
logic testing accuracy beyond the current
gold standard synovial fluid culture.
In the current study, the synovial cell
percentage of PMN cells was not statisti-
cally significant in comparing the septic
arthritis and nonseptic arthritis groups
(P>.05). In addition, the LR of septic
arthritis was 0.76 with 80% PMN cells
and 0.99 with 90% PMN cells. This con-
trasts with other studies that reported a
cell count of at least 90% suggests sep-
tic arthritis with an LR of 3.4 (95% CI,
2.8-4.2), whereas a PMN cell count less
than 90% lowers the likelihood (LR, 0.34;
95% CI, 0.25-0.47).6
It is unclear why the
percentage of PMN cells in the synovial
fluid analysis from the current study was
not significant. This may be due to the fact
that noninfectious inflammatory arthropa-
thies can cause an elevated PMN percent-
age in synovial fluid or that patients who
are immunosuppressed with septic arthri-
tis may not develop a sufficient immune
response.
This investigation showed that labora-
tory parameters (ESR, C-reactive protein,
and serum WBC count) and body temper-
Figure 3: Graph showing correlation of synovial cell count with Charlson Comorbidity Index (septic ar-
thritis cases).
Figure 4: Graph showing frequency distribution of the number of positive criteria in septic and nonseptic
arthritis patients; criteria were fever greater than 100.4°F, inability to bear weight, erythrocyte sedimenta-
tion rate greater than 40 mm/h, serum leukocyte count greater than 12 cells/mm3
, and absolute synovial
leukocyte count greater than 64,000 cells/µL.
e661

6. Copyright © SLACK Incorporated
n Feature Article
ature were not statistically different be-
tween the 2 cohorts, which may indicate
that they have limited clinical utility in
the diagnostic work-up of septic arthritis.
The low PPV of the Kocher criteria or the
addition of synovial leukocyte count as a
fifth criterion results in a high false-posi-
tive proportion of tests. Although there is
a high NPV (>90%), using this criteria in
adults is not recommended as the sequelae
of a missed septic arthritis (false negative)
are catastrophic. However, these values
may be useful during long-term treatment
monitoring.2,7,10
The hematocrit count was
significantly lower and the platelet count
was significantly higher in the septic ar-
thritis cohort (P<.05), which is consistent
with infection being a common cause of
thrombocytosis.31
The low hematocrit
may reflect a shift in the hematopoietic
system during an infectious disease pro-
cess, although further studies are needed.
The CCI correlated weakly with cul-
ture-positive septic arthritis (R2
=0.11)
(Figure 3). The overall trend was a slight
decrease in synovial fluid WBC count
with an increase in the CCI, although
outliers were seen with some high syno-
vial fluid WBC counts in patients with
high CCI scores. Septic arthritis patients
with multiple comorbidities may have
a blunted immune response (Figure 3),
thus serum or synovial leukocytosis alone
should not be relied on, particularly in
immunocompromised patients. A recent
meta-analysis compared the relative risk
of septic arthritis in the setting of multiple
comorbidities,6
demonstrating increasing
risk with recent joint surgery excluding
arthroplasty (LR, 6.9; 95% CI, 3.8-12.0),
age older than 80 years (LR, 3.5; 95% CI,
1.8-7.0), rheumatoid arthritis (LR, 2.5;
95% CI, 2.0-3.1), skin infection (LR, 2.8;
95% CI, 1.7-4.5), diabetes mellitus (LR,
2.7; 95% CI, 1.0-6.9), and HIV (LR, 1.7;
95% CI, 1.0-2.8).8
There were several limitations to the
current study, including the retrospective
design, which may introduce confounding
factors. In addition, this study was con-
ducted at a single institution, which could
result in spectrum bias skewing toward
higher-severity illnesses and falsely el-
evating sensitivity. However, the results of
this study may still be applicable to simi-
lar socioeconomic patient populations and
similar institutions to determine which
specific factors are most predictive of sep-
tic arthritis. Negative cultures were attrib-
uted to other causes based on clinical pre-
sentation and other synovial or serologic
markers (eg, rheumatoid, gout, reactive
arthritis, Lyme disease, and exacerbation
of osteoarthritis). Relatively high false-
negative rates also are known to occur in
clinical practice as some bacteria such as
Kingella kingae and certain mycobacte-
rium species may require an extended in-
cubation time. Antibiotics administration
prior to arthrocentesis was not completely
documented in all cases, which precluded
a complete analysis of possible effects of
culture-negative results.
Although studies support clinical pre-
diction algorithms for early diagnosis of
septic arthritis of the hip in children, there
is no current model in place for the adult
population. An algorithm described by
Kocher et al9
combined 4 independent
predictors to assist in differentiating be-
tween septic arthritis and transient syno-
vitis in the hips of children.12
Adult septic
arthritis, however, has many confounding
factors, and this study demonstrates that
the Kocher criteria are of limited utility in
adult patients. Therefore, arthrocentesis
should be used for diagnosis with par-
ticular attention to the synovial leukocyte
count in combination with the patient’s
clinical presentation and inflammatory
markers.
Conclusion
This study emphasizes the importance
of synovial fluid analysis and clinical
examination in the work-up of septic ar-
thritis. There is little utility in the Kocher
criteria for adult patients, and serum labo-
ratory values frequently are elevated non-
specifically. Older age was more common
in the septic arthritis group, and synovial
fluid WBC count trended down slightly
with an increase in patient CCI score.
Synovial PMN percentage was not sig-
nificantly different between the septic and
nonseptic arthritis groups. Synovial fluid
WBC counts greater than 64,000 cells/µL
yielded the optimal combination of sensi-
tivity and specificity, and should be seen
as a value above which a patient almost
certainly has a true infection. Lower val-
ues may represent a more mixed picture
and warrant further examination to avoid
the morbidity of unnecessary irrigation
and debridement. Further studies analyz-
ing larger populations and modalities such
as polymerase chain reaction and other
synovial fluid parameters such as lactate,
proteins, and glucose may enhance the
ability to accurately diagnosis septic ar-
thritis in a timely manner.
References
1. Newman JH. Review of septic arthritis
throughout the antibiotic era. Ann Rheum
Dis. 1976; 35(3):198-205.
2. Mathews CJ, Weston VC, Jones A, Field M,
Coakley G. Bacterial septic arthritis in adults.
Lancet. 2010; 375(9717):846-855.
3. Sharff KA, Richards EP, Townes JM. Clinical
management of septic arthritis. Curr Rheu-
matol Rep. 2013; 15(6):332.
4. Khan FY, Abu-Khattab M, Baagar K, et al.
Characteristics of patients with definite septic
arthritis at Hamad General Hospital, Qatar: a
hospital-based study from 2006 to 2011. Clin
Rheumatol. 2013; 32(7):969-973.
5. Haustedt N, Schmidt H, Loock J, Wollen-
haupt J. Bacterial septic arthritis as rheuma-
tological emergency [in German]. Z Rheuma-
tol. 2012; 71(4):288-296.
6. Margaretten ME, Kohlwes J, Moore D, Bent
S. Does this adult patient have septic arthri-
tis? JAMA. 2007; 297(13):1478-1488.
7. Uckay I, Tovmirzaeva L, Garbino J, et al.
Short parenteral antibiotic treatment for adult
septic arthritis after successful drainage. Int J
Infect Dis. 2013; 17(3):e199-e205.
8. Goldenberg DL. Septic arthritis. Lancet.
1998; 351(9097):197-202.
9. Kocher MS, Mandiga R, Zurakowski D,
Barnewolt C, Kasser JR. Validation of a clini-
cal prediction rule for the differentiation be-
tween septic arthritis and transient synovitis
of the hip in children. J Bone Joint Surg Am.
2004; 86(8):1629-1635.
e662

7. JULY/AUGUST 2016 | Volume 39 • Number 4
n Feature Article
10. Al-Nammari SS, Gulati V, Patel R, Bejjanki
N, Wright M. Septic arthritis in haemodialy-
sis patients: a seven-year multi-centre review.
J Orthop Surg (Hong Kong). 2008; 16(1):54-
57.
11. Geirsson AJ, Statkevicius S, Vikingsson
A. Septic arthritis in Iceland 1990-2002:
increasing incidence due to iatrogenic in-
fections. Ann Rheum Dis. 2008; 67(5):638-
643.
12. Gupta MN, Sturrock RD, Field M. A pro-
spective 2-year study of 75 patients with
adult-onset septic arthritis. Rheumatology
(Oxford). 2001; 40(1):24-30.
13. Hariharan P, Kabrhel C. Sensitivity of eryth-
rocyte sedimentation rate and C-reactive
protein for the exclusion of septic arthritis
in emergency department patients. J Emerg
Med. 2011; 40(4):428-431.
14. Shah K, Spear J, Nathanson LA, McCauley
J, Edlow JA. Does the presence of crystal ar-
thritis rule out septic arthritis? J Emerg Med.
2007; 32(1):23-26.
15. Ungprasert P, Kaewpoowat Q, Ratapano S,
Srivali N, Bischof EF Jr. Presence of crystals
is not an evidence of absence of infection. Am
J Emerg Med. 2013; 31(2):455.
16. Abdullah S, Young-Min SA, Hudson SJ,
Kelly CA, Heycock CR, Hamilton JD. Gross
synovial fluid analysis in the differential di-
agnosis of joint effusion. J Clin Pathol. 2007;
60(10):1144-1147.
17. de Jonge R, Brouwer R, Smit M, et al. Au-
tomated counting of white blood cells in sy-
novial fluid. Rheumatology (Oxford). 2004;
43(2):170-173.
18. Gallagher EJ. Evidence-based emergency
medicine/editorial: the problem with sensi-
tivity and specificity. Ann Emerg Med. 2003;
42(2):298-303.
19. Kocher MS, Zurakowski D, Kasser JR. Differ-
entiating between septic arthritis and transient
synovitis of the hip in children: an evidence-
based clinical prediction algorithm. J Bone
Joint Surg Am. 1999; 81(12):1662-1670.
20. Li SF, Cassidy C, Chang C, Gharib S, Tor-
res J. Diagnostic utility of laboratory tests
in septic arthritis. Emerg Med J. 2007;
24(2):75-77.
21. Li SF, Henderson J, Dickman E, Darzynkie-
wicz R. Laboratory tests in adults with mono-
articular arthritis: can they rule out a septic
joint? Acad Emerg Med. 2004; 11(3):276-
280.
22. McGillicuddy DC, Shah KH, Friedberg RP,
Nathanson LA, Edlow JA. How sensitive is
the synovial fluid white blood cell count in
diagnosing septic arthritis? Am J Emerg Med.
2007; 25(7):749-752.
23. Trampuz A, Hanssen AD, Osmon DR, Man-
drekar J, Steckelberg JM, Patel R. Synovial
fluid leukocyte count and differential for the
diagnosis of prosthetic knee infection. Am J
Med. 2004; 117(8):556-562.
24. Yang S, Ramachandran P, Hardick A, et al.
Rapid PCR-based diagnosis of septic arthritis
by early Gram-type classification and patho-
gen identification. J Clin Microbiol. 2008;
46(4):1386-1390.
25. Yu KH, Luo SF, Liou LB, et al. Concomi-
tant septic and gouty arthritis: an analysis
of 30 cases. Rheumatology (Oxford). 2003;
42(9):1062-1066.
26.
Shmerling RH, Delbanco TL, Tosteson
AN, Trentham DE. Synovial fluid tests:
what should be ordered? JAMA. 1990;
264(8):1009-1014.
27. Dubost JJ, Soubrier M, Sauvezie B. Pyogenic
arthritis in adults. Joint Bone Spine. 2000;
67(1):11-21.
28. Ernst AA, Weiss SJ, Tracy LA, Weiss NR.
Usefulness of CRP and ESR in predicting
septic joints. South Med J. 2010; 103(6):522-
526.
29. Carpenter CR, Schuur JD, Everett WW, Pines
JM. Evidence-based diagnostics: adult septic
arthritis. Acad Emerg Med. 2011; 18(8):781-
796.
30.
Kortekangas P, Aro HT, Tuominen J,
Toivanen A. Synovial fluid leukocytosis in
bacterial arthritis vs. reactive arthritis and
rheumatoid arthritis in the adult knee. Scand
J Rheumatol. 1992; 21(6):283-288.
31. Santhosh-Kumar CR, Yohannan MD, Higgy
KE, al-Mashhadani SA. Thrombocytosis in
adults: analysis of 777 patients. J Intern Med.
1991; 229(6):493-495.
32. Soderquist B, Jones I, Fredlund H, Vikerfors
T. Bacterial or crystal-associated arthritis?
Discriminating ability of serum inflam-
matory markers. Scand J Infect Dis. 1998;
30(6):591-596.
33. Krey PR, Bailen DA. Synovial fluid leukocy-
tosis: a study of extremes. Am J Med. 1979;
67(3):436-442.
e663