E D I T O R I A L C O M M E N TA R Y
Bone of Contention: Diagnosing Diabetic Foot Osteomyelitis
Benjamin A. Lipsky
Veterans Affairs Puget Sound Health Care System and University of Washington, Seattle
(See the article by Dinh et al. on pages 519–27)
Development of osteomyelitis of the foot
is a potentially catastrophic event for a
person with diabetes. The high success
rates achieved with antimicrobial therapy
for most infectious diseases have not yet
been achieved for bone infections because
of their unique physiological and anatomical characteristics [1]. When a foot ulcer
becomes infected and the infection
spreads to bone, the risk of limb amputation, with its substantial associated morbidity and mortality, is dramatically increased [2]. Furthermore, diabetic foot
osteomyelitis often requires surgical therapy and/or prolonged antibiotic therapy.
Because the key to successful management
is early diagnosis, making an accurate diagnosis of this entity is crucial. Unfortunately, it is also difficult.
Two main issues complicate making a
correct diagnosis of osteomyelitis in the
diabetic foot [3]. First, as with other types
of bone infection, it usually takes a couple
of weeks before there is sufficient loss of
bone to be apparent on plain radiographs.
Second, patients with longstanding diabetes often have peripheral neuropathy,
which may both obscure clinical sympReceived 28 April 2008; accepted 1 May 2008;
electronically published 7 July 2008.
Reprints or correspondence: Dr. Benjamin A. Lipsky, VA
Puget Sound HCS, S-111-PCC, 1660 S. Columbian Way,
Seattle, WA 98108 ([email protected]).
Clinical Infectious Diseases 2008; 47:528–30
This article is in the public domain, and no copyright is
claimed.
1058-4838/2008/4704-0014
DOI: 10.1086/590012
toms of infection [4] and predispose to
neuro-osteoarthropathy. This noninfectious entity, often called Charcot foot, can
be difficult to differentiate from bone infection [5]. Examination of a bone sample,
with microbiological or histopathologic
evaluation, is generally accepted as the criterion standard for diagnosis of osteomyelitis [6, 7]. Unfortunately, this safe and
relatively simple procedure is not widely
used and can yield results that are either
false positive (caused by specimen contamination during the procedure) or false
negative (caused by prior antibiotic therapy or erroneous sampling of an uninfected area). Thus, many investigations
have undertaken a search of clinical, laboratory, or imaging findings that may help
in the diagnosis of osteomyelitis.
In this issue of Clinical Infectious Diseases, Dinh et al. [8] present the results of
a meta-analysis of studies examining the
diagnostic accuracy of various clinical and
imaging methods for diabetic patients
with a foot ulcer. They elected to include
only studies that used histopathologic examination or culture of a bone specimen
as the reference for diagnosis of osteomyelitis. Although this criterion adds rigor
to their findings, it allowed them to select
only 9 studies for their analysis. By contrast, a systematic review of diagnostic
tests for diabetic foot osteomyelitis (published since the submission of the article
by Dinh et al. [8]) that included some
patients from whom no bone specimen
was obtained reported data from 21
528 • CID 2008:47 (15 August) • EDITORIAL COMMENTARY
publications [9]. This study and the one
by Dinh et al. [8] are otherwise very similar, seeking articles addressing the same
question for the same types of patients
over the same period. Surprisingly and for
unclear reasons, only 4 of the same studies
were selected by both groups for inclusion
in their analyses.
Thus, which evidence is useful for diagnosis of diabetic foot osteomyelitis?
Both Dinh et al. [8] and Butalia et al. [9]
concluded that the presence of exposed or
visible bone correlated with bone infection, but this conclusion was based on
only 2 studies. There was insufficient data
to support the value of any other clinical
finding, except perhaps the presence of a
foot ulcer with a size 12 cm2. Butalia at
al. [9] reviewed the value of laboratory
tests and concluded—again based on only
2 studies—that an erythrocyte sedimentation rate 170 mm/h significantly increased the probability of osteomyelitis.
With regard to imaging studies, both reviews concluded that MRI is the most accurate of the available tests. Plain radiography and WBC radionuclide scans are
moderately helpful, but bone scans are too
nonspecific to be useful. It is not easy to
directly compare the findings of the 2
studies, because the systematic review by
Butalia et al. [9] reported likelihood ratios
for the various diagnostic tests, and the
meta-analysis by Dinh et al. [8] provided
pooled diagnostic ORs and Q* values (i.e.,
a summary receiver operating characteristic that is less affected by heterogeneity).
It is worth considering diagnostic methods that neither study discussed, because
these methods have not been subject to
rigorous investigation. Experienced clinicians have advocated some clinical findings that may suggest osteomyelitis. These
include the presence of a break in the
skin—especially a chronic ulcer that is overlying a bony prominence—that affects
the forefoot (or the heel) rather than the
midfoot and that is deep [3]. Similarly, an
ulcer that is not healing (or especially,
deepening) despite appropriate care and
pressure off-loading suggests underlying
osteomyelitis [10]. Although both reviews
recommended the probe-to-bone test, the
test must be performed as described in the
studies demonstrating its usefulness (i.e.,
after debridement of the soft-tissue wound
and with a sterile metal—not a wooden
or plastic—probe). Also, as with other diagnostic tests, the performance characteristics of the probe-to-bone test depend on
the pretest probability of osteomyelitis in
the tested population [11]. With regard to
available laboratory tests, leukocytosis is
infrequent in patients with diabetic foot
osteomyelitis [12], but C-reactive protein
measurement may be useful, because the
C-reactive protein level is often elevated
in patients with bone infection but is normal in patients with Charcot foot [13, 14].
More recently, the serum procalcitonin
level has been shown to be a useful diagnostic marker of diabetic foot infection
[15, 16], but additional investigations are
required to determine the value of this test,
especially for diagnosis of osteomyelitis.
With regard to imaging tests, there are
some promising diagnostic approaches. It
may be possible to overcome the lack of
sensitivity of negative plain radiograph
findings for a patient with an acute softtissue wound by providing appropriate
treatment (including for any infection)
and then repeating the radiography assessment a few weeks later. Negative follow-up radiograph findings make the
presence of osteomyelitis unlikely, and the
development of new findings of bony erosion suggests that is osteomyelitis present
[10]. Of course, newer diagnostic methods
are continually being evaluated. One
method that is particularly promising is
the positron emission tomography scan
with 18F-fluorodeoxyglucose imaging
[17]. Studies have revealed that this
method can detect clinically unsuspected
osteomyelitis [14] and can accurately distinguish osteomyelitis from Charcot foot
[18].
A different approach to this diagnostic
dilemma is to develop a consensus scheme
that integrates the results of a range of
clinical, laboratory, and imaging findings.
This technique has been used in several
clinical situations, such as the Duke criteria for endocarditis, in which there is not
a single criterion sufficiently reliable for
making a diagnosis. To that end, the International Working Group on the Diabetic Foot appointed an expert advisory
group to suggest criteria for the diagnosis
of diabetic foot osteomyelitis that could
be used in future research [19]. The group
stratified levels of diagnostic certainty,
based on the posttest probabilities of various diagnostic tests (depending on their
relative values), into 4 categories of likelihood: definite (190%), probable (51%–
90%), possible (10%–50%), and unlikely
(!10%). In addition to using an individual
criterion, they proposed combinations of
test results that would determine the diagnostic category. The scheme may be useful for initial decisions regarding whether
additional diagnostic testing is needed or
whether initiation of empirical antibiotic
therapy is appropriate. It also allows for
changing the level of diagnostic certainty
over time as the course of infection
evolves. Of course, this proposed scheme
should currently only be used for research
purposes and must undergo validation for
use in clinical trials.
The meta-analysis by Dinh et al. [8] is
a useful review of the current methods for
diagnosis of osteomyelitis of the foot in
patients with diabetes. Make no bones
about it, with the combination of some
promising new diagnostic tests, the systematic review by Butalia et al. [9], and
the proposed research consensus scheme
from the International Working Group on
the Diabetic Foot, we are finally approaching the point of having the ability to accurately diagnose this relatively frequent
and potentially devastating infection.
Acknowledgments
Potential conflicts of interest. B.A.L.: no
conflicts.
References
1. Lew DP, Waldvogel FA. Osteomyelitis. Lancet
2004; 364:369–79.
2. Armstrong DG, Wrobel J, Robbins J. Are diabetes-related wounds and amputations worse
than cancer? Int Wound J 2007; 4:286–7.
3. Berendt AR, Lipsky BA. Is this bone infected
or not? Differentiating neuro-osteoarthropathy from osteomyelitis in the diabetic foot.
Curr Diab Rep 2004; 4:424–9.
4. Edmonds M, Foster A. The use of antibiotics
in the diabetic foot. Am J Surg 2004; 187:
25S–8S.
5. Soysal N, Ayhan M, Guney E. Differential diagnosis of Charcot arthropathy and osteomyelitis. Neuro Endocrinol Lett 2007; 28:
556–9.
6. Lipsky BA. Osteomyelitis of the foot in diabetic patients. Clin Infect Dis 1997; 25:
1318–26.
7. Lipsky BA, Berendt AR, Deery HG, et al.; Infectious Diseases Society of America. Diagnosis and treatment of diabetic foot infections.
Clin Infect Dis 2004; 39:885–910.
8. Dinh MT, Abad CL, Safdar N. Diagnostic accuracy of the physical examination and imaging tests for osteomyelitis underlying diabetic foot ulcers: meta-analysis. Clin Infect Dis
2008; 47:519–27 (in this issue).
9. Butalia S, Palda VA, Sargeant RJ, Detsky AS,
Mourad O. Does this patient with diabetes
have osteomyelitis of the lower extremity?
JAMA 2008; 299:806–13.
10. Hartemann-Heurtier A, Senneville E. Diabetic
foot osteomyelitis. Diabetes Metab 2008; 34:
87–95.
11. Lavery LA, Armstrong DG, Peters EJ, Lipsky
BA. Probe-to-bone test for diagnosing diabetic
foot osteomyelitis: reliable or relic? Diabetes
Care 2007; 30:270–4.
12. Armstrong DG, Lavery LA, Sariaya M, Ashry
H. Leukocytosis is a poor indicator of acute
osteomyelitis of the foot in diabetes mellitus.
J Foot Ankle Surg 1996; 35:280–3.
13. Petrova NL, Moniz C, Elias DA, BuxtonThomas M, Bates M, Edmonds ME. Is there
a systemic inflammatory response in the acute
Charcot foot? Diabetes Care 2007; 30:997–8.
14. Schwegler B, Stumpe KD, Weishaupt D, et al.
EDITORIAL COMMENTARY • CID 2008:47 (15 August) • 529
Unsuspected osteomyelitis is frequent in persistent diabetic foot ulcer and better diagnosed
by MRI than by 18F-FDG PET or 99mTcMOAB. J Intern Med 2008; 263:99–106.
15. Uzun G, Solmazgul E, Curuksulu H, et al.
Procalcitonin as a diagnostic aid in diabetic
foot infections. Tohoku J Exp Med 2007; 213:
305–12.
16. Jeandrot A, Richard JL, Combescure C, et al.
Serum procalcitonin and C-reactive protein
concentrations to distinguish mildly infected
from non-infected diabetic foot ulcers: a pilot
study. Diabetologia 2008; 51:347–52.
17. Kumar R, Basu S, Torigian D, Anand V,
Zhuang H, Alavi A. Role of modern imaging
techniques for diagnosis of infection in the era
of 18F-fluorodeoxyglucose positron emission
tomography. Clin Microbiol Rev 2008; 21:
209–24.
18. Basu S, Chryssikos T, Houseni M, et al. Potential role of FDG PET in the setting of diabetic neuro-osteoarthropathy: can it differ-
530 • CID 2008:47 (15 August) • EDITORIAL COMMENTARY
entiate uncomplicated Charcot’s neuroarthropathy from osteomyelitis and soft-tissue
infection? Nucl Med Commun 2007; 28:
465–72.
19. Berendt AR, Peters EJG, Bakker K, et al. Diabetic foot osteomyelitis: a progress report on
diagnosis and a systematic review of treatment. Diabetes Metab Res Rev 2008; 24(Suppl
1):S145–61.