ACR Appropriateness Criteria®

Clinical Condition: Suspected Osteomyelitis in Patients with Diabetes Mellitus

Variant 1: Soft tissue edema without ulcer or neuroarthropathy.

Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.

Variant 2: Ulcer with no exposed bone without neuroarthropathy.

Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.

Variant 3: Ulcer with exposed bone without neuroarthropathy.

Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.

Variant 4: Neuropathy without ulcer

Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.

Variant 5: Neuroarthropathy with ulcer without exposed bone.

Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.

Variant 6: Neuroarthropathy with ulcer with exposed bone.

Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.

Summary of Literature Review

Through the last 50 years there has been much written about the diabetic foot, with little consensus as to whether, when, and what imaging is appropriate. This overview will summarize some of the work and draw conclusions based on the available data. Several clinical situations will be discussed in which osteomyelitis or diabetic pedal disease is suspected, but clinical findings differ because of the presence or absence of edema ulceration and neuropathy.

Please note that although several of the variants have similar recommendations, they do present as unique clinical scenarios.

Soft-Tissue Edema without Ulceration

First, the probability of having osteomyelitis in a diabetic foot without evidence of ulceration is extremely low. Whether there is or is not soft-tissue swelling, these patients have almost no incidence of osteomyelitis and a low incidence of septic arthritis, but some frequency of soft-tissue infections. The only situation in which such a patient can have osteomyelitis is the presence of a "hidden" ulcer that has granulated over and may appear healed. In that situation the risk of osteomyelitis is still extremely low, since the ulcer would not have granulated over if osteomyelitis were present. Therefore, without a clinically apparent ulcer, the role of imaging might be to diagnose neuropathic disease or to see if there is soft-tissue infection only.

Neuropathy without Ulcer

A more difficult question is whether it is the neuroarthropathy or the soft-tissue infection that is causing the soft-tissue swelling. In a patient who has neuroarthropathy, the risk of infection is usually low without ulceration. Radiography can be used as a screening examination. Computed tomography (CT) may pick up neuroarthropathy, which may not be apparent radiographically and may be the cause of the swelling and pain (mimicking infection). CT can rarely exclude the diagnosis of osteomyelitis definitively if there is no edema in the marrow (fat is visible).

Scintigraphy is of indeterminate insensitivity and specificity, whether it is bone scan, indium or indium with sulfur colloid, or even positron emission tomography (PET). Flow images are the best discriminators of infection, but remain imperfect. Magnetic resonance imaging (MRI) likely has the best clinical results in this scenario with or without contrast, but the yield is going to be low in this clinical group of patients, and it is costly.

There is some importance in diagnosing neuropathic disease prior to radiographic changes, as these patients will be treated with altered footwear and orthotics to prevent the progression to deformity. Scintigraphy is, however, extremely sensitive to early neuropathic disease, long before radiographic changes are present. MRI is less sensitive but is a better test if there is a possibility of soft-tissue infection.

Ulcer with Exposed Bone

If an ulcer is present, the risk of infection is quite high, and almost invariable if the ulcer reaches bone. The role of imaging would be to confirm the infection and show extent. Radiography will show the infection, however late. Bone scan is quite nonspecific. Surprisingly, indium scan, even when combined with sulfur colloid marrow imaging, has low specificity, although if the ulcer is away from the joint these techniques are better. MRI has high specificity and sensitivity both with and without contrast. Ultrasound (US) may have promise in long bones but, to date, data about its utility in diagnosing the diabetic foot are quite limited. PET results are similarly poor, as this technique shows metabolic activity primarily and therefore is not specific.

Ulcer with Neuropathy and Exposed Bone

In patients who have diabetes and secondary neuroarthropathy, the infection is usually over an osseous abnormality with an ulcer. If the ulcer tracks down to bone, the risk of osteomyelitis is extremely high, perhaps even higher than in the preceding situation where there is an ulcer without neuropathic deformity. The overall role of imaging therefore, is more to determine the extent of the disease than to definitively diagnose it. Therefore, most authors do not advocate scintigraphy in this situation because of its relative poor spatial resolution for extent of disease; similar conclusions apply to PET.

Similarly, indium-labeled WBC (white blood cell) scanning with or without bone marrow scanning has only mixed sensitivity and specificity for osteomyelitis with neuropathy and yields poor anatomic extent of infection. Radiography has a high specificity but low sensitivity. US is unproven. CT will show the neuroarthropathic disease but not much else. MRI should be performed to determine extent of disease. T1 and fat-suppressed sequences are complementary, and contrast may or may not be used. The use of contrast is more to see the extent of the disease as well as the extent of vascularity, rather than to diagnose infections. Contrast may also help identify necrotic or poorly perfused regions, and to aid in surgical planning.

Summary and Recommendations

If a patient has an ulcer that extends to bone, there is quite likely, but not invariably, osteomyelitis. The best way to confirm this diagnosis and determine the extent of disease is with MRI. If there is no ulcer and there is still a clinical suspicion of infection, MRI is the test of choice. However, conventional radiographs should be done simultaneously in both situations. In indeterminate cases, aspiration and biopsy would be the next step.

If there is soft-tissue swelling the question is, "Is there early neuropathic disease or infection present?" Radiographs should be performed first. If the radiographs are normal, another test should be performed. If the suspicion of infection is low, the next test should probably be a three-phase bone scan. If there is a modest risk of infection, MRI is probably indicated.

Anticipated Exceptions

The following are available:

The National Guideline Clearinghouse™ (NGC) does not develop, produce, approve, or endorse the guidelines represented on this site.

All guidelines summarized by NGC and hosted on our site are produced under the auspices of medical specialty societies, relevant professional associations, public or private organizations, other government agencies, health care organizations or plans, and similar entities.