EVIDENCE-BASED ANSWER:
Magnetic resonance imaging (MRI) has a higher sensitivity and specificity (90% and 79%) than plain radiography (54% and 68%) for diagnosing diabetic foot osteomyelitis. MRI performs somewhat better than any of several common tests—probe to bone (PTB), erythrocyte sedimentation rate (ESR) >70 mm/hr, C-reactive protein (CRP) >14 mg/L, procalcitonin >0.3 ng/mL, and ulcer size >2 cm2—although PTB has the highest specificity of any test and is commonly used together with MRI. No studies have directly compared MRI with a combination of these tests, which may assist in diagnosis (strength of recommendation [SOR]: B, meta-analysis of cohort trials and individual cohort and case control trial).
Experts recommend obtaining plain films when considering diabetic foot ulcers to evaluate for bony abnormalities, soft tissue gas, and foreign body; MRI should be considered in most situations when infection is suspected (SOR: B, evidence-based guidelines).
EVIDENCE SUMMARY
One-fifth of patients with diabetes who have foot ulcerations will develop osteomyelitis.1,2 Most cases of diabetic foot osteomyelitis result from the spread of a foot infection to underlying bone.2
MRI has highest sensitivity, probe to bone test is most specific
A meta-analysis3 of 9 cohort trials (8 prospective, 1 retrospective) of 612 patients with diabetes and a foot ulcer examined the accuracy of diagnostic methods for osteomyelitis (TABLE3,4). MRI had the highest sensitivity (90%), followed by bone scan (81%). Bone scan was the least specific (28%), however. Plain film radiography had the lowest sensitivity (54%). A PTB test was highly specific (91%) but had moderate sensitivity (60%). (PTB involves inserting a sterile, blunt stainless steel probe into an ulcerated lesion. If the probe comes to a hard stop, considered to be bone, the test is positive.)
A meta-analysis of 21 prospective and retrospective trials with 1027 diabetic patients with foot ulcers or suspected osteomyelitis found that ulcer size >2 cm2, PTB, and ESR >70 mm/hr were helpful in making the diagnosis.4
Combining ESR with ulcer size increases specificity
A prospective trial of 46 diabetic patients hospitalized with a foot infection examined the accuracy of a combination of clinical and laboratory diagnostic features in patients with diabetic foot osteomyelitis that had been diagnosed by MRI or histopathology.5 (Twenty-four patients had osteomyelitis, and 22 didn’t.)