Clinical Question: Which test is most accurate in diagnosing chronic osteomyelitis?
Setting: Various (meta-analysis)
Study Design: Meta-analysis (randomized controlled trials)
Synopsis: Researchers searched MEDLINE, EMBASE, and Current Contents for studies including at least 10 adults evaluated for chronic osteomyelitis with noninvasive diagnostic tests (i.e., radiography, positron emission tomography [PET], leukocyte scintigraphy, bone scintigraphy, gallium scintigraphy, magnetic resonance imaging, or computed tomography) who also had histology, culture results, or clinical follow-up. Because one of the hallmarks of a good diagnostic test is its ability to distinguish among similar conditions, it is noteworthy that they excluded studies that evaluated septic or aseptic loosening of prosthetic joints. The researchers only included studies that provided enough detail to calculate the characteristics of the various diagnostic tests.
Two reviewers independently determined study eligibility and extracted the data. Any discrepancies were settled by consensus. They also evaluated each study’s methodologic quality on six criteria. When they found excessive variability in the data across studies, they did subgroup analyses to determine if the tests perform differently in axial and peripheral skeletal infections.
A total of 23 studies with 1,269 patients were included. The methodologic quality was highly variable; nonetheless, the researchers pooled the data for all studies. PET, compared with any other test, was the most sensitive (96 percent) and specific (91 percent). This translates to a positive likelihood ratio of 10.7 (95% confidence interval [CI], 4.6 to 19.8) and a negative likelihood ratio of 0.04 (95% CI, 0.01 to 0.15). Some combinations of tests (i.e., bone scintigraphy plus leukocyte scintigraphy or bone scintigraphy plus gallium scintigraphy) had comparable sensitivity. These estimates are likely to be biased because poor-quality studies were included.
Bottom Line: PET is the most accurate test to diagnose chronic osteomyelitis. (Level of Evidence: 3a)