Medical Imaging and Laboratory Analysis of Diagnostic Accuracy in 107 Consecutive Hospitalized Patients With Diabetic Foot Osteomyelitis and Partial Foot Amputations

Abstract

The primary aim of our study was to compare the preoperative diagnostic accuracy of plain radiographic findings with the accuracy of magnetic resonance imaging (MRI) findings for diabetic foot osteomyelitis in hospitalized patients who underwent first-time partial foot amputations with confirmed histopathological specimens positive for osteomyelitis. Second, it was desired to determine whether certain variables within the initial clinical presentation and preoperative laboratory findings were associated with more accurate diagnosis of diabetic foot osteomyelitis in this study population. Finally, it was desired to determine the most common bacterial organisms found in bone and soft-tissue cultures taken intraoperatively and to determine how often the same organism was found in both. After applying the inclusion and exclusion criteria to the initial 329 patients identified through chart review, the final sample size for further analysis was n =107. In this study, after adjusting for the effects of covariates such as age, erythrocyte sedimentation rate (ESR) and C-reactive protein, plain radiographs seemed to have statistically more significant power than MRI in predicting and diagnosing diabetic foot osteomyelitis. In addition, higher ESR values were confirmed to predict a higher chance of positive diagnosis for diabetic foot osteomyelitis. Furthermore, the presence of positive bacterial identification from intraoperative bone cultures did not always indicate true osteomyelitis on histopathological examination. Levels of Evidence: Level II: Diagnostic study