Current issues of ACP Journal Club are published in Annals of Internal Medicine


The identification of palpable bone in diabetic pedal ulcers by blunt sterile probe was indicative of osteomyelitis

ACP J Club. 1995 Nov-Dec;123:74. doi:10.7326/ACPJC-1995-123-3-074

Source Citation

Grayson ML, Gibbons GW, Balogh K, Levin E, Karchmer AW. Probing to bone in infected pedal ulcers. A clinical sign of underlying osteomyelitis in diabetic patients. JAMA. 1995 Mar 1;273:721-3.



To evaluate the bedside technique of probing infected pedal ulcers for palpable bone for the detection of osteomyelitis in patients with diabetes.


A blinded comparison of gentle probing of foot ulcers with histologic examination.


A U.S. tertiary care center.


75 patients (mean age 60 y, 69% men) who were hospitalized and had a total of 76 infected pedal ulcers. Patients with nonhealed recent surgical wounds or exposure of the adjacent bone during debridement were excluded.

Description of test and diagnostic standard

The ulcer was probed with a sterile, blunt, 14.0-cm, 5-F, stainless steel eye probe for palpable bone (rock-hard structure at the ulcer base). Open ulcers were probed before debridement, and ulcers covered by an eschar were probed after removal of the eschar. If bone was palpable, exposed during debridement, or resected, biopsy specimens were taken. A pathologist who was unaware of the probe test results did the histologic examination. Osteomyelitis was diagnosed if evidence showed inflammatory cells within the bone, fibrosis of intertrabecular soft tissue, and destruction or necrosis of bone and reactive new bone formation. Radiologic osteomyelitis and bone appearance at surgery were used for diagnosis if bone was not available for histologic examination. Absence of osteomyelitis required healing of the ulcer and no recurrence after abbreviated antibiotic therapy.

Main outcome measures

Sensitivity, specificity, and likelihood ratios.

Main results

Osteomyelitis was diagnosed in 50 (66%) of the 76 ulcers, 46 times by histologic examination. In the remaining 4 ulcers, the diagnosis was made on the basis of roentgenographic changes (n = 1), by clinical and roentgenographic criteria (n = 1), and by clinical criteria (n = 2). The bone was palpable by probe in 33 of the 50 osteomyelitis-associated ulcers . Of the 26 ulcers without osteomyelitis, the bone was palpable by probe in 4 . Sensitivity, specificity, and likelihood ratios are shown in the Table.


The identification of palpable bone in infected pedal ulcers by blunt, sterile probe was indicative of osteomyelitis in patients with diabetes.

Source of funding: Roerig Division, Pfizer, Inc.

For article reprint: Dr. A.W. Karchmer, Division of Infectious Diseases, New England Deaconess Hospital, 185 Pilgrim Road, Boston, MA 02215, USA. FAX 617-632-0766.

Table. Test characteristics of probing to bone in diabetic pedal ulcers for detecting osteomyelitis*

Outcome Sensitivity (95% CI) Specificity (CI) +LR -LR
Osteomyelitis 66% (51 to 79) 85% (65 to 96) 4.3 0.4

*LRs defined in Glossary; LRs and CI calculated from data in article.


When patients with diabetes develop ulcers on their feet, osteomyelitis that is underlying the ulcers increases the risk that these patients will eventually need an amputation (1). Osteomyelitis is commonly diagnosed with imaging procedures, but these procedures have poor sensitivity and specificity and are expensive (2).

The patients in the study by Grayson and colleagues represent the most severely affected; they were included because they had "severe, limb-threatening foot infection." The investigators therefore found a higher prevalence of bone exposed by probing and of osteomyelitis than did previous investigators (1). This study did not evaluate the reproducibility of the technique for probing to bone, nor did it assess whether probing to bone is a better predictor of osteomyelitis than the physician's overall clinical judgment.

The sensitivity, specificity, and likelihood ratios of probing to bone are better than those of plain radiograph and bone scan but are worse than those of 111indium scanning and magnetic resonance imaging (2).

To illustrate the use of probing, if a physician determined that a patient with diabetes who had a foot ulcer had a 50% probability of osteomyelitis before a probe was passed, the presence of exposed bone would increase the probability to 81% and the absence would decrease it to 29%. In this situation, the presence of exposed bone on probing into the ulcer might lead the clinician to treat the patient for osteomyelitis; the absence of exposed bone might lead to conservative therapy or further diagnostic evaluation. In situations in which the pretest probability of osteomyelitis is very low or very high, however, probing to bone changes the absolute probability of osteomyelitis much less. Therefore, probing to bone is useful (if not definitive) when the clinician determines that the probability of osteomyelitis is intermediate but should not overrule the clinician's strongly held judgment about the presence or absence of osteomyelitis.

David Edelman, MD
Durham Veterans Affairs Medical CenterDurham, North Carolina, USA


1. Newman LG, Waller J, Palestro CJ, et al. Unsuspected osteomyelitis in diabetic foot ulcers. Diagnosis and monitoring by leukocyte scanning with indium in 111 oxyguinoline. JAMA. 1991;266:1246-51.

2. Eckman MH, Greenfield S, Mackey WC, et al. Foot infections in diabetic patients. Decision and cost-effectiveness analyses. JAMA. 1995;273:712-20.