Does complete resection of infected bone improve clinical outcomes in patients with diabetic foot osteomyelitis?

Abstract

The objective of the study was to compare outcomes in patients with complete surgical resection versus partial resection of diabetic foot osteomyelitis (OM). A post hoc analysis of 171 patients with OM was performed using data from two randomized clinical trials. OM was confirmed with bone culture or histopathology. Surgical culture specimens were obtained from resected bone and sent for histopathology and microbiology. Residual osteomyelitis (RO) was defined as a positive resected margin on culture or histopathology. No residual osteomyelitis (NRO) was defined as no growth from bone culture and no histopathological inflammation in the biopsy of the resection margin. Data from the 12-month follow-up were used to determine clinical outcomes. During the index hospitalization, NRO patients had significantly shorter duration of antibiotic therapy (NRO 21.0, 13.0-38.0 vs. RO 37.0, 20.8-50.0, p <0.01) and more amputations than patients with RO (NRO 89.9% vs. RO 60.9%, p <0.01). During the 12-month follow-up, patients with NRO also had significantly shorter duration of antibiotic therapy (NRO 42, 21.0-66.5 vs. RO 50.5, 35.0-75.0, p = 0.02). During the 12-month follow-up, there was no difference in ulceration at the same site (NRO 3.7%, RO 4.3% p = 0.85), hospitalization (NRO 32.6%, RO 34.8%, p = 0.76), total re-infections (NRO 25.3%, RO 29.3%, p = 0.56), re-infection with osteomyelitis (NRO 13.3% vs. 13.5%, p = 0.36), amputation (NRO 8.8%, RO 5.4%, p = 0.86) and time to wound healing in days (NRO 94, 41.0-365 vs. RO 106, 42.8-365, p = 0.77). Successful treatment of osteomyelitis was achieved by 86.7% and 86.5% of patients. During the index hospitalization, patients with no residual osteomyelitis had more amputations and were treated with antibiotics for a shorter duration. During the 12-month follow-up, patients with no residual osteomyelitis had shorter durations of antibiotics. There were no differences in re-infection, amputation, re-ulceration or hospitalization. Level of evidence: 1.

Keywords: amputation; diabetes; infection; osteomyelitis; ulcer.

FIGURE 1

Twelve‐month outcome comparison of patients with residual osteomyelitis and no residual osteomyelitis. This bar chart compares the 12‐month clinical outcomes in patients with residual osteomyelitis and no residual osteomyelitis. There were no differences in clinical outcomes after hospital discharge.

FIGURE 2

Kaplan–Meier survival plot comparing the time to heal in patients with residual osteomyelitis and no residual osteomyelitis. Kaplan–Meier survival analysis of reinfection by margin status. The Y‐axis represents the proportion healed, and the X‐axis represents the study period (days). The comparison in re‐infection in patients with residual osteomyelitis and no residual osteomyelitis demonstrates that there is no significant difference in time to healing between the two groups. The average time until healing for NRO = 73.7 ± 61.4 versus RO = 75.4 ± 65.1.

FIGURE 3

Kaplan–Meier survival analysis of reinfection by margin status. The Y‐axis represents the proportion without reinfected, and the X‐axis represents the study period (days). The comparison in re‐infection in patients with residual osteomyelitis and no residual osteomyelitis demonstrates that there is no significant difference in time to re‐infection between the two groups. The average time until healing for RO was 98.7 ± 86.6 and NRO = 72.8 ± 49.4. Kaplan–Meier survival analysis comparing the time to re‐infection in patients with a residual osteomyelitis and no residual osteomyelitis in demonstrates that there is no significant difference in time to re‐infection between the two groups.