The association between bacteria and outcome and the influence of sampling method, in people with a diabetic foot infection

Abstract

Purpose: Different bacteria lead to divers diabetic foot infections (DFIs), and some bacteria probably lead to higher amputation and mortality risks. We assessed mortality and amputation risk in relation to bacterial profiles in people DFI and investigated the role of sampling method.

Methods: We included people (> 18 years) with DFI in this retrospective study (2011-2020) at a Dutch tertiary care hospital. We retrieved cultures according to best sampling method: (1) bone biopsy; (2) ulcer bed biopsy; and (3) swab. We aggregated data into a composite determinant, consisting of unrepeated bacteria of one episode of infection, clustered into 5 profiles: (1) Streptococcus and Staphylococcus aureus; (2) coagulase-negative Staphylococcus, Cutibacterium, Corynebacterium and Enterococcus; (3) gram-negative; (4) Anaerobic; and (5) less common gram-positive bacteria. We calculated Hazard Ratio's (HR's) using time-dependent-Cox regression for the analyses and investigated effect modification by sampling method.

Results: We included 139 people, with 447 person-years follow-up and 459 episodes of infection. Sampling method modified the association between bacterial profiles and amputation for profile 2. HR's (95% CI's) for amputation for bacterial profiles 1-5: 0.7 (0.39-1.1); stratified analysis for profile 2: bone biopsy 0.84 (0.26-2.7), ulcer bed biopsy 0.89 (0.34-2.3), swab 5.9*(2.9-11.8); 1.3 (0.78-2.1); 1.6 (0.91-2.6); 1.6 (0.58-4.5). HR's (95% CI's) for mortality for bacterial profiles 1-5: 0.89 (0.49-1.6); 0.73 (0.38-1.4); 2.6*(1.4-4.8); 1.1(0.58-2.2); 0.80(0.19-3.3).

Conclusions: In people with DFI, there was no association between bacterial profiles in ulcer bed and bone biopsies and amputation. Only in swab cultures, low-pathogenic bacteria (profile 2), were associated with a higher amputation risk. Infection with gram-negative bacteria was associated with a higher mortality risk. This study underlined the possible negative outcome of DFI treatment based on swabs cultures.

Keywords: Amputation; Bacteria; Diabetes mellitus; Diabetic foot infection; Mortality.

Fig. 1

Step 1: cultures obtained from one participant during the course of 1 infection episode. There were 3 cultures obtained during this episode. 1. Swab: E.coli. 2. Ulcer bed: S. aureus and bone biopsy: E. coli. 3. Swab: Corynebacterium and bone biopsy: S. aureus. Step 2. Selecting best sampling method. Bone biopsy prevails over swab and ulcer bed biopsy, we therefore discarded the results of swab and ulcer bed biopsy (red crosses). Step 3. We created a composite determinant consisting of bacteria from best sampling method, and in which we removed duplicate bacterial species, e.g., if a S. aureus was cultured twice during one infective episode, we only counted this bacterial species ones (unrepeated bacterial species). After deduplication of bacterial species we grouped these bacteria into one of the five profiles. S. aureus causes acute, usually mild, infections and is grouped with Streptococcus in profile 1, and E. coli is a gram-negative bacterium, causing usually severe infection and is grouped in profile 3. For the analyses we used profile 1 and profile 3 in this example