Associations of diabetes mellitus with orthopaedic infections.

BACKGROUND: Clinical experience suggests that a high proportion of orthopaedic infections occur in persons with diabetes. METHODS: We reviewed several databases of adult patients hospitalized for orthopaedic infections at Geneva University Hospitals from 2004 to 2014 and retrieved 2740 episodes of infection. RESULTS: Overall, diabetes was noted in the medical record for 659 (24%) of these cases. The patients with, compared with those without, diabetes had more than five times more foot infections (274/659 [42%] vs 155/2081 [7%]; p < 0.01) and a significantly higher serum C-reactive protein level at admission (median 96 vs 70 mg/L; p < 0.01). Diabetic patients were older (median 67 vs 52 years; p < 0.01), more often male (471 [71%] vs 1398 [67%]; p = 0.04), and had more frequent polymicrobial infections (219 [37%] vs 353 [19%]; p < 0.01), including more gram-negative non-fermenting rods (90 [15%] vs 168 [9%]; p < 0.01). Excluding foot infections from these analyses did not change the statistically significant differences. Diabetes was present in 17% of all infected orthopaedic patients without foot involvement. In Geneva canton, the overall prevalence of diabetes is estimated at 5.1%, while we have found that the prevalence is 13% in our hospitalized adults. CONCLUSIONS: Diabetes is present in 24% of all adult patients hospitalized for surgery for an orthopaedic infection, a prevalence that is several times higher than for the general population and twice as high as that for the population of hospitalized patients. Compared with non-diabetics, patients with diabetes have significantly more infections that are polymicrobial, including gram-negative non-fermenting rods.

Administration of antibiotic agents before intraoperative sampling in orthopedic infections alters culture results.

UNLABELLED: Many physicians and surgeons think that prescribing antibiotics before intraoperative sampling does not alter the microbiological results. METHODS: Case-control study of adult patients hospitalized with orthopedic infections. RESULTS: Among 2740 episodes of orthopedic infections, 1167 (43%) had received antibiotic therapy before surgical sampling. Among these, 220 (19%) grew no pathogens while the proportion of culture-negative results in the 2573 who had no preoperative antibiotic therapy was only 6%. By multivariate analyses, pre-operative antibiotic exposure was associated with significantly more culture-negative results (odds ratio 2.8, 95% confidence interval 2.1-3.7), more non-fermenting rods and skin commensals (odds ratio 2.8 and 3.0, respectively). Even a single pre-operative dose of antibiotic was significantly associated with subsequent culture-negative results (19/93 vs. 297/2350; χ²-test, p = 0.01) and skin commensals (17/74 vs. 274/2350; p = 0.01) compared to episodes without preceding prophylaxis. CONCLUSIONS: Prior antibiotic use, including single-dose prophylactic administrations, is three-fold associated with culture-negative results, non-fermenting rods and resistant skin commensals.