Comparison of the microbiology and antibiotic treatment among diabetic and nondiabetic patients hospitalized for cellulitis or cutaneous abscess

Abstract

Background: Among diabetics, complicated skin infections may involve gram-negative pathogens; however, the microbiology of cellulitis and cutaneous abscess is not well established.

Objective: To compare the microbiology and prescribing patterns between diabetics and nondiabetics hospitalized for cellulitis or abscess.

Design: Secondary analysis of 2 published retrospective cohorts.

Setting/patients: Adults hospitalized for cellulitis or abscess, excluding infected ulcers or deep tissue infections, at 7 academic and community facilities.

Methods: Microbiological findings and antibiotic use were compared among diabetics and nondiabetics. Multivariable logistic regression was performed to identify factors associated with exposure to broad gram-negative therapy, defined as receipt of at least 2 calendar days of β-lactamase inhibitors, second- to fifth-generation cephalosporins, fluoroquinolones, carbapenems, tigecycline, aminoglycosides, or colistin.

Results: Of 770 total patients with cellulitis or abscess, 167 (22%) had diabetes mellitus. Among the 38% of cases with a positive culture, an aerobic gram-positive organism was isolated in 90% of diabetics and 92% of nondiabetics (P = 0.59); aerobic gram-negative organisms were isolated in 7% and 12%, respectively (P = 0.28). Overall, diabetics were more likely than nondiabetics to be exposed to broad gram-negative therapy (54% vs 44% of cases, P = 0.02). By logistic regression, diabetes mellitus was independently associated with exposure to broad gram-negative therapy (odds ratio: 1.66, 95% confidence interval: 1.15-2.40).

Conclusion: In cases of cellulitis or abscess associated with a positive culture, gram-negative pathogens were not more common among diabetics compared with nondiabetics. However, diabetics were overall more likely to be exposed to broad gram-negative therapy suggesting this prescribing practice may not be not warranted.

Figure 1

Study schematic