Ciprofloxacin use and susceptibility of Gram-negative organisms to quinolone and non-quinolone antibiotics.

OBJECTIVES: A Trust strategy to reduce ciprofloxacin use was implemented at a University hospital. This study aimed to investigate whether the susceptibility of Gram-negative organisms (GNO) to alternative antimicrobials (co-amoxiclav, doxycycline, aztreonam, piperacillin/tazobactam, meropenem and gentamicin) changed, and whether there was any relationship between GNO susceptibility to these antimicrobials and ciprofloxacin usage. METHODS: The first isolate of each GNO from blood cultures, sputum and urine of hospitalized adults, between January 2008 and August 2009, was included. Antibiotic usage and GNO susceptibility were investigated using linear regression. The association between defined daily dose/1000 occupied bed days (DDD/1000 OBD) and susceptibility was assessed using Pearson correlation and linear regression. RESULTS: Ciprofloxacin use decreased significantly by 4.37 DDD/1000 OBD per month [95% confidence interval (CI) 2.99, 5.75; P < 0.001], while aztreonam and gentamicin use increased significantly (aztreonam: 0.22 DDD/1000 OBD increase per month; 95% CI 0.10, 0.34; P = 0.001; gentamicin: 0.46 DDD/1000 OBD increase per month; 95% CI 0.12, 0.79; P = 0.01). There was no change in meropenem, co-amoxiclav, doxycycline or piperacillin/tazobactam use. When DDD/1000 OBD for all non-quinolone antimicrobials were pooled, use increased significantly by 3.33 DDD/1000 OBD per month (95% CI 0.79, 5.87; P = 0.013). There were 5410 GNO isolates. A significant increase was recorded in the proportion of GNO susceptible to ciprofloxacin (0.55% increase in susceptibility per month; 95% CI 0.38, 0.72; P < 0.001), aztreonam (1.87% susceptibility increase per month; 95% CI 1.18, 2.55; P < 0.001), piperacillin/tazobactam (0.18% susceptibility increase per month; 95% CI 0.03, 0.33; P = 0.021), meropenem (0.27% susceptibility increase per month; 95% CI 0.08, 0.47; P = 0.009) and gentamicin (0.17% susceptibility increase per month; 95% CI 0.04, 0.29; P = 0.011). An inverse association between ciprofloxacin use and susceptibility to ciprofloxacin (P < 0.001), piperacillin/tazobactam (P = 0.12), aztreonam (P= 0.002), meropenem (P = 0.015) and gentamicin (P = 0.034) is suggested. CONCLUSIONS: These data demonstrate reduced ciprofloxacin usage and concomitant increasing GNO susceptibility to ß-lactams. While definitive evidence of a causal relationship is beyond the capability of a single-centre study, the results suggest that reducing quinolone exposure may exert a favourable effect on the quinolone, ß-lactam and gentamicin susceptibility of GNO.