This article is more than 5 years old. Information may no longer be current.
Antibiotic mixing, cycling strategies fail to reduce resistance in ICUs
Click Here to Manage Email Alerts
Researchers examining the impact of antibiotic mixing and antibiotic cycling strategies in ICUs across Europe found that neither intervention significantly changed the overall prevalence of antibiotic-resistant, gram-negative bacteria.
Pleun Joppe van Duijin, MD, of the department of epidemiology at Julius Center for Health Sciences and Primary Care, Utrecht, Netherlands, and colleagues reported in The Lancet Infectious Diseases that their findings suggest that “one strategy cannot be recommended over another, and neither strategy can be recommended in ICUs.”
Both antibiotic mixing and cycling interventions involve a scheduled alternation of first-line empirical therapy to increase diverse antibiotic use, thereby reducing resistance, according to the researchers. In antibiotic cycling, a specific antibiotic is recommended as first-line therapy for all patients who require treatment during a prespecified period. After this period, another antibiotic with different selective properties becomes the recommended treatment. In antibiotic mixing, antibiotics with different selective properties are alternated after each patient receives treatment.
Because data from previous clinical trials evaluating the interventions provide “inconclusive results,” van Duijin and colleagues conducted a cluster-randomized crossover trial at eight ICUs in Belgium, France, Germany, Portugal and Slovenia to better understand their impact in the ICU setting.
After a 4-month baseline period in which ICUs provided standard care, the facilities were randomly assigned to implement a 9-month mixing or cycling intervention. Then, after a 1-month “washout period,” the alternative intervention was implemented for another 9 months. During the cycling interventions, ICUs interchanged the use of third- or fourth-generation cephalosporins, piperacillin-tazobactam and carbapenems as preferred empirical therapy every 6 weeks. During the mixing interventions, the three different antibiotic groups were rotated after each consecutively treated patient. The primary endpoint was the prevalence of antibiotic-resistant, gram-negative bacteria in respiratory and perineal swabs collected during monthly point prevalence visits at each ICU.
From June 2011 to February 2014, 4,069 patients were admitted to an ICU during cycling periods and 4,707 were admitted during mixing periods. Among these patients, 745 enrolled during a cycling period and 853 enrolled during a mixing period were present during a point prevalence survey and included the in the primary analysis.
The prevalence of antibiotic-resistant, gram-negative bacteria was 28% at baseline, 23% during cycling periods and 22% during mixing periods (adjusted incidence rate ratio = 1.039; 95% CI, 0.837-1.291). There was no significant difference in all-cause, in-ICU mortality, which was 11% during baseline, 11% during cycling and 12% during mixing. Patient and ICU characteristics such as hand hygiene adherence, prevalence of isolation precautions and staffing were similar among the facilities, according to the researchers.
In a related editorial, Philipp Scheutz, MD, of the University Department of Medicine, Kantonsspital Aarau, and the University of Basel, Switzerland, and Robert Eric Beardmore, PhD, professor at Biosciences University of Exeter, said the results “strongly suggest” that antibiotic cycling does not provide a benefit over antibiotic mixing, and there is an “urgent need” for new strategies to effectively reduce resistance.
“Until then, we need to reinforce the patient-level tools that are available, including (among others) improved hand hygiene and better selection of patients in need of antibiotics by host-response markers such as procalcitonin and other pathogenic markers,” they wrote. – by Stephanie Viguers
Disclosures: Beardmore is supported by a UK EPSRC Healthcare Technology Impact Fellowship and previously received research support from AstraZeneca. Scheutz is supported by the Forschungsrat of the Kantonsspital Aarau and the Swiss National Science Foundation, and reports receiving research support from Abbott, BioMerieux, BRAHMS/Thermofisher, Nestle, Novo Nordisk and Roche. van Duijn reports no relevant financial disclosures. Please see the study for all other authors’ relevant financial disclosures.
Perspective
Back to Top
van Duijn and colleagues conducted a study that is well-designed, exploring in eight ICUs two potential interventions to minimize the likelihood of developing antimicrobial resistance among gram-negative pathogens — antibiotic cycling and antibiotic mixing.
To put this study in context, we have abundant published evidence that antibiotic use with a given drug or drug class will promote resistance to the agent(s) used among bacteria that are colonizing or causing infection in that patient population. A perusal of the ESAC-Net and EARS-Net surveillance networks from Europe provides visual ecological evidence of this association. With this in mind, we would do well to remember that there is no substitute for decreasing antimicrobial use and choosing narrow spectrum-of-activity agents whenever feasible.
van Duijn and colleagues have given us some of the best data and analysis to date on two antimicrobial stewardship strategies to minimize the development of resistance when broad-spectrum, gram-negative agents are used. It is notable as the authors point out that without a significant change in the overall volume of antibiotics used, their study successfully described the impact of cycling and mixing strategies.
Given the rates of antibiotic use and resistance, it is not surprising that these interventions were unsuccessful. Their findings are consistent with other studies of clinical interventions and recent analyses (eg, Beardmore and colleagues). Though both interventions seemed to show a modest, nonsignificant impact on the prevalence of extended-spectrum beta-lactamases (genotypic or phenotypic), the observed overall prevalence of drug resistance in gram-negative pathogens was highly variable (0% to > 75%), and the data preclude stratifying by hospitals with high and low baseline prevalence. The authors built a model adjusting for hand hygiene compliance and the proportion of short-stay patients; however, being a pragmatic trial, there may be other important confounders such as patient acuity and conditions (eg, transplant) that may obscure a true difference in the interventions.
Two aspects of the study design may offer the only realistic potential that repeating this endeavor may yield different results. First, it is plausible that the bacterial resistance ecology of a patient population may take more than 9 months of intervention to see a significant change in antimicrobial resistance (eg, Molina and colleagues). Second, the use of monthly prevalence screening may not be a precise estimate of the development (and potential transmission/nosocomial acquisition) of and subsequent infection with antimicrobial-resistant organisms in their ICUs or institution.
In short, the authors were correct when they stated, “the epidemiology of antibiotic resistance in ICUs is complex.” Their data suggest that antimicrobial stewards are likely to find a logistically challenging intervention such as antibiotic cycling or mixing will demonstrate minimal, if any, impact on the pattern of resistance in their facility.
References:
Beardmore RE, et al. Mol Biol Evol. 2017;doi:10.1093/molbev/msw292.
Molina J, et al. Clin Infect Dis. 2017;doi:10.1093/cid/cix692.
Click Here to Manage Email Alerts