By Richard R. Watkins, MD, MS, FACP, FIDSA

Associate Professor of Internal Medicine, Northeast Ohio Medical University; Division of Infectious Diseases, Cleveland Clinic Akron General, Akron, OH

Dr. Watkins reports that he has received research support from Allergan.

SYNOPSIS: A randomized study that included eight intensive care units in Europe found no reduction in mortality or carriage of antibiotic-resistant gram-negative pathogens with antibiotic cycling compared to antibiotic mixing.

SOURCE: van Duijn PJ, Verbrugghe W, Jorens PG, et al. The effects of antibiotic cycling and mixing on antibiotic resistance in intensive care units: A cluster-randomised crossover trial. Lancet Infect Dis 2018;18:401-409.

Preserving the effectiveness of antibiotics is crucial for maintaining the tremendous advances in medicine that have occurred over the past century. Two strategies that aim to protect antibiotics include antibiotic cycling, defined as prescribing one drug during a set period followed by a rotation to another drug for another set period, and antibiotic mixing, in which a drug from an alternative class is chosen for each consecutive patient. van Duijn and colleagues sought to determine which of these strategies is better for reducing carriage of antibiotic-resistant gram-negative bacteria in patients in intensive care units (ICUs).

The study was multicenter with a cluster-randomized, crossover design including medical, surgical, or mixed ICUs from several European countries. Following a four-month baseline period, ICUs were randomized to two nine-month periods of intervention (cycling or mixing) separated by a washout period of one month. The antibiotics used covered gram-negative pathogens and included cephalosporins, piperacillin-tazobactam, and carbapenems, but the selection of medication was left up to the treating physician. With mixing, the empiric treatment choice changed with each consecutive treatment course. With cycling, the preferred treatment was changed every six weeks. Investigators measured the unit-wide prevalence of antibiotic-resistant gram-negative bacteria by monthly screening cultures of the oropharynx and perineum of all patients in the ICU.

Antibiotic resistance was defined as carriage of Enterobacteriaceae harboring ESBL genes, phenotypical resistance to piperacillin-tazobactam, or carbapenemases for Acinetobacter spp. and Pseudomonas aeruginosa. The primary endpoint was the change in the prevalence unit wide of carriage, which could be from either culture site. Notably, it was possible for a single patient to be screened multiple times if they had a prolonged ICU stay. Secondary endpoints included mortality and length of ICU stay.

Three ICUs were assigned to the mixing approach followed by cycling and five were assigned to the cycling approach followed by mixing. Carbapenems were the most frequently used agents. During cycling, the prevalence of antibiotic-resistant gram-negative bacteria was 23%; during mixing, it was 22% (P = 0.64). There were no significant differences in the prevalence for specific bacterial species. The study antibiotics accounted for 42% of antibiotics used in the cycling group and 43% in the mixing group. In the ICU, mortality was 11% at baseline, 11% during cycling, and 12% during mixing (P = 0.38). Finally, there was no difference in the length of ICU stay between the mixing and cycling groups.

The researchers noted two major deviations from protocol. One involved an ICU in which the swabs were not collected during the final three months of the study, and the data from this period were excluded from the analysis. The other occurred when an outbreak of carbapenem-resistant Klebsiella pneumoniae made it impossible to adhere to the study protocol. Therefore, the investigators prolonged the washout period until the outbreak ended, which was five months, and antibiotic policy returned to what it had been before the outbreak.


The usefulness of antibiotic cycling and mixing has been debated for more than 20 years. The study by van Duijn and colleagues was well designed and strongly supports the idea that antibiotic cycling is not better than mixing for preventing the emergence and spread of antibiotic-resistant gram-negative bacteria. The investigators took care to limit confounding variables that could have influenced the results, such as the use of non-study antibiotics, infection control practices including hand hygiene, patient case mix, and changes to the proportion of patients who were colonized with antibiotic-resistant pathogens on admission to the ICU. Thus, since the study was high quality, it seems unlikely that another similarly designed trial would contradict these findings.

Given that the baseline mortality in the ICU did not change with cycling, it is reasonable to conclude that other antibiotic stewardship and infection control strategies should be given higher priority than either antibiotic cycling or mixing. For example, until an institution’s hand hygiene compliance rate reaches 100% for all staff members, there is room for improvement. Perhaps just as important is the ongoing need to reduce the total volume of antibiotics prescribed in ICUs. This should lead to reduced selection pressure for the development of antibiotic-resistant bacteria.

The finding that carbapenems were the most frequently prescribed agents in the study is concerning given the ongoing spread of carbapenem-resistant pathogens, such as Acinetobacter baumannii. The study was conducted between June 2011 and February 2014 and it is unclear whether carbapenem use has changed in European ICUs since then.

One limitation of the study is that it was conducted in ICUs in five European countries, so the findings may not be applicable to other settings. Another is that it was underpowered to detect resistance at the species level. Finally, the sample size did not meet the initial value for calculating the main outcome, although the investigators attributed this to not taking into account the crossover design, which overestimated the sample size needed.

In summary, van Duijn and colleagues have demonstrated convincingly that antibiotic cycling does not reduce antibiotic-resistant gram-negative bacteria in the ICU setting. Therefore, antibiotic stewardship efforts should focus on different strategies to mitigate the spread of antibiotic-resistant pathogens and promote the judicious use of antibiotics.