Authors Study Design Setting Protocol Outcomes Indicator
Toltzis P et al 2002
Controlled trial
Neonatal ICU
Monthly cycling of gentamicin, piperacillin-tazobactam and ceftazidime for suspected infections due to Gram-negative pathogens versus standard practice in the control group (usually ampicillin and gentamicin for suspected infection at birth, vancomycin and gentamicin for hospital-acquired infection, ampicillin and cefotaxime for meningitis, and piperacillin-tazobactam for necrotizing enterocolitis) No de-escalation Typing to assess clonality of bacterial isolates PRIMARY Similar incidence of colonization with resistant bacilli to any antibiotic Similar incidence of colonization with resistant bacilli to the rotated antibiotics (even when only data regarding clonally discordant isolates were considered) OTHER On-cycle antibiotic use 84.3% for the rotation team Predominant use of gentamicin in the control team Similar overall antibiotic use Similar length of stay
Unit-wide surveillance cultures
Cadena J et al 2007
Before-and-after
Haematology-Oncology Unit
Cycling of piperacillin-tazobactam and cefepime for the empirical therapy of neutropenic fever every three months versus standard practice during a baseline period (not further clarified) Potential of de-escalation not clarified No typing of bacterial isolates to assess clonality PRIMARY Inconclusive changes in relevant susceptibilities of Enterobacterales and P. aeruginosa Decrease in ampicillin-susceptible Enterococcus spp, erythromycin- and clindamycin-susceptible S. aureus OTHER Increase in cefepime and piperacillin-tazobactam consumption index from 0.003 to 0.88 Increase in cefepime use
Unit-wide clinically indicated cultures
Bennett KM et al 2007
Before-and-after
Surgical ICU
Cycling of piperacillin-tazobactam, imipenem, ceftazidime and ciprofloxacin every month for the empirical treatment of suspected Gram-negative infections (Ciprofloxacin discarded later) versus standard practice during a baseline period (not further clarified) De-escalation permitted No typing of bacterial isolates to assess clonality PRIMARY Increase in piperacillin-tazobactam and ceftazidime-susceptible P. aeruginosa proportions; No changes for the Medical ICU (Used as a comparison unit) Inconclusive changes for E. coli and K. pneumoniae in the Surgical ICU; Increase in piperacillin-tazobactam-resistant E. coli proportions and inconclusive changes for K. pneumoniae in the Medical ICU OTHER No information provided regarding secondary outcomes
Unit-wide clinically indicated cultures
Smith R et al 2008
Before-and-after
Surgical ICU
Cycling of vancomycin and linezolid for suspected Gram-positive infections every three months versus primary vancomycin use during a baseline period De-escalation permitted No typing of bacterial isolates to assess clonality PRIMARY Decrease in MRSA incidence rates during cycling Similar VRE incidence rates OTHER Similar percentage of in-hospital deaths according to initial empirical therapy Similar incidence rates of C. difficile colitis
Unit-wide clinically indicated cultures
Nijssen S et al 2009
Prospective comparative cross-over trial
2 ICUs (Medical ICU and Neurosurgery ICU) Weekly cycling of ceftriaxone, amoxicillin-clavulanate and levofloxacin or ciprofloxacin as empirical treatment versus the homogeneous administration of ciprofloxacin or levofloxacin No de-escalation Typing of isolates to exclude clonal outbreaks PRIMARY Higher colonization rates for ciprofloxacin-resistant isolates (including ciprofloxacin-resistant cephalosporin-resistant isolates) during the homogeneous period Similar colonization rates for cephalosporin-resistant Enterobacteriaceae OTHER Similar overall antibiotic use Higher ciprofloxacin use during the homogeneous period Lower third-generation cephalosporin use during the homogeneous period
Unit-wide surveillance cultures
Raineri E et al 2010
Before-and-after
2 ICUs
Cycling of piperacillin-tazobactam, fluoroquinolones, carbapenems, cefepime/ceftazidime every three months for the empirical treatment of VAP versus standard practice in a baseline period (most commonly piperacillin-tazobactam or levofloxacin) No de-escalation No typing of bacterial isolates to assess clonality PRIMARY Similar incidence of VAP due to antibiotic-resistant bacteria Decrease in cefepime- and aminoglycoside-resistant P. aeruginosa isolates Decrease in cefazolin-resistant K. pneumoniae and E. coli isolates from No other conclusive changes OTHER On-cycle antibiotic use 83% in Unit 1 and 88% in Unit 2 Increase in carbapenem and extended-spectrum penicillin use Decrease in aminoglycoside, fluoroquinolone, 3GC and 4GC use Similar mortality rates
Respiratory cultures derived from Ventilator-associated Pneumonia cases
Cumpston A et al 2012
Before-and-after
Blood and Marrow Transplantation Unit
Pre-cycling period: No prophylaxis for neutropenia;* Piperacillin-tazobactam for the empirical treatment of febrile neutropenia Period A: Cycling of imipenem, cefepime plus+ tobramycin and piperacillin-tazobactam plus tobramycin every eight months for the empirical treatment of febrile neutropenia; Levofloxacin as prophylaxis for neutropenia* Period B: Cycling of agents every three months; Addition of tobramycin in the imipenem arm; Levofloxacin as prophylaxis for neutropenia* *Addition of vancomycin at the discretion of the clinician De-escalation permitted No typing to assess clonality of bacterial isolates PRIMARY Increase in quinolone-resistant Enterobacterales incidence rates Increase in VRE incidence rates No other conclusive changes in resistance patterms OTHER Decrease in vancomycin use Similar use of cefepime, piperacillin-tazobactam and imipenem across the four most recent years of cycling Decrease in incidence rate of Klebsiella and E. coli bacteremia and candidemia Similar morbidity and mortality incidence rates
Unit-wide blood cultures
Chong Y et al 2013
Before-and-after
Haematology Unit
Monthly cycling of piperacillin-tazobactam, ciprofloxacin, meropenem and cefepime for the empirical treatment of neutropenic fever versus the homogeneous use of cefepime during a baseline period Potential of de-escalation not clarified No typing of bacterial isolates to assess clonality PRIMARY Blood isolates: Decrease in cefepime-resistant isolate incidence from 6/13 (70% of those were ESBLs) to 01/14 (p=0.007); Decrease in ciprofloxacin-resistant isolate incidence Stool isolates: Decrease in ESBL and ciprofloxacin-resistant E. coli incidence OTHER Similar mortality rates 65.9% decrease in unit-wide cefepime-use
Blood and stool cultures from patients with neutropenic fever
Teranishi H et al 2017
Before-and-after
Paediatric Haematology Unit
Monthly cycling of piperacillin-tazobactam, meropenem and cefepime versus the homogeneous prescription of cefpirome as empirical treatment for neutropenic fever during a baseline period No de-escalation No typing of bacterial isolates to assess clonality PRIMARY Blood isolates: Decrease in ESBL incidence from 5/15 to 0/15 isolates (p< 0.05) Nasal and stool isolates: Decrease in ESBL incidence from 15/33 to 0/33 isolates (p<0.01) Similar MRSA and VRE incidence OTHER No information provided regarding secondary outcomes
Blood, nasal and stool cultures from patients with neutropenic fever
Tsukayama D et al 2004
Comparative trial
ICU
Cycling of ciprofloxacin or levofloxacin plus clindamycin or metronidazole with piperacillin-tazobactam every four months as first-line empirical treatment De-escalation permitted Typing to assess clonality of bacterial isolates PRIMARY No correlation between particular antibiotic class consumption and onset of resistance OTHER Off-cycle antibiotic use not drastically reduced
Unit-wide surveillance units
Van Loon H et al 2005
Comparative trial
ICU
Cycling of levofloxacin plus aminoglycoside with beta-lactam plus aminoglycoside (cefpirome in one cycle and piperacillin-tazobactam in the other) every four months for suspected Gram-negative infections No de-escalation No typing of bacterial isolates to assess clonality PRIMARY Colonization rates for Gram-negative bacteria resistant to levofloxacin higher in periods of exposure Colonization rates for Gram-negative bacteria resistant to cefpirome similar between periods of exposure and non-exposure Colonization rates for Gram-negative bacteria resistant to piperacillin-tazobactam higher in periods of exposure OTHER On-cycle antibiotic use 88.5%-100%
Unit-wide surveillance cultures
Ginn A et al 2012
Comparative trial
2 ICUs
Cycling of piperacillin-tazobactam and cefepime for the empirical therapy of sepsis every four months Potential of de-escalation not clarified Typing of isolates to exclude clonal outbreaks PRIMARY Proportion of admissions complicated by antibiotic-resistant isolates higher in cefepime cycles Proportion of admissions complicated by MRSA higher in cefepime cycles OTHER Similar risk of admissions complicated by any infection On-cycle antibiotic use>60% of total use Off-cycle antibiotic use<15% of total use
Unit-wide clinically indicated cultures
Martinez J et al 2006
Comparative cross-over trial
2 ICUs
1st arm: Cycling of cefepime (or ceftazidime), ciprofloxacin, carbapenems, and piperacillin-tazobactam every month for suspected Pseudomonas infections 2ndarm: Successive administration of these agents to consecutive patients Potential of de-escalation not clarified Combination therapy permitted No typing to assess clonality of bacterial isolates PRIMARY Higher proportion of patients colonised with cefepime-resistant P. aeruginosa during mixing Inconclusively higher proportion of ceftazidime and carbapenem-resistant P. aeruginosa during mixing (p=. 0.06 and 0.07 respectively) No other significant differences with regard to other Gram-negatives species OTHER Higher mortality rates during cycling only for Unit 2 Similar mortality rates during cycling and mixing for Unit 1 Higher use of carbapenems and piperacillin-tazobactam and lower use of cephalosporins during mixing
Unit-wide surveillance cultures
Van Duijn PJ et al 2018
Cluster randomised cross-over trial
Multi-centre ICU
Cycling of 3GC (or 4GC), carbapenems and piperacillin-tazobactam every six weeks versus mixing those agents (administering those successively to consecutive patients) for empirical treatment of suspected Gram-negative infections De-escalation permitted Combination therapy permitted No typing to assess clonality of bacterial isolates PRIMARY Similar prevalence of antibiotic-resistant Gram-negative bacteria Similar incidence rate ratio of antibiotic-resistant Gram-negative bacteria adjusted for hand hygiene compliance, patient-sex and proportion of short-stay patients Similar prevalence of ESBLs, piperacillin-tazobactam- or carbapenem-resistant non-fermenters OTHER Similar mortality rates and similar length of stay during periods of mixing and cycling Similar overall use of antibiotics and similar use of study antibiotics between study periods Three times higher use of on-cycle antibiotics compared to off-cycle use
Unit-wide surveillance cultures
Jayashree M et al 2020
Comparative trial
Paediatric ICU
Period 1: Mixing piperacillin-tazobactam, imipenem and cefepime (administering those successively to consecutive patients) for suspected Gram-negative infections Period 2: Cycling the aforementioned agents every month De-escalation permitted Combination therapy permitted No typing to assess clonality of bacterial isolates PRIMARY Higher percentage of resistant isolates during baseline period than in mixing, cycling and washout periods Similar percentage of resistant isolates during mixing and cycling OTHER Similar mortality rates between periods Similar episodes of healthcare-associated infections during mixing and cycling but lower than baseline Similar overall use of antibiotics between all phases
Unit-wide surveillance cultures