Received: 2 March 2019 | Accepted: 19 March 2019

DOI: 10.1111/aas.13382

REVIEW

Piperacillin/tazobactam vs carbapenems for patients with

bacterial infection: Protocol for a systematic review

Marie Warrer Petersen1 | Anders Perner1 | Fredrik Sjövall2 |

Andreas Bender Jonsson1 | Morten Steensen1 | Jakob Steen Andersen1 |
Michael Patrick Achiam3 | Niels Frimodt‐Møller4 | Morten Hylander Møller1

1
Department of Intensive Care, Copenhagen
University Hospital Rigshospitalet, Introduction: Early empirical broad‐spectrum antimicrobial therapy is recommended
Copenhagen, Denmark for patients with severe infections, including sepsis. β‐lactam/β‐lactamase inhibitor
2
Department of Perioperative
combinations or carbapenems are often used to ensure coverage of likely pathogens.
Medicine, Skåne University Hospital, Malmö,
Sweden Piperacillin/tazobactam is proposed as a carbapenem‐sparing agent to reduce the
3
Department of Surgical incidence of multidrug‐resistant bacteria and superinfections. In the recently pub‐
Gastroenterology, Copenhagen University
Hospital Rigshospitalet, Copenhagen,
lished MERINO trial, increased mortality from piperacillin/tazobactam was suggested
Denmark in patients with bacteraemia with resistant Escherichia coli or Klebsiella species.
4
Department of Clinical Whether these findings also apply to empirical piperacillin/tazobactam in patients
Microbiology, Copenhagen University
Hospital Rigshospitalet, Copenhagen, with other severe infections, including sepsis, is unknown. We aim to assess the ben‐
Denmark efits and harms of empirical and definitive piperacillin/tazobactam vs carbapenems
Correspondence for patients with severe bacterial infections.
Marie Warrer Petersen, MD, Department of Methods and analysis: This protocol has been prepared according to the Preferred
Intensive Care 4131, Copenhagen University
Hospital Rigshospitalet, Inge Lehmanns Vej Reporting Items for Systematic Review and Meta‐Analysis Protocols statement, the
5, DK – 2100 Copenhagen. Cochrane Handbook and the Grading of Recommendations, Assessment,
Email: mariewpetersen@gmail.com
Development, and Evaluation approach. We will include randomised clinical trials as‐
Financial support
This research project was funded by the
sessing piperacillin/tazobactam vs carbapenems in patients with severe bacterial in‐
Ehrenreich's foundation. The foundation fections of any origin. The primary outcome will be all‐cause short‐term
was not involved in the development of the
protocol.
mortality ≤ 90 days. Secondary outcomes will include all‐cause long‐term mortal‐
ity > 90 days, adverse events, quality of life, use of life support, secondary infections,
antibiotic resistance, and length of stay. We will conduct meta‐analyses, including
pre‐planned subgroup and sensitivity analyses for all assessed outcomes. The risk of
random errors in the meta‐analyses will be assessed by trial sequential analysis.

1 | I NTRO D U C TI O N Early empirical broad‐spectrum antimicrobial therapy covering

Sepsis is a common and life‐threatening condition1; 30%‐40% of pa‐
tients in the intensive care unit (ICU) suffer from sepsis. 2,3
The World
Health Organization estimates that 6 million deaths annually are
caused by sepsis, making sepsis the third most common cause of death
worldwide. 4
PROSPERO registration number: CRD42019123052
likely pathogens is recommended for patients with sepsis, as both
delayed and inadequate empirical antimicrobial therapy is associated
with increased mortality and morbidity.1 The choice of antimicrobials
depends on the presumed primary source of infection.1 Most often,
broad‐spectrum therapy with either β‐lactam/β‐lactamase inhibitor
combinations (BL/BLIs) (eg, piperacillin/tazobactam, ticarcillin/clavu‐
lanate) or carbapenems (eg, meropenem, ertapenem, imipenem) is
recommended to ensure coverage of the most likely pathogens.1

Acta Anaesthesiol Scand. 2019;1–6. wileyonlinelibrary.com/journal/aas

© 2019 The Acta Anaesthesiologica Scandinavica | 1
Foundation. Published by John Wiley & Sons Ltd
2 |
Worldwide excessive use of broad‐spectrum antibiotics is asso‐
ciated with increased antimicrobial resistance.5 Particularly, the high
use of carbapenems is believed to propose a great threat to global
health by increasing the proportion of multidrug‐resistant bacteria.6-8
Piperacillin/tazobactam and other BLBLI are often proposed as carbap‐
enem‐sparing agents to limit the spread of multidrug‐resistant bacteria
and to reduce the incidence of bacterial and fungal superinfections.6,9,10
Several observational studies have compared the efficacy of
BLBLIs with carbapenems for severe bacterial infections, including
bacteraemia caused by extended‐spectrum β‐lactamase (ESBL)‐pro‐
ducing Enterobacteriaceae,11-22 pneumonia, 23 urinary tract infec‐
tions (UTIs), 21,24 intra‐abdominal infections (IAIs), 25-28 acute pelvic
infections (APIs), 25-27 and skin and skin‐structure infections (SSIs), 25-
27
and most have demonstrated equality in mortality and treatment
success between BLBLIs and carbapenems.11-13,17-23,25-27,29
BLBLIs have previously been compared to carbapenems in sys‐
tematic reviews with meta‐analyses in patients with bacteraemia with
ESBL‐producing Enterobacteriaceae,30-32 severe infections,33,34 and
complicated IAIs, APIs and SSIs.35 Most have suggested no difference
in mortality between groups for both empirical and definitive treat‐
ments, proposing BLBLIs as a safe alternative to carbapenems for var‐
30-33,35
ious severe infections. The effect of definitive treatment with
piperacillin/tazobactam vs meropenem in patients with bacteraemia
with ceftriaxone‐nonsusceptible Escherichia coli or Klebsiella species
(spp.) was assessed in a recent non‐inferiority randomised clinical trial
36
(RCT) by Harris et al (the MERINO trial). Disturbingly, the authors
were unable to demonstrate non‐inferiority, as the 30‐day mortality
was more than 3 times higher in patients allocated to piperacillin/
tazobactam (12.3%), as compared to those allocated to meropenem
(3.7%).36 Of note, all bacteria were susceptible in vitro to both pip‐
eracillin/tazobactam and meropenem.36 Whether these findings also
apply to empirical use of piperacillin/tazobactam in other severe bac‐
terial infections, including sepsis, is unknown.
1.1 | Objective
We aim to assess the benefits and harms of empirical and definitive
piperacillin/tazobactam vs carbapenems for patients with severe
bacterial infections of any origin. We hypothesise that empirical and
definitive piperacillin/tazobactam is inferior to carbapenems in se‐
vere bacterial infections.
2 | M E TH O DS
This protocol has been registered in the International Prospective
Register of Systematic Reviews (PROSPERO) (trial registration num‐
ber: CRD42019123052) on 21 February 2019. We prepared and re‐
ported this protocol according to the Preferred Reporting Items for
Systematic Review and Meta‐Analysis Protocols (PRISMA‐P) state‐
ment and recommendations from the Cochrane Collaboration.37,38
A PRISMA‐P checklist is provided in Appendix 1 in supplementary
materials.
PETERSEN et al.
2.1 | Types of studies
We will include RCTs assessing piperacillin/tazobactam vs carbapen‐
ems in patients with severe bacterial infections of any origin. We will
include RCTs regardless of publication status, language, and blinding.
Quasi‐randomised and cross‐over trials will not be included.
2.2 | Types of participants
We will assess RCTs conducted in adult patients (as defined in the
included trials) with suspected or confirmed severe bacterial infec‐
tions of any origin. A severe bacterial infection will be defined as any
infection requiring hospitalisation. We will exclude RCTs conducted
in animals or children.
2.3 | Types of interventions
We will include RCTs assessing empirical or definitive treatment with
piperacillin/tazobactam vs carbapenems (ie, meropenem, ertap‐
enem, imipenem, doripenem) in any timing, dosing, formulation, and
duration. Multiple intervention and control groups will be permit‐
ted. We will allow patients to be randomised to other antimicrobial
agents in both the intervention and control group, but any additional
antimicrobial agent must be identical in both arms. Empirical treat‐
ment will be defined as any antibiotic treatment commenced due to
clinical suspicion of infection before microbiological results are avail‐
able, whereas definitive treatment will be defined as any antibiotic
treatment commenced according to microbiological results.39
2.4 | Types of outcome measures
We will exclusively assess patient‐important outcomes.40 Each as‐
sessed outcome will be ranked as critical, important but not criti‐
cal, or of limited importance for decision‐making according to the
Grading of Recommendations, Assessment, Development and
Evaluation (GRADE) methodology.41
2.4.1 | Primary outcome measure
All‐cause short‐term mortality ≤ 90 days, including in‐ICU and in‐
hospital mortality (critical outcome).
2.4.2 | Secondary outcome measures
1. All‐cause long‐term mortality > 90 days (critical outcome)
2. Adverse events (as defined in the original trials) at the longest fol‐
low‐up (critical outcome)
3. Quality of life (as defined in the original trials) at the longest fol‐
low‐up (critical outcome)
4. Days alive without/duration of mechanical ventilation (important,
but not critical outcome)
5. Days alive without/duration of renal replacement therapy (impor‐
tant, but not critical outcome)
PETERSEN et al.
6. Days alive without/duration of vasopressors/inotropes (impor‐
tant, but not critical outcome)
7. Secondary infections (ie, new infectious origin or clinical deterio‐
ration of the original infection with a different organism com‐
pared to the original infection) at longest follow‐up (important,
but not critical outcome)
8. Selection of resistant bacteria (eg, enterococci, C difficile) or fungi
(Candida spp) at longest follow‐up (important, but not critical
outcome)
9. Days alive out of hospital/hospital length of stay (LOS) (if both are
available, we will report the outcome for which most data exist)
(important, but not critical outcome)
If outcomes are available at multiple time points, we will provide data
for the outcome at longest follow‐up. If both days alive without and
duration of life support (secondary outcome measures 4‐6) are avail‐
able, we will report the outcome for which the most data exist.
2.5 | Search methods for identification of studies
We will perform systematic searches in MEDLINE (PubMed inter‐
face, 1966 onwards), Embase (OVID interface, 1947 onwards), The
Cochrane Central Register of Controlled Trials (CENTRAL, Wiley in‐
terface, current issue) and Epistemonikos, and we will search online
trial registers for ongoing trials (clinicaltrials.gov, Controlled trials,
European Clinical Trial Database). Moreover, the reference lists of
relevant records will be reviewed to identify potentially eligible trials.
The tentative search strategy in MEDLINE is provided in Appendix 2
in supplementary materials.
2.6 | Data collection and analysis
2.6.1 | Selection of studies
A minimum of two authors will independently screen titles and ab‐
stracts of all identified records and review potentially eligible re‐
cords in full text. Disagreements will be resolved by discussion and
consensus among all authors.
2.6.2 | Data extraction and management
A minimum of two authors will independently extract data using a
standardised data extraction form in duplicate. Extracted data items
will include trial setting, risk of bias, population, intervention(s) (ie, an‐
timicrobial agents, dose, timing, formulation, duration), comparator(s)
(ie, antimicrobial agents, dose, timing, formulation, duration), and
data on the predefined outcome measures. Disagreements will be
resolved by discussion and consensus among all authors.
2.6.3 | Measures of treatment effect
For dichotomous outcomes, we will report relative risks (RRs) with
corresponding 95% confidence intervals (CIs) and for continuous
| 3
outcomes, we will report mean differences (MDs) with corresponding
standard deviations (SDs). If available, we will conduct intention‐to‐
treat rather than per‐protocol analysis. Moreover, we will use both
random and fixed effect models for our meta‐analyses and report the
analysis with the most conservative treatment effect estimate.42
2.6.4 | Assessment of risk of bias in included studies
A minimum of two authors will independently assess the risk of bias
of the included trials according to the recommendations from the
Cochrane Collaboration (Appendix 3 in supplementary materials).38
For each included trial, we will assess the following bias domains: (1)
random sequence generation, (2) allocation concealment, (3) blind‐
ing of participants and personnel, (4) blinding of outcome assessors,
(5) incomplete outcome data, (6) selective outcome reporting, and
(7) other sources of bias, including financial bias, baseline imbalance,
and early stopping.38,43-45 Disagreements will be resolved by discus‐
sion and consensus among all authors.
Trials will be adjudicated as having an overall low risk of bias if all
bias domains are regarded as having a low risk of bias. If 1 or more
bias domains are regarded as having an unclear or high risk of bias,
the trial will be adjudicated as having an overall high risk of bias.
The complete risk of bias assessment for all trials will be sum‐
marised in a figure. Moreover, we will provide a supplementary table
with support for the judgement of each bias domain for each individ‐
ual trial (Appendix 4 in supplementary materials).
2.6.5 | Assessment of risk of random errors
We will conduct trial sequential analysis (TSA) to assess the risk of
random errors and spurious findings due to repetitive testing in the
cumulative meta‐analyses.46 TSA widens the CI and quantifies the
strength of evidence by calculating the information size required for
a conclusive meta‐analysis.46 For the analysis, we will apply trial se‐
quential monitoring boundaries according to an a priori 15% relative
risk reduction, alpha 5%, beta 10%, and a control event proportion
as per the control arm of the included trials.
2.6.6 | Dealing with missing data
We will contact trial authors for supplementary data in case data
are missing. All missing data will be reported. We will conduct a sen‐
sitivity analysis with (1) best‐worst and (2) worst‐best scenarios if
patients are lost to follow up 47:
1. Best‐worst scenario: all patients lost to follow up in the intervention
group have not had the outcome of interest, and all patients lost to
follow up in the control group have had the outcome of
interest.47
2. Worst‐best scenario: all patients lost to follow up in the interven‐
tion group have had the outcome of interest, and all patients lost to
follow up in the control group have not had the outcome of
interest.47
4 |
2.6.7 | Assessment of heterogeneity
We will calculate inconsistency (I2) and diversity factor (D2) to assess
heterogeneity across trials.48
2.6.8 | Assessment of small trial bias
We will test for funnel plot symmetry to assess the risk of small
trial bias (reporting bias) if 10 or more trials are included in the
49
meta‐analyses.
2.6.9 | Data synthesis
We will use Review Manager (RevMan version 5.3) to conduct
both meta‐analyses and the predefined subgroup analyses. For
the TSA, we will use the TSA software (version 0.9.5.10 beta)
from Copenhagen Trial Unit (available from www.ctu.dk). For the
primary outcome measures, a p‐value of 0.05 will be considered
statistically significant. For the secondary outcome measures, we
will apply multiplicity adjustment corresponding to the number of
secondary outcomes assessed in the cumulative meta‐analyses to
42
control the family‐wise error rate. Correspondingly, a p‐value
of 0.01 will be considered statistically significant if data for all 9
secondary outcome measures are available. If data are unavailable
for 1 or more outcome measures, we will calculate the p‐value by
dividing 0.05 with the value halfway between 1 and the number of
outcome measures (1 secondary outcome: P = 0.05; 2 secondary
outcomes: P = 0.033; 3 secondary outcomes: P = 0.25 etc).42
2.6.10 | Subgroup analysis
We will conduct the following predefined subgroup analyses, if data
for both subgroups are available50,51:
1. Trials with an overall high vs low risk of bias. We hypothesise
increased harm of piperacillin/tazobactam compared to car‐
bapenem in trials with an overall low risk of bias.
2. Trials conducted in ICU vs non‐ICU patients. We hypothesise in‐
creased harm from piperacillin/tazobactam compared to carbap‐
enem in trials conducted in ICU patients.
3. Trials in patients with vs without immunosuppression (eg, hema‐
tological malignancy, human immunodeficiency virus). We hy‐
pothesise increased harm from piperacillin/tazobactam compared
to carbapenem in patients with immunosuppression.
4. Trials conducted in patients with different primary site of infec‐
tions (ie, central nervous system vs respiratory tract vs gastroin‐
testinal tract vs urinary tract vs gynaecological vs bone and joints
vs skin and skin‐structure). We hypothesise a difference in the
balance between benefits and harms of piperacillin/tazobactam
vs carbapenem in patients with different primary site of
infections.
5. Trials assessing empirical vs definitive treatment: We hypothesise
a difference in the balance between benefits and harms of
PETERSEN et al.
piperacillin/tazobactam vs carbapenem in trials assessing empiri‐
cal and definitive treatment.
6. Trials assessing different types of carbapenems: We hypothesise
a difference in the balance between benefits and harms of pipera‐
cillin/tazobactam vs carbapenem in patients receiving different
carbapenems.
We will use the chi‐squared test to assess the statistical heterogene‐
ity across patient subgroups (test of interaction). A p‐value of 0.10
will be considered statistically significant.
2.6.11 | Sensitivity analysis
We will apply empirical continuity correction in the zero event trials.52
2.6.12 | Assessment of the overall
quality of evidence
The quality of evidence for each outcome measures will be assessed
according to the GRADE recommendations.41 We will downgrade
the quality of evidence for risk of bias, inconsistency, indirectness,
imprecision, and publication bias to adjudicate the overall quality of
evidence as very low, low, moderate, or high.41
3 | D I S CU S S I O N
This protocol has been planned and reported according to the
PRISMA‐P statement.37 We will conduct the review according to
the PRISMA statement and recommendations from the Cochrane
Collaboration and GRADE methodology.38,41,53 Other strengths
include the assessment of patient‐important outcomes, the prede‐
fined subgroup and sensitivity analyses, and the use of TSA to assess
the risk of random errors.46
The planned review holds limitations too. We expect some
clinical heterogeneity among the included trials regarding setting,
population, intervention(s), and comparator(s). To assess this clinical
heterogeneity, we plan to perform pre‐planned subgroup analyses
of trials assessing different primary infections, empirical vs defini‐
tive treatment and different types of carbapenems. For the TSA, we
have defined an a priori relative risk reduction of 15% as clinically
relevant with a control event proportion as per the control arm of
the included trials. Consequently, our findings will not be able to rule
out a smaller treatment effect (<15%).
4 | CO N C LU S I O N S
Broad‐spectrum therapy with piperacillin/tazobactam or carbapen‐
ems is commonly used in the empirical or definitive treatment of se‐
vere bacterial infections, including sepsis.1 In the recently published
MERINO trial, harm from piperacillin/tazobactam was suggested in
patients with bacteraemia with ESBL‐producing E. coli or Klebsiella
PETERSEN et al.
spp.36 This outlined review will provide an important update on the
balance between the benefits and harms of empirical and definitive
piperacillin/tazobactam vs carbapenems in patients with severe bac‐
terial infections of any origin.
ETHICS
Ethical approval will not be required as only published data will be used.
AC K N OW L E D EG M E N T S
This research project was funded by the Ehrenreich's foundation.
C O N FL I C T S O F I N T E R E S T
The Department of Intensive Care 4131 at Copenhagen University
Hospital Rigshospitalet has received support for research from
Ferring Pharmaceutical, Fresenius Kabi, and CSL Behring.
AU T H O R C O N T R I B U T I O N S
MWP and MHM drafted the protocol. All authors made critical revi‐
sions for the manuscript. MHM developed the idea for the review
and is the guarantor of the review.
ORCID
Marie Warrer Petersen https://orcid.org/0000-0003-1127-9599
Anders Perner https://orcid.org/0000-0002-4668-0123
Andreas Bender Jonsson https://orcid.org/0000-0003-4611-5543
Morten Hylander Møller https://orcid.org/0000-0002-6378-9673
REFERENCES
1. Rhodes A, Evans LE, Alhazzani W, et al. Surviving Sepsis Campaign:
International Guidelines for Management of Sepsis and Septic
Shock: 2016. Crit Care Med. 2017;45:486‐552.
2. Vincent JL, Sakr Y, Sprung CL, et al. Sepsis occurrence in acutely
Ill patients I. Sepsis in European intensive care units: results of the
SOAP study. Crit Care Med. 2006;34:344‐353.
3. Vincent J‐L, Marshall JC, Ñamendys‐Silva SA, et al. Assessment of
the worldwide burden of critical illness: the intensive care over na‐
tions (ICON) audit. Lancet. Respir Med. 2014;2:380‐386.
4. Seventieth World Health Assembly A70/13. Improving the preven‐
tion, diagnosis and clinical management of sepsis. http://apps.who.
int/gb/ebwha/​pdf_files/​WHA70/​A70_13-en.pdf?ua=1. Provisional
agenda item 12.2 edn, 2017.
5. Luyt CE, Brechot N, Trouillet JL, Chastre J. Antibiotic stewardship in
the intensive care unit. Crit Care. 2014;18:480.
6. McLaughlin M, Advincula MR, Malczynski M, Qi C, Bolon M,
Scheetz MH. Correlations of antibiotic use and carbapenem re‐
sistance in Enterobacteriaceae. Antimicrob Agents Chemother.
2013;57:5131‐5133.
7. Chang HJ, Hsu PC, Yang CC, et al. Risk factors and outcomes of
carbapenem‐nonsusceptible Escherichia coli bacteremia: a matched
case‐control study. J Microbiol Immunol Infect. 2011;44:125‐130.
| 5
8. Meyer E, Schwab F, Schroeren‐Boersch B, Gastmeier P. Dramatic
increase of third‐generation cephalosporin‐resistant E. coli in
German intensive care units: secular trends in antibiotic drug use
and bacterial resistance, 2001 to 2008. Crit Care. 2010;14:R113.
9. Eljaaly K, Enani MA, Al‐Tawfiq JA. Impact of carbapenem ver‐
sus non‐carbapenem treatment on the rates of superinfection: a
meta‐analysis of randomized controlled trials. J Infect Chemother.
2018;24:915‐920.
10. Harris PN, Tambyah PA, Paterson DL. beta‐lactam and beta‐lact‐
amase inhibitor combinations in the treatment of extended‐spec‐
trum beta‐lactamase producing Enterobacteriaceae: time for a
reappraisal in the era of few antibiotic options? Lancet Infect Dis.
2015;15:475‐485.
11. Ko JH, Lee NR, Joo EJ, et al. Appropriate non‐carbapenems are
not inferior to carbapenems as initial empirical therapy for bacte‐
remia caused by extended‐spectrum beta‐lactamase‐producing
Enterobacteriaceae: a propensity score weighted multicenter co‐
hort study. Eur J Clin Microbiol Infect Dis. 2018;37:305‐311.
12. Gudiol C, Royo‐Cebrecos C, Abdala E, et al. Efficacy of beta‐lac‐
tam/beta‐lactamase inhibitor combinations for the treatment of
bloodstream infection due to extended‐spectrum‐beta‐lactamase‐
producing Enterobacteriaceae in hematological patients with neu‐
tropenia. Antimicrob Agents Chemother. 2017;61;e00164‐17.
13. Gutierrez‐Gutierrez B, Perez‐Galera S, Salamanca E, et al. A multina‐
tional, preregistered cohort study of beta‐lactam/beta‐lactamase in‐
hibitor combinations for treatment of bloodstream infections due to
extended‐spectrum‐beta‐lactamase‐producing Enterobacteriaceae.
Antimicrob Agents Chemother. 2016;60:4159‐4169.
14. Ng TM, Khong WX, Harris PN, et al. Empiric piperacillin‐tazobac‐
tam versus carbapenems in the treatment of bacteraemia due to ex‐
tended‐spectrum beta‐lactamase‐producing Enterobacteriaceae.
PLoS ONE. 2016;11:e0153696.
15. Ofer‐Friedman H, Shefler C, Sharma S, et al. Carbapenems versus
piperacillin‐tazobactam for bloodstream infections of nonurinary
source caused by extended‐spectrum beta‐lactamase‐producing
Enterobacteriaceae. Infect Control Hosp Epidemiol. 2015;36:981‐985.
16. Tamma PD, Han JH, Rock C, et al. Carbapenem therapy is associ‐
ated with improved survival compared with piperacillin‐tazobactam
for patients with extended‐spectrum beta‐lactamase bacteremia.
Clin Infect Dis. 2015;60:1319‐1325.
17. Harris PN, Yin M, Jureen R, et al. Comparable outcomes for β‐
lactam/β‐lactamase inhibitor combinations and carbapenems in
definitive treatment of bloodstream infections caused by cefotax‐
ime‐resistant Escherichia coli or Klebsiella pneumoniae. Antimicrob
Resist Infect Control. 2015;4:14.
18. Tsai HY, Chen YH, Tang HJ, et al. Carbapenems and piperacillin/
tazobactam for the treatment of bacteremia caused by extended‐
spectrum beta‐lactamase‐producing Proteus mirabilis. Diagn
Microbiol Infect Dis. 2014;80:222‐226.
19. Rodriguez‐Bano J, Navarro MD, Retamar P, Picon E, Pascual A.
beta‐Lactam/beta‐lactam inhibitor combinations for the treatment
of bacteremia due to extended‐spectrum beta‐lactamase‐produc‐
ing Escherichia coli: a post hoc analysis of prospective cohorts. Clin
Infect Dis. 2012;54:167‐174.
20. Kang CI, Park SY, Chung DR, Peck KR, Song JH. Piperacillin‐ta‐
zobactam as an initial empirical therapy of bacteremia caused by
extended‐spectrum beta‐lactamase‐producing Escherichia coli and
Klebsiella pneumoniae. J Infect. 2012;64:533‐534.
21. Yoon YK, Kim JH, Sohn JW, Yang KS, Kim MJ. Role of piperacillin/
tazobactam as a carbapenem‐sparing antibiotic for treatment of
acute pyelonephritis due to extended‐spectrum beta‐lactamase‐
producing Escherichia coli. Int J Antimicrob Agents. 2017;49:410‐415.
22. Tang CQ, Li JQ, Xia ZF, et al. Empirical therapy of bloodstream in‐
fections caused by extended‐spectrum beta‐lactamase‐producing
Enterobacteriaceae. Zhonghua Yi Xue Za Zhi. 2016;96:2076‐2080.
6 |
23. Karino F, Miura K, Fuchita H, et al. Efficacy and safety of pipera‐
cillin/tazobactam versus biapenem in late elderly patients with
nursing‐ and healthcare‐associated pneumonia. J Infect Chemother.
2013;19:909‐915.
24. Dizbay M, Ozger HS, Karasahin O, Karasahin EF. Treatment efficacy
and superinfection rates in complicated urinarytract infections
treated with ertapenem or piperacillin tazobactam. Turk J Med Sci.
2016;46:1760‐1764.
25. Solomkin J, Teppler H, Graham DR, et al. Treatment of polymicro‐
bial infections: post hoc analysis of three trials comparing ertape‐
nem and piperacillin‐tazobactam. J Antimicrob Chemother. 2004;53
Suppl 2:ii51‐7.
26. Tellado J, Woods GL, Gesser R, McCarroll K, Teppler H. Ertapenem
versus piperacillin‐tazobactam for treatment of mixed anaerobic
complicated intra‐abdominal, complicated skin and skin structure,
and acute pelvic infections. Surg Infect (Larchmt). 2002;3:303‐314.
27. Teppler H, McCarroll K, Gesser RM, Woods GL. Surgical infections
with enterococcus: outcome in patients treated with ertapenem ver‐
sus piperacillin‐tazobactam. Surg Infect (Larchmt). 2002;3:337‐349.
28. Niinikoski J, Havia T, Alhava E, et al. Piperacillin/tazobactam versus
imipenem/cilastatin in the treatment of intra‐abdominal infections.
Surg Gynecol Obstet. 1993;176:255‐261.
29. Tang C, Li J, Xia Z, et al. Beta‐lactam/beta‐lactamase inhibitors versus
carbapenems for the treatment of urinary tract infections caused by
extended‐spectrum beta‐lactamase‐producing bacteria: A retrospec‐
tive cohort study. ACCP 2013 ANNUAL MEETING Albuquerque, New
Mexico: Pharmacotherapy Publications Inc., 2013:e213‐e14.
30. Vardakas KZ, Tansarli GS, Rafailidis PI, Falagas ME. Carbapenems
versus alternative antibiotics for the treatment of bacteraemia
due to Enterobacteriaceae producing extended‐spectrum beta‐
lactamases: a systematic review and meta‐analysis. J Antimicrob
Chemother. 2012;67:2793‐2803.
31. Harris PN, Wei JY, Shen AW, et al. Carbapenems versus alternative
antibiotics for the treatment of bloodstream infections caused by
Enterobacter, Citrobacter or Serratia species: a systematic review
with meta‐analysis. J Antimicrob Chemother. 2016;71:296‐306.
32. Sfeir MM, Askin G, Christos P. Beta‐lactam/beta‐lactamase inhib‐
itors versus carbapenem for bloodstream infections due to ex‐
tended‐spectrum beta‐lactamase‐producing Enterobacteriaceae:
systematic review and meta‐analysis. Int J Antimicrob Agents.
2018;52:554‐570.
33. Shiber S, Yahav D, Avni T, Leibovici L, Paul M. beta‐Lactam/beta‐
lactamase inhibitors versus carbapenems for the treatment of sep‐
sis: systematic review and meta‐analysis of randomized controlled
trials. J Antimicrob Chemother. 2015;70:41‐47.
34. Edwards SJ, Clarke MJ, Wordsworth S, Emmas CE. Carbapenems
versus other beta‐lactams in treating severe infections in intensive
care: a systematic review of randomised controlled trials. Eur J Clin
Microbiol Infect Dis. 2008;27:531‐543.
35. An MM, Zou Z, Shen H, et al. Ertapenem versus piperacillin/tazo‐
bactam for the treatment of complicated infections: a meta‐analysis
of randomized controlled trials. BMC Infect Dis. 2009;9:193.
36. Harris P, Tambyah PA, Lye DC, et al. Effect of piperacillin‐tazobac‐
tam vs meropenem on 30‐day mortality for patients with E. coli or
Klebsiella pneumoniae bloodstream infection and ceftriaxone resis‐
tance: a randomized clinical trial. JAMA. 2018;320:984‐994.
37. Moher D, Shamseer L, Clarke M, et al. Preferred reporting items for
systematic review and meta‐analysis protocols (PRISMA‐P) 2015
statement. Syst Rev. 2015;4:1.
38. Higgins J, Green S. Cochrane Handbook for Systematic Reviews of
Interventions Version 5.1.0 [updated March 2011]. The Cochrane
Collaboration. www.cochr​ane-handb​ook.org. 2011.
PETERSEN et al.
39. Leekha S, Terrell CL, Edson RS. General principles of antimicrobial
therapy. Mayo Clin Proc. 2011;86:156‐167.
40. Ciani O, Buyse M, Garside R, et al. Comparison of treatment effect
sizes associated with surrogate and final patient relevant outcomes
in randomised controlled trials: meta‐epidemiological study. BMJ.
2013;346:f457.
41. Atkins D, Best D, Briss PA, et al. Grading quality of evidence and
strength of recommendations. BMJ. 2004;328:1490.
42. Jakobsen JC, Wetterslev J, Winkel P, Lange T, Gluud C. Thresholds
for statistical and clinical significance in systematic reviews with
meta‐analytic methods. BMC Med Res Methodol. 2014;14:120.
43. Lundh A, Sismondo S, Lexchin J, Busuioc OA, Bero L. Industry
sponsorship and research outcome. Cochrane Database Syst Rev.
2012;12:MR000033.
44. Corbett MS, Higgins JP, Woolacott NF. Assessing baseline imbal‐
ance in randomised trials: implications for the Cochrane risk of bias
tool. Res Synth Methods. 2014;5:79‐85.
45. Tyson JE, Pedroza C, Wallace D, D'Angio C, Bell EF, Das A. Stopping
guidelines for an effectiveness trial: what should the protocol spec‐
ify? Trials. 2016;17:240.
46. Wetterslev J, Thorlund K, Brok J, Gluud C. Trial sequential analysis
may establish when firm evidence is reached in cumulative meta‐
analysis. J Clin Epidemiol. 2008;61:64‐75.
47. Akl EA, Kahale LA, Agoritsas T, et al. Handling trial participants with
missing outcome data when conducting a meta‐analysis: a system‐
atic survey of proposed approaches. Syst Rev. 2015;4:98.
48. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta‐
analysis. Stat Med. 2002;21:1539‐1558.
49. Sterne JA, Sutton AJ, Ioannidis JP, et al. Recommendations for ex‐
amining and interpreting funnel plot asymmetry in meta‐analyses of
randomised controlled trials. BMJ. 2011;343:d4002.
50. Sun X, Briel M, Busse JW, et al. Credibility of claims of subgroup
effects in randomised controlled trials: systematic review. BMJ.
2012;344:e1553.
51. European Medicines Agency. Guideline on the investigation of sub‐
groups in 4 confirmatory clinical trials. https​://www.ema.europa.
eu/docum​e nts/scien​t ific-guide​line/draft-guide​line-inves​t igat​ion-
subgr​oups-confi​rmato​r y-clini​c al-trials_en.pdf. 2014.
52. Sweeting MJ, Sutton AJ, Lambert PC. What to add to nothing? Use
and avoidance of continuity corrections in meta‐analysis of sparse
data. Stat Med. 2004;23:1351‐1375.
53. Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred
reporting items for systematic reviews and meta‐analyses: the
PRISMA statement. Int J Surg. 2010;8:336‐341.
SUPPORTING INFORMATION
Additional supporting information may be found online in the
Supporting Information section at the end of the article.
How to cite this article: Petersen MW, Perner A, Sjövall F, et
al. Piperacillin/tazobactam vs carbapenems for patients with
bacterial infection: Protocol for a systematic review. Acta
Anaesthesiol Scand. 2019;00:1–6. https​://doi.org/10.1111/
aas.13382​