lable at ScienceDirect

Clinical Microbiology and Infection xxx (2016) 1e8

Contents lists avai

Clinical Microbiology and Infection

journal homepage: www.cl inicalmicrobiologyandinfect ion.com

Review

Antibiotic de-escalation for bloodstream infections and pneumonia:systematic review and meta-analysis

M. Paul 1, *, 3, Y. Dickstein 1, 3, A. Raz-Pasteur 1, 2

1) Infectious Diseases Institute, Rambam Health Care Campus and The Ruth and Bruce Rappaport Faculty of Medicine, Technion, Israel Institute ofTechnology, Haifa, Israel2) Medicine A, Rambam Health Care Campus, Haifa, Israel

a r t i c l e i n f o

Article history:Received 7 March 2016Received in revised form23 May 2016Accepted 24 May 2016Available online xxx

Editor: A. Huttner

Keywords:Antibiotic treatmentBiasDe-escalationEmpirical antibiotic treatmentMeta-analysisObservational studyRandomized controlled trialSusceptibility testingSystematic review

* Corresponding author. M. Paul, Director, InfectiouHealth Care Campus, Haifa, 3109601, Israel.

E-mail address: paulm@post.tau.ac.il (M. Paul).3 Authors M. Paul and Y. Dickstein contributed equ

http://dx.doi.org/10.1016/j.cmi.2016.05.0231198-743X/© 2016 European Society of Clinical Micro

Please cite this article in press as: Paul M, emeta-analysis, Clinical Microbiology and Inf

a b s t r a c t

Antibiotic de-escalation is an appealing strategy in antibiotic stewardship programmes. We aimed toassess its safety and effects using a systematic review and meta-analysis. We included randomizedcontrolled trials (RCTs) and observational studies assessing adults with bacteraemia, microbiologicallydocumented pneumonia or severe sepsis, comparing between antibiotic de-escalation and no de-escalation. De-escalation was defined as changing an initially covering antibiotic regimen to a nar-rower spectrum regimen based on antibiotic susceptibility testing results within 96 hours. The primaryoutcome was 30-day all-cause mortality. A search of published articles and conference proceedings waslast updated in September 2015. Crude and adjusted ORs with 95% CI were pooled in random-effectsmeta-analyses. Sixteen observational studies and three RCTs were included. Risk of bias related toconfounding was high in the observational studies. De-escalation was associated with fewer deaths inthe unadjusted analysis (OR 0.53, 95% CI 0.39e0.73), 19 studies, moderate heterogeneity. In the adjustedanalysis there was no significant difference in mortality (adjusted OR 0.83, 95% CI 0.59e1.16), 11 studies,moderate heterogeneity and the RCTs showed non-significant increased mortality with de-escalation (OR1.73, 95% 0.97e3.06), three trials, no heterogeneity. There was a significant unadjusted association be-tween de-escalation and survival in bacteraemia/severe sepsis (OR 0.45, 95% CI 0.30e0.67) andventilator-associated pneumonia (OR 0.49, 95% CI 0.26e0.95), but not with other pneumonia (OR 0.97,95% CI 0.45e2.12). Only two studies reported on the emergence of resistance with inconsistent findings.Observational studies suggest lower mortality with antibiotic susceptibility testing-based de-escalationfor bacteraemia, severe sepsis and ventilator-associated pneumonia that was not demonstrated in RCTs.M. Paul, CMI 2016;▪:1© 2016 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier Ltd. All

rights reserved.

Background

Bloodstream infections and pneumonia requiring hospitaliza-tion are responsible for significant morbidity and mortality, withmortality rates ranging between 27% and 54% [1,2]. Numerousstudies have pointed to the importance of appropriate empiricantibiotics in reducingmortality for severe infections [3,4]. This hadled to the widespread use of broad-spectrum drugs as first-linetreatment, potentially contributing to the increase in bacterialresistance to antibiotics since, without intervening, empirical

s Diseases Institute, Rambam

ally to the study.

biology and Infectious Diseases. Pu

t al., Antibiotic de-escalationection (2016), http://dx.doi.o

therapy is frequently continued. Several strategies have evolved tolimit the appearance and spread of such organisms, among themantibiotic de-escalation [5,6].

De-escalation (also termed streamlining) refers to tailoring ofempirical antibiotic treatment to the susceptibilities of the bacteriaisolated, selecting the narrowest spectrum antibiotic. It can followany empirical treatment, but is also applied with a policy of initialbroad-spectrum treatment mainly in intensive-care units (ICUs). De-escalation might be easier to implement in antibiotic stewardshipprogrammes than interventions targeting empirical antibiotics [5,7].More information is available at the latter time-point, the patient'scourse is known and there is time for discussion and consideration.Promoting de-escalation entails increasing awareness among allantibiotic prescribers and education regarding antibiotic hierarchyor local preferences for targeted antibiotic treatment.

blished by Elsevier Ltd. All rights reserved.

for bloodstream infections and pneumonia: systematic review andrg/10.1016/j.cmi.2016.05.023

M. Paul et al. / Clinical Microbiology and Infection xxx (2016) 1e82

A Cochrane review published in 2013 found insufficient evi-dence to recommend for or against de-escalation in adults withsepsis after a review of the literature failed to reveal randomizedcontrolled trials (RCTs) testing the intervention [8]. RCTs constitutethe reference standard design to assess such an intervention.However, as de-escalation has been appraised in many observa-tional studies, systematically reviewing them and appraising therisk of bias in observational studies might prove useful to guidepractice and further research. We aimed to assess the outcome ofde-escalation therapy in patients with bloodstream infections, se-vere sepsis and pneumonia.

Materials and methods

We included RCTs and prospective or retrospective observa-tional studies, conducted in non-ICU and ICU settings. Patients,interventions, comparisons and outcomes are summarized in theSupplementary material (Table S1). We included adults 18 years ofage and older with pneumonia, bacteraemia and severe sepsis/septic shock with microbiologically documented infections, whoreceived appropriate empirical antibiotic treatment. Bloodstreaminfections had to be defined as clinically significant using validdefinitions to exclude contaminants and pneumonia had to bedefined using valid clinical and microbiological definitions [9].

The studies had to compare de-escalation therapy versuscontinued empiric antibiotic therapy. De-escalation was defined aschanging an initially appropriate (covering) antimicrobial therapyto a narrower spectrum regimen based on culture results within 96hours. A narrower spectrum regimen was defined as downgradingfrom a broad spectrum to a narrower spectrum agent within thesame antibiotic class, changing a broad-spectrum antibiotic to anarrower-spectrum antibiotic of a different class (e.g. vancomycinto oxacillin), or discontinuation of one or more drugs of a combinedregimen. Downgrading antibiotics from a broad to a narrow spec-trum necessitates a hierarchy of antibiotics; we documentedwhether a hierarchy was used and accepted the study definitionsfor de-escalation and antibiotic hierarchy, as long as defined byantibiotic susceptibility testing and compatible with our defini-tions. We separated between studies in which empiric antibiotictreatment was intentionally broad-spectrum and those that did notspecifically direct the empirical regimen.

The primary outcome assessed was all-cause mortality at 30days. If not reported, we used all-causemortality at the end of studyfollow up. Secondary outcomes included clinical failure, as definedin the study, examined at the end of treatment; duration of hospitaland ICU stay; duration of antibiotic treatment; resistance devel-opment and superinfections, defined as secondary clinically sig-nificant infections developing within a 30-day follow up. Antibioticresistance development was assessed as isolation of bacteriaresistant to the antibiotics given to the patient in clinical or sur-veillance samples; and as isolation of MDR bacteria of epidemio-logical significance that were not present initially, including:extended-spectrum b-lactamase-producing bacteria, methicillin-resistant Staphylococcus aureus, vancomycin-resistant Enter-obacteriaceae and carbapenem-resistant Enterobacteriaceae. Weassessed adverse effects of antibiotic therapy, including nausea orvomiting, rash, antibiotic-associated diarrhoea, renal failure andhepatotoxicity and Clostridium-difficile infection.

We conducted a broad search for randomized and observationalstudies in PubMed, The Cochrane Library from inception untilSeptember 2015, and conference proceedings for the last 3 years ofthe ECCMID and ICAAC. In addition, we examined the bibliogra-phies of identified trials as well as previous systematic reviews. Norestrictions on language, date of publication or publication statuswere applied. We tailored the following search string by database:

Please cite this article in press as: Paul M, et al., Antibiotic de-escalationmeta-analysis, Clinical Microbiology and Infection (2016), http://dx.doi.o

(blood stream infection OR bloodstream infection OR bacteremiaOR sepsis OR septic shock OR pneumonia) AND (de-escalation ORde-escalate OR streamlining OR streamline OR targeted OR tar-geting OR narrowing OR narrow) AND (antibiotic OR antibiotics).

We applied the Cochrane Risk of Bias Assessment Tool for Non-Randomized Studies of Interventions (ACROBAT-NRSI, http://bmg.cochrane.org/cochrane-risk-bias-assessment-tool-non-randomized-studies-interventions-acrobat-nrsi) to both observa-tional studies and RCTs, addressing the outcome of all-cause mor-tality. The tool includes signalling questions in seven domains ofbias, to which the responses are yes, probably yes, probably no andno. RCTs could achieve low risk of bias for all domains, dependingon randomization methods. Low risk of bias in an observationalstudy implies that it is comparable to a well-performed RCT. Wetailored the tool to our review (see Supplementary material,Appendix S1).

Two reviewers independently applied inclusion/ exclusioncriteria and extracted all data. Data were compared and differenceswere resolved by discussion. We extracted crude mortality ratesand adjusted effect estimates in observational studies. For contin-uous outcomes we computed means and standard deviations fromthe data reported in the study using methods specified in theCochrane Handbook and Wan et al. [10,11].

Adjusted risk ratios were converted to odds ratios and thehazard ratio reported in one trial was assumed to represent the riskratio following the proportional hazards assumption. We compiledcrude ORs or absolute mean differences using a random effectsmeta-analysis and adjusted ORs (or ORs computed from RCTs) us-ing an inverse variable random effects meta-analysis. Heteroge-neity was assessed using a chi-square test (p <0.1) and the I2 test(>50%). Analyses were conducted in REVIEW MANAGER 5.3 [12].

Results

Our literature review identified 558 potential articles for eval-uation. Of these, 19 fulfilled review eligibility criteria and wereincluded in the analysis [13e31] (Fig. 1). One study contributed totwo analyses of different, non-overlapping periods in the study[24]. The studies comprised a total of 3973 patients, all adults, witha study mean or median age between 51 and 71 years. Eight studiesenrolled patients with bacteria, two addressed severe sepsis orseptic shock in the ICU and 14 studies enrolled patients withpneumonia (Table 1). Four studies included community-acquiredinfections only [14,17,25,29], eight included exclusively hospital-acquired infections [13,19,22,23,27,28,30,31] and six studiesincluded both [15,18,20,21,24,26]. De-escalation was performedaccording to MDI in all of the included studies and was definedvariably (see Supplementary material, Table S2): ten defined de-escalation by narrowing the spectrum and 13 by stopping one ormore drugs of a combination. In five studies on hospital-acquiredpneumonia de-escalation was performed from a broad-spectrumempirical regimen [13,19,22,30,31]. All-cause mortality was re-ported in all studies. The majority (11) reported 28- to 30-day all-cause mortality, whereas the others reported on in-hospital, in-ICU, 90 days [26] or in relation to completion of antibiotic treatment[13].

Risk of bias assessment

Three trials were randomized [17,22,26], whereas the remainderwere observational studies. The RCTs did not score low risk of biasfor all items, one trial failing to match groups for important con-founders [26] and two open-label trials did not report on co-interventions [17,26] (Fig. 2). Overall, 11/19 studies reported onpre-defined important confounders (age, renal function baseline

for bloodstream infections and pneumonia: systematic review andrg/10.1016/j.cmi.2016.05.023

Records iden fied through database searching (n = 525)

Addi onal records iden fied through other sources (n = 33)

Records a er duplicates removed(n = 549)

Records screened(n = 549)

Records excluded(n = 519)

Full-text ar cles assessed for eligibility (n = 30)

Full-text ar cles excluded(n = 11)

Not all pa ents with microbiological documenta on/ de-escala on not based on AST [1–6] No comparator group [7,8] Review [9,10] No clinical data on de-escala on [11]

Studies included in qualita ve synthesis

(n = 19)

Studies included in quan ta ve synthesis

(n = 19)

Fig. 1. Study flow; references refer to supplementary Appendix S2.

M. Paul et al. / Clinical Microbiology and Infection xxx (2016) 1e8 3

and acute, haemodynamic status, respiratory status, functional sta-tus, cardiac disease, see Supplementary material, Appendix S1) andperformed an adjusted analysis. The mean severity-of-illness scores(e.g. APACHE II) were well-matched between the study groups in 13of the studies [13e15,18e20,22,25,27,29,30]. However, items scoringat high risk in most studies included confounder assessment,matching for confounders and co-interventions (Fig. 2).

Primary outcome: mortality

Overall, the pooled OR for mortality with de-escalation of allstudies using crude, unadjusted, results from the observationalstudies was 0.53 (95% CI 0.39e0.73), OR <1 favouring de-escalation(Fig. 3). The difference was statistically significant in studies ofbacteraemic patients (OR 0.45, 95% CI 0.30e0.67, ten studies) andnon-significant for pneumonia (OR 0.64, 95% CI 0.39e1.06, ninestudies). Both subgroups and the overall analysis had moderateheterogeneity (I2 47e69%). Heterogeneity in the pneumonia anal-ysis could be partially explained by the type of pneumonia: forventilator-associated pneumonia there was a significant advantageto de-escalation (OR 0.49, 95% CI 0.26e0.95), whereas in studiesincluding mostly non-ventilated pneumonia there was no signifi-cant difference between groups (OR 0.97, 95% CI 0.45e2.12; seeSupplementary material, Fig. S1), with no differences observedwhen empirical treatment was broad-spectrum or non-directed.However, heterogeneity remained also in the stratified pneu-monia analysis. ORs were similar for studies in which the mortalityrate was above the median (mean 32%) and in those with lowermortality (mean 16%).

Please cite this article in press as: Paul M, et al., Antibiotic de-escalationmeta-analysis, Clinical Microbiology and Infection (2016), http://dx.doi.o

The adjusted analysis showed a smaller association between de-escalation and survival that was no longer statistical significant(adjusted OR 0.83, 95% CI 0.59e1.16) and in the subgroup of fourstudies assessing patients with pneumonia there was no difference(adjusted OR 1.17, 95% CI 0.73e1.86), without significant hetero-geneity (Fig. 4).

Pooling of the three RCTs, assessing different types of infec-tion (community-acquired pneumonia, hospital-acquired pneu-monia and severe sepsis in ICU) and different de-escalationdefinitions showed increased mortality with de-escalationwithout statistical significance, OR 1.73, 95% CI 0.97e3.06),without heterogeneity (I2¼ 0%). Overall, the large associationbetween de-escalation and survival observed crudely wassmaller in adjusted analyses of these observational studies andnon-existent in RCTs (Fig. 5).

Secondary outcomes

Clinical failure was variably defined and reported in only sixstudies [13,18e20,22,24]. Significantly fewer patients with bacter-aemia experienced treatment failure with de-escalation (unad-justed OR 0.44, 95% CI 0.26e0.82, three studies), withoutheterogeneity. For pneumonia there was no difference betweengroups (OR 1.30, 95% CI 0.76e2.23, three studies), without het-erogeneity. Only one of the studies in the pneumonia analysis wasrandomized and the others did not report an adjusted analysis.There were no significant differences between groups for hospitalor ICU length of stay, whether only patients discharged alive wereconsidered or when all patients were analysed. The duration of

for bloodstream infections and pneumonia: systematic review andrg/10.1016/j.cmi.2016.05.023

Table 1Characteristics of included studies

Study ID Location Study years N patients Type of infectiona Infection definitionsb Place ofacquisition

ICU or non-ICUsettingc

Alvarez-Lerma 2006 [13] Spain 2000e2001 213 HAP CDC criteria with positive BAL,TA or PBS or Legionella antigen

Hospital ICU

Carugati 2015 [14] multicenter,35 countries

2005 261 CAP with BSI New infiltrate on CXR withclinical criteria and bacteraemiawith plausible pathogen

Community Both

Cremers 2014 [15] The Netherlands 2001e2011 275 BSI Blood-culture-proven pneumococcal infection Hospital/Community

Both

Eachampati 2009 [16] USA 2005e2007 134 VAP Clinical suspicion of pneumonia-positive BAL Hospital ICUFalguera 2010 [17] Spain 2006e2008 177 CAP Infiltrate on CXR with clinical criteria Community Non-ICUGarnacho-

Montero 2014 [18]Spain 2008e2012 628 Severe sepsis/

septic shock, anyEstablished criteria for severesepsis and septic shock

Hospital/Community

ICU

Giantsou 2007 [31] Greece NS 143 VAP Physician diagnosis with BAL or CXR Hospital ICUJoffe 2008 [19] USA 2000e2005 412 VAP New or persistent infiltrate on CXR

with clinical criteriaHospital ICU

Khasawnehpneumonia 2014 [20]

USA 2008 60 Pneumoniawith BSI

Same pathogen in blood and respiratorysample and pneumonia(non-defined) documented in medical chart

Hospital/Community

Both

Khasawneh UTI 2014 [21] USA 2008 65 UTI with BSI Same pathogen isolated inblood and urine cultures

Hospital/Community

Both

Kim 2012 [22] Korea 2004e2006 108 HAP New and persistent infiltrate onCXR with clinical criteria

Hospital ICU

Kollef 2006 [23] USA 2003e2004 394 VAP New and persistent infiltrate onCXR with clinical criteria

Hospital ICU

Koupetori 2014 [24] Greece 2006e2013 94þ 129 BSI BSI with at least two SIRS criteria Hospital/Community

Both

Lee 2015 [25] Taiwan 2005e2012 189 BSI community-onset monomicrobialEnterobacteriaceae BSI

Community Both

Leone 2014 [26] France 2012e2013 116 Severe sepsis,any

Established definitions for severe sepsis Hospital/Community

ICU

Rello 2004 [27] Spain 2000e2001 331 VAP New, persistent infiltrate onCXR with purulentrespiratory secretions and clinical criteria

Hospital ICU

Shime 2011 [29] Japan 2004e2009 201 BSI Monomicrobial bacteraemia caused byantibiotic-susceptible bacteria

Community Both

Shime 2013 [28] Japan 2006e2011 49 BSI Monomicrobial bacteraemiacaused by typically MDRGram-negative bacteria

Hospital Both

Soo Hoo 2005 [30] USA 1998e2002 106 HAP New, persistent infiltrate onCXR with purulent respiratorysecretions and clinical criteria

Hospital Both

a Types of infection: BSI, bloodstream infection; CAP, community-acquired pneumonia; HAP, hospital-acquired pneumonia; UTI, urinary tract infection; VAP, ventilator-associated pneumonia.

b BAL, bronchoalveolar lavage; BSI, bloodstream infection; CDC, Centers for Disease Control; CSR, chest X-ray; MDR, multidrug-resistant bacteria; PBS, protected brushspecimen; SIRS, sepsis inflammatory response syndrome; TA, tracheal aspirate.

c The studies' settings: ICU, intensive care unit.

M. Paul et al. / Clinical Microbiology and Infection xxx (2016) 1e84

antibiotic treatment was significantly longer with de-escalation instudies of bacteraemia, with a mean difference of 3.06 days (95% CI1.98e4.14), four studies, with no heterogeneity (I2¼ 0%). However,survival in the de-escalation groupwas higher, as shown above, andtreatment duration reported in the studies was not adjusted forsurvival. With pneumonia, there was no difference in antibiotictreatment duration (four studies).

Only two studies compared the emergence of resistance. Onereported that the emergence of extended-spectrum b-lactamase-producing bacteria occurred less often in the de-escalation groupthan in the non-de-escalation group (1/86 (1.16%) versus 9/103(8.7%)) [25], whereas the other showed no difference betweengroups in the detection of various multi-drug resistant bacteria(methicillin-resistant S. aureus, carbapenem-resistant Gram-nega-tives, extended-spectrum b-lactamase-producing bacteria andStenotrophomonas maltophilia [22]. Superinfections were assessedin two studies of ventilator-associated pneumonia in the ICU[19,26] with no significant pooled difference between groups. Asingle RCT reported on any adverse effect and found a rate of 8/88(9%) in the de-escalation group versus 16/89 (18%) in the non-de-

Please cite this article in press as: Paul M, et al., Antibiotic de-escalationmeta-analysis, Clinical Microbiology and Infection (2016), http://dx.doi.o

escalation group [17] and a single RCT reported on C. difficile-associated diarrhea with no events in either group [26].

Discussion

In a systematic review of studies examining the strategy ofantibiotic susceptibility testing-guided antibiotic de-escalationfrom covering empirical treatment for bacteraemia, pneumonia orsevere sepsis/septic shock, we foundmajor differences in results forobservational studies and RCTs. Observational studies reported asignificant and large unadjusted association between de-escalationand lower mortality (OR 0.53, 95% CI 0.39e0.73) and most claimedthis as their conclusion. The adjusted effect estimate from obser-vational studies was smaller and non-existent when the analysiswas limited to three RCTs that showed an opposite trend (OR 1.73,95% CI 0.97e3.06). It is likely that the crude lower mortality withde-escalation in observational studies is due to bias. Physicianswere probably more likely to change to a narrow-spectrum anti-biotic in patients who had improved at the time when culture re-sults became available. It is difficult to measure and adjust for this

for bloodstream infections and pneumonia: systematic review andrg/10.1016/j.cmi.2016.05.023

Study Study design

Coh

ort s

elec

tion

Exp

osur

e

asse

ssm

ent

Con

foun

der

mat

chin

g

Con

foun

der

asse

ssm

ent

Out

com

e

asse

ssm

ent

Miss

ing

data

Co-

inte

rven

tions

Alvarez-Lerma 2006 Prospective DY DY DY DN DN PY PN

Carugati 2015 Retrospective DY PN PY PN PY DN PN

Cremers 2014 Retrospective DY PN PY PN DN DY PN

Eachampati 2009 Retrospective DY DN PY PN PN DN PN

Falguera 2010 RCT DY DY PY PY DY DY PN

Garnacho-Montero 2014 Prospective DY PY PY DY DY DY PY

Giantsou 2007 Prospective DY DY DN DN DY PY DN

Joffe 2008 Retrospective DY DY PY DY DY DY DN

Khasawneh pneumonia

2014 Retrospective DY PN DN DN PN DY PN

Khasawneh UTI 2014 Retrospective DY PN DN DN DN DY PN

Kim 2012 RCT DY DY DY PY DY PY DY

Kollef 2006 Prospective PN DY DN DN DN DN PN

Koupetori 2014 Prospective DY PY PY PY DY DY PN

Lee 2015 Retrospective DY PN PY DN PY DY PN

Leone 2014 RCT DY DY DN DN DY PY PN

Rello 2004 Prospective DY PY DN DN DN PN PN

Shime 2011 Retrospective DY PN DN DN DN DY PN

Shime 2013 Retrospective DY PN DN DN PY PY PN

Soo Hoo 2005 Retrospective DN PN PY DN PY DY DN

DY Definitely yes (low risk of bias)

PY Probably yes

PN Probably no

DN Definitely no (high risk of bias)

Fig. 2. Risk of bias in included studies; RCT, randomized controlled trial; Prospective/retrospective, refers to observational cohort studies.

M. Paul et al. / Clinical Microbiology and Infection xxx (2016) 1e8 5

trend in clinical status to which the treating physicians react. RCTs,despite risk for confounding in one of the studies, are profoundlydifferent from the observational studies in that patient selection isnot based on the initial response to treatment or source control, anda protocol-guided intervention is applied. RCTs are the onlyappropriate design to assess the effect of de-escalation on survival.

We do not actually expect a mortality effect with de-escalation.It should be enough to show that de-escalation is not associatedwith adverse consequences (mortality or clinical failure) to adopt it.Our main expectation from de-escalation is to limit the selectionand development of bacteria resistant to broad-spectrum antibi-otics in the individual treated. Unfortunately, only 2/19 of thestudies reported on isolation of resistant bacteria with versuswithout de-escalation and none performed active surveillance forresistance assessment. Hence, we do not have evidence showingthat de-escalation actually results in less resistance. The effects ofde-escalation on resistance in the unit, ward or hospital and inother patients can be demonstrated only in studies comparing the

Please cite this article in press as: Paul M, et al., Antibiotic de-escalationmeta-analysis, Clinical Microbiology and Infection (2016), http://dx.doi.o

different strategies at the unit, ward or hospital level; such studieswere not identified. This could be achieved in cluster-randomizedtrials, cross-over and beforeeafter studies or other comparativestudies designed at the cluster level.

The large clinical heterogeneity in the patients assessed andinterventions applied in the RCTs included in our review does notallow strong confidence in the effect estimate of their meta-analysis suggesting higher mortality with de-escalation. De-esca-lation could result in highermortality if the isolated bacteria are notthe actual, or only, bacteria causing the infection. This is more likelyto happen with pneumonia where respiratory samples might notrepresent well the aetiology of the pneumonia, and the diagnosis ofventilator-associated pneumonia as the source of infection is nottotally specific [32,33]. Indeed, in the observational studies, theassociation between de-escalation and survival observed amongpatients with bacteraemia was not observed for patients withpneumonia. Another possibility is that the narrow-spectrum anti-biotics were older and less effective than the empirical, newer

for bloodstream infections and pneumonia: systematic review andrg/10.1016/j.cmi.2016.05.023

Fig. 3. Unadjusted analysis for all-cause mortality for de-escalation (DE) versus non-de-escalation (non-DE) antibiotic strategies.

Fig. 4. Adjusted analysis for all-cause mortality for de-escalation (DE) versus non-de-escalation (non-DE) antibiotic strategies; RCTs are denoted by a star in the forest plot.

M. Paul et al. / Clinical Microbiology and Infection xxx (2016) 1e86

Please cite this article in press as: Paul M, et al., Antibiotic de-escalation for bloodstream infections and pneumonia: systematic review andmeta-analysis, Clinical Microbiology and Infection (2016), http://dx.doi.org/10.1016/j.cmi.2016.05.023

Fig. 5. Pooled odds ratios for all-cause mortality with de-escalation (DE) versus non-de-escalation (non-DE) by study design.

M. Paul et al. / Clinical Microbiology and Infection xxx (2016) 1e8 7

agents. Differences between antibiotics probably exist and cannotbe appreciated from the non-inferiority trials used for theirapproval [34,35].

The main limitation of our analysis concerns the heterogeneityof the patient populations, types of infection and the intervention.A systematic review recently summarized the variability in thedefinitions of de-escalation, despite having limited study inclusionto those conducted in the ICU [36]. A consensus statement on b-lactam hierarchy might prove useful to standardize future de-escalation studies [37]. Heterogeneity in the unadjusted mortalityanalysis was partially explained by the type of infection andperhaps, thereby, by the degree of certainty in the pathogens asresponsible for the infection that directed de-escalation. Severity ofinfection and the empirical antibiotic regimen (broad-spectrum ornot) probably did not introduce heterogeneity into our analyses.We restricted inclusion to patients treated with appropriateempirical antibiotic treatment to standardize an important deter-minant of survival from sepsis [4].

In summary, observational studies show lower mortalityfollowing antibiotic de-escalation guided by culture results amongpatients with bacteraemia, pneumonia or severe sepsis, whereasthree small RCTs favoured no de-escalation without reaching sta-tistical significance. We have no information on the effects of de-escalation on resistance in the individual treated or in the envi-ronment of the patient. Observational studies need to focus on theecological impact of de-escalation on resistance rather than onmortality. RCTs are needed to assess the safety of antibiotic de-escalation, especially in the context of pneumonia and to assessthe effects of a policy implementing antibiotic de-escalation onresistance.

Transparency Declaration

The authors declare that they have no conflicts of interest.

Appendix A. Supplementary data

Additional Supporting Information may be found in the onlineversion of this article http://dx.doi.org/10.1016/j.cmi.2016.05.023.

References

[1] Paul M, Greub G. The hidden killer: are we improving the management ofbacteremia? Clin Microbiol Infect 2015;21:291e4.

[2] Weir DL, Majumdar SR, McAlister FA, Marrie TJ, Eurich DT. The impact ofmultimorbidity on short-term events in patients with community-acquiredpneumonia: prospective cohort study. Clin Microbiol Infect 2015;21:264e7e264e213.

[3] Leibovici L, Shraga I, Drucker M, Konigsberger H, Samra Z, Pitlik SD. Thebenefit of appropriate empirical antibiotic treatment in patients with blood-stream infection. J Intern Med 1998;244:379e86.

[4] Paul M, Shani V, Muchtar E, Kariv G, Robenshtok E, Leibovici L. Systematicreview and meta-analysis of the efficacy of appropriate empiric antibiotictherapy for sepsis. Antimicrob Agents Chemother 2010;54:4851e63.

[5] Kaye KS. Antimicrobial de-escalation strategies in hospitalized patients withpneumonia, intra-abdominal infections, and bacteremia. J Hosp Med2012;7(Suppl 1):S13e21.

Please cite this article in press as: Paul M, et al., Antibiotic de-escalationmeta-analysis, Clinical Microbiology and Infection (2016), http://dx.doi.o

[6] Laxminarayan R, Duse A, Wattal C, Zaidi AK, Wertheim HF, Sumpradit N, et al.Antibiotic resistance-the need for global solutions. Lancet Infect Dis 2013;13:1057e98.

[7] Schulz L, Osterby K, Fox B. The use of best practice alerts with the develop-ment of an antimicrobial stewardship navigator to promote antibiotic de-escalation in the electronic medical record. Infect Control Hosp Epidemiol2013;34:1259e65.

[8] Silva BN, Andriolo RB, Atallah AN, Salomao R. De-escalation of antimicrobialtreatment for adults with sepsis, severe sepsis or septic shock. CochraneDatabase Syst Rev 2013;3:CD007934.

[9] Horan TC, Gaynes RP. Surveillance of nosocomial infections. In: Mayhall CG,editor. Hospital Epidemiology and Infection Control. 3rd ed. Philadelphia:Lippincott Williams & Wilkins; 2004. p. 1659e702.

[10] Higgins JPT, Green S, editors. Cochrane Handbook for Systematic Reviews ofInterventions Version 5.1.0 [updated March 2011]. The Cochrane Collabora-tion; 2011. Available from, www.cochrane-handbook.org.

[11] Wan X, Wang W, Liu J, Tong T. Estimating the sample mean and standarddeviation from the sample size, median, range and/or interquartile range.BMC Med Res Methodol 2014;14:135.

[12] Review Manager (RevMan) [Computer program]. Version 5.3. Copenhagen:The Nordic Cochrane Centre, The Cochrane Collaboration; 2014.

[13] Alvarez-Lerma F, Alvarez B, Luque P, Ruiz F, Dominguez-Roldan JM,Quintana E, et al. Empiric broad-spectrum antibiotic therapy of nosocomialpneumonia in the intensive care unit: a prospective observational study. CritCare 2006;10:R78.

[14] Carugati M, Franzetti F, Wiemken T, Kelly R, Peyrani P, Blasi F, et al. De-escalation therapy among bacteraemic patients with community-acquiredpneumonia. Clin Microbiol Infect 2015;21:936e911e38.

[15] Cremers AJ, Sprong T, Schouten JA, Walraven G, Hermans PW, Meis JF, et al.Effect of antibiotic streamlining on patient outcome in pneumococcal bac-teraemia. J Antimicrob Chemother 2014;69:2258e64.

[16] Eachempati SR, Hydo LJ, Shou J, Barie PS. Does de-escalation of antibiotictherapy for ventilator-associated pneumonia affect the likelihood of recurrentpneumonia or mortality in critically ill surgical patients? J Trauma 2009;66:1343e8.

[17] Falguera M, Ruiz-Gonzalez A, Schoenenberger JA, Touzon C, Gazquez I,Galindo C, et al. Prospective, randomised study to compare empirical treat-ment versus targeted treatment on the basis of the urine antigen results inhospitalised patients with community-acquired pneumonia. Thorax 2010;65:101e6.

[18] Garnacho-Montero J, Gutierrez-Pizarraya A, Escoresca-Ortega A, Corcia-Palomo Y, Fernandez-Delgado E, Herrera-Melero I, et al. De-escalation ofempirical therapy is associated with lower mortality in patients with severesepsis and septic shock. Intensive Care Med 2014;40:32e40.

[19] Joffe AR, Muscedere J, Marshall JC, Su Y, Heyland DK. Canadian Critical CareTrials G. The safety of targeted antibiotic therapy for ventilator-associatedpneumonia: a multicenter observational study. J Crit Care 2008;23:82e90.

[20] Khasawneh FA, Karim A, Mahmood T, Ahmed S, Jaffri SF, Mehmood M. Safetyand feasibility of antibiotic de-escalation in bacteremic pneumonia. InfectDrug Resist 2014;7:177e82.

[21] Khasawneh FA, Karim A, Mahmood T, Ahmed S, Jaffri SF, Tate ME, et al.Antibiotic de-escalation in bacteremic urinary tract infections: potential op-portunities and effect on outcome. Infection 2014;42:829e34.

[22] Kim JW, Chung J, Choi SH, Jang HJ, Hong SB, Lim CM, et al. Early use of imi-penem/cilastatin and vancomycin followed by de-escalation versus conven-tional antimicrobials without de-escalation for patients with hospital-acquired pneumonia in a medical ICU: a randomized clinical trial. Crit Care2012;16:R28.

[23] Kollef MH, Morrow LE, Niederman MS, Leeper KV, Anzueto A, Benz-Scott L,et al. Clinical characteristics and treatment patterns among patients withventilator-associated pneumonia. Chest 2006;129:1210e8.

[24] Koupetori M, Retsas T, Antonakos N, Vlachogiannis G, Perdios I, Nathanail C,et al. Bloodstream infections and sepsis in Greece: over-time change ofepidemiology and impact of de-escalation on final outcome. BMC Infect Dis2014;14:272.

[25] Lee CC, Lee NY, Chen PL, Hong MY, Chan TY, Chi CH, et al. Impact of antimi-crobial strategies on clinical outcomes of adults with septic shock andcommunity-onset Enterobacteriaceae bacteremia: de-escalation is beneficial.Diagn Microbiol Infect Dis 2015;82:158e64.

[26] Leone M, Bechis C, Baumstarck K, Lefrant JY, Albanese J, Jaber S, et al. De-escalation versus continuation of empirical antimicrobial treatment in severe

for bloodstream infections and pneumonia: systematic review andrg/10.1016/j.cmi.2016.05.023

M. Paul et al. / Clinical Microbiology and Infection xxx (2016) 1e88

sepsis: a multicenter non-blinded randomized noninferiority trial. IntensiveCare Med 2014;40:1399e408.

[27] Rello J, Vidaur L, Sandiumenge A, Rodriguez A, Gualis B, Boque C, et al. De-escalation therapy in ventilator-associated pneumonia. Crit Care Med2004;32:2183e90.

[28] Shime N, Kosaka T, Fujita N. De-escalation of antimicrobial therapy for bac-teraemia due to difficult-to-treat Gram-negative bacilli. Infection 2013;41:203e10.

[29] Shime N, Satake S, Fujita N. De-escalation of antimicrobials in the treatment ofbacteraemia due to antibiotic-sensitive pathogens in immunocompetent pa-tients. Infection 2011;39:319e25.

[30] Soo Hoo GW, Wen YE, Nguyen TV, Goetz MB. Impact of clinical guidelines inthe management of severe hospital-acquired pneumonia. Chest 2005;128:2778e87.

[31] Giantsou E, Liratzopoulos N, Efraimidou E, Panopoulou M, Alepopoulou E,Kartali-Ktenidou S, et al. De-escalation therapy rates are significantly higherby bronchoalveolar lavage than by tracheal aspirate. Intensive Care Med2007;33:1533e40.

[32] Fabregas N, Ewig S, Torres A, El-Ebiary M, Ramirez J, de La Bellacasa JP, et al.Clinical diagnosis of ventilator associated pneumonia revisited: comparative

Please cite this article in press as: Paul M, et al., Antibiotic de-escalationmeta-analysis, Clinical Microbiology and Infection (2016), http://dx.doi.o

validation using immediate post-mortem lung biopsies. Thorax 1999;54:867e73.

[33] Mondi MM, Chang MC, Bowton DL, Kilgo PD, Meredith JW, Miller PR. Pro-spective comparison of bronchoalveolar lavage and quantitative deep trachealaspirate in the diagnosis of ventilator associated pneumonia. J Trauma2005;59:891e5. discussion 895e6.

[34] Yahav D, Lador A, Paul M, Leibovici L. Efficacy and safety of tigecycline: asystematic review and meta-analysis. J Antimicrob Chemother 2011;66:1963e71.

[35] Yahav D, Paul M, Fraser A, Sarid N, Leibovici L. Efficacy and safety of cefepime:a systematic review and meta-analysis. Lancet Infect Dis 2007;7:338e48.

[36] Tabah A, Cotta MO, Garnacho-Montero J, Schouten J, Roberts JA, Lipman J,et al. A Systematic review of the definitions, determinants, and clinical out-comes of antimicrobial de-escalation in the intensive care unit. Clin Infect Dis2016;62:1009e17.

[37] Weiss E, Zahar JR, Lesprit P, Ruppe E, Leone M, Chastre J, et al. Elaboration of aconsensual definition of de-escalation allowing a ranking of beta-lactams. ClinMicrobiol Infect 2015;21(7):649e1e649e10.

for bloodstream infections and pneumonia: systematic review andrg/10.1016/j.cmi.2016.05.023