REVIEW

Guideline Recommendations for EmpiricalAntimicrobial Therapy: An Appraisal of ResearchEvidence for Clinical Decision-Making in Ethiopia

Balew Arega . Asnake Agunie . Abraham Minda . Amdemeskel Mersha .

Alazar Sitotaw . Getachew Weldeyohhans . Ayele Teshome .

Kelemework Adane . Getachew Mengistu

Received: May 1, 2020 / Published online: June 22, 2020� The Author(s) 2020

ABSTRACT

Introduction: The rapid spread of drug resis-tance is forcing standard treatment guidelines(STGs) to become more appropriate with dueconsideration of the evidence on the antimi-crobial resistance (AMR) situation in Ethiopia.Therefore, we aimed to assess the local AMRrecommendations in the STGs for empiricalantibacterial prescriptions for the five commoninfectious syndromes. We also determined thequality of AMR reviews conducted in thecountry.Methods: We conducted a review of the STGsused in the health centers, general hospitals,

and primary hospitals in Ethiopia and assessedthe AMR recommendations in STGs for empiri-cal antibacterial prescriptions for community-acquired pneumonia (CAP), urinary tract infec-tion (UTI), tonsillopharyngitis, acute otitismedia (AOM), and bacterial dysentery. Next, weperformed an overview of AMR reviews pub-lished in Ethiopia. We used the MEDLINE/PubMed, Embase, Cochrane Library, and Goo-gle Scholar databases to identify AMR reviews.The review characteristics were extracted. Wealso evaluated the quality of each included AMRreview using a measurement tool to assess thesystematic review scale (AMSTAR 2).Results: A total of 6 STGs and 12 AMR reviewsconducted in the country were included. Thechoice of empirical antibacterials for similarinfectious syndromes (and editions) was com-parable across the three levels of the health caresystem. None of the STGs evaluated includedthe local AMR recommendations for empiricalantibacterial prescriptions for five commoninfectious syndromes. Of all the AMR reviewsincluded, 75% had low and below method-ologic quality, and none had a high-qualityscore using the AMSTAR 2 tool.Conclusion: Standard treatment guidelines didnot consider local AMR recommendations forempirically prescribing antibacterials for com-mon infectious syndromes. The AMR reviewspublished in the country produced poormethodologic quality evidence for clinicalapplications. This highlights the need to

Digital Features To view digital features for this articlego to https://doi.org/10.6084/m9.figshare.12388859.

Electronic supplementary material The onlineversion of this article (https://doi.org/10.1007/s40121-020-00308-3) contains supplementary material, which isavailable to authorized users.

B. Arega (&) � A. Agunie � A. Minda � A. Mersha �A. Sitotaw � G. Weldeyohhans � A. TeshomeYekatit 12 Hospital Medical College, Addis Ababa,Ethiopiae-mail: balewmlt@gmail.com

K. AdaneMeklle University, College of Health Science,Mekelle, Ethiopia

G. MengistuDebere Markos University, College of HealthScience, Debre Markos, Ethiopia

Infect Dis Ther (2020) 9:451–465

https://doi.org/10.1007/s40121-020-00308-3

improve the methodologic quality to providethe best available evidence for clinical decision-making and curb the ongoing AMR in Ethiopia.Trial registration: Retrospectively registered(15/07/2020).

Keywords: Antimicrobial resistance; Ethiopia;Treatment guideline

Key Summary Points

Why carry out this study?

The national standard treatmentguidelines (STGs) have become moreappropriate because of using evidence onthe antibacterial resistance (AMR)situation in Ethiopia.

Despite the growing number of STGs andAMR reviews in the country, acomprehensive evaluation of STGs’ AMRrecommendations and the quality ofreviews on AMR was lacking.

This review describes the level ofrecommendations related to local AMRevidence in the STGs for empiricalantibacterial prescriptions for fivecommon infectious syndromes and thequality of the AMR reviews publishedsince the national STGs were launched.

What was learned from the study?

None of the STGs evaluated included thelocal AMR recommendations for empiricalantibiotic prescriptions for the fivecommon infectious syndromes. Morethan 75% of the reviews on AMR in thecounty also had low and below quality,and none had a high-quality scoreaccording to the AMSTAR 2 tool.

Reviews published on AMR producedmethodologically poor quality evidencefor clinical applications. This highlightsthe need to improve the quality to providethe best available evidence for clinicaldecision-making and curb the ongoingAMR in the country.

INTRODUCTION

Antimicrobial resistance (AMR) has emerged asone of the principal public health problems ofthe twenty-first century, threatening the effec-tive prevention and treatment of ever-increas-ing bacterial infections [1]. The problem iscommon mainly in low-income countrieswhere microbiologic diagnostics are often non-existent or weak and alternative antimicrobialagents are not commonly available [2].

The rapid spread of AMR has led to stan-dardized treatment guidelines (STGs) being tai-lored according to the local resistance patternsof a given area [3]. The guidelines supportclinical decision-making through a consensualprocess based on locally generated evidence onAMR and facilitate the best available choice ofantibacterial agent, preventing relevant resis-tances [4]. This is effective when the guidelinesare developed with the best available localresearch evidence from AMR reviews [system-atic reviews with or without meta-analysis (SRsand MAs)] [5]. The reviews produce strong datafor clinical applications when carried out usingrandomized clinical trials [6]. However, whetherAMR reviews of observational studies generateimportant clinical evidence rather thanimprove the academic ranks of the reviewers is asignificant concern [7].

Bacterial infections are the major cause ofdeath in Ethiopia, and antibacterial resistancethreatens the management of bacterial infec-tion in the health care setting [8]. The firstnational AMR baseline survey (2009) revealedthat most bacteria causing common infectionsin humans showed considerable resistance tocommonly used first-line antibacterial agents[9]. The country’s Medicine and Health CareAdministration and Control Authority(FMHACA), currently called the EFDA (EthiopiaFood and Drug Administration), has establishedSTGs for the three levels of the health care sys-tem (health centers, primary and general hos-pitals) since 2004 and has an updated thirdedition [10]. The guidelines aimed to ensure theeffective and safe use of medicines, containinghealth care costs, and preventing antimicrobialresistance. However, a comprehensive

452 Infect Dis Ther (2020) 9:451–465

evaluation of the local recommendations rela-ted to AMR for empirical antibacterial prescrip-tion for common infectious syndromes in theSTGs was lacking. Furthermore, information onthe quality of reviews on AMR, the potentialsources of the recommendations to prepareSTGs, in Ethiopia is not available. Therefore, thepurpose of this study was twofold. First, itassessed the STGs’ local antibacterial resistancerecommendation level for empirically prescrib-ing antibacterial agents for community-ac-quired pneumonia (CAP), urinary tractinfections (UTIs), tonsillopharyngitis, bacterialdysentery, and acute otitis media (AOM). Sec-ond, it evaluated the methodologic quality ofreviews on AMR published in Ethiopia using theAMSTAR 2 tool considering their use in clinicalapplications.

METHODS

This is a review of STGs used in the three healthcare levels (health centers, primary and generalhospitals) and an overview of reviews on AMRrelated to bacteria isolated from human samplesin Ethiopia.

Search, Inclusion, and Data Extractionof STGs

We retrieved all STGs published up to October2019 from the FMHACA (currently EPDA) officeand their official web site in Ethiopia [11]. Weincluded the latest two editions for each healthcare level related to five common infectioussyndromes [10, 12– 16].

We extracted data about the types of infec-tious syndrome (CAP, UTIs, AOM tonsil-lopharyngitis, and bacterial dysentery), editionsof STGs, and types of empirical antibacterialchoices (first line, second line, or alternative).We also extracted the level of recommendationsfor empirical antibacterial prescriptions for theincluded clinical infection syndromes. Thelevels were classified into three categories assatisfactory, partially satisfactory, and unsatis-factory based on epidemiologic and resistancepattern data [5] (Table 1). Two authors (BA andGM) carried out the STG search and data

extraction. We resolved the one-to-one dis-agreements that occurred during the search anddata extraction by consulting a third person.

Searching Reviews on AMR

We included reviews on AMR (SRs and MAs)published in Ethiopia. We used the MEDLINE/PubMed, Embase, Cochrane Library, and Goo-gle Scholar databases to search the originalreviews. A hand search was also done for cross-reference lists of the identified original reviews.A flow chart showing the selection of reviewswas made following the Preferred ReportingItems for Systematic Reviews and Meta-Analyses(PRISMA) statement guideline [17]. An elec-tronic search was performed using a combina-tion of keywords for the MEDLINE/PubMeddatabase using the Medical Science Heading(MeSH) terms: [‘‘bacterial’’(MeSHTerms) OR‘‘bacteria’’(AllFields)] OR [‘‘microorgan-isms’’(MeSHTerms) OR [‘‘microorganisms’’ (AllFields) AND ‘‘antimicrobial resistance’’ (AllFields)] OR ‘‘antimicrobial resistance’’ [MeSHTerms] OR ‘‘drug-resistance’’ [All Fields] OR‘‘drug-resistance’’ [MeSH Terms] AND (‘‘meta-analysis and systemic review’’ [MeSH Terms] OR[‘‘meta-analysis and systemic review’’ (All Fields)OR ‘‘systemic review’’ (All Fields)] OR ‘‘systemicreview’’ (MeSHterm)] AND [‘‘Ethiopia’’ (MeSHTerms) OR ‘‘Ethiopia’’ (All Fields)]. The reviews

Table 1 Hierarchy of recommendations related toantimicrobial resistance in empirical therapy

Satisfactory: An alternative empirical antibiotic

recommendation was supported by our country-

specific resistance patterns findings

Partial satisfactory: An alternative empirical antibiotic

therapy recommendation was supported by

inconsistent resistance patterns, which is not a

country-specific finding

Unsatisfactory: The empirical antibiotic

recommendation did not support any resistance

patterns or was not justified by country-specific

resistance patterns

Infect Dis Ther (2020) 9:451–465 453

were limited to the human category. The datasearch was done between 1 September 2019 and15 October 2019.

AMR Review Inclusion and DataExtraction

We included AMR reviews fulfilling the follow-ing criteria: (1) conducted in Ethiopia, (2) pub-lished and written in English, (3) took samplesfrom humans (isolated from either diseased orhealthy individuals), and (4) reported theantibacterial resistance. We excluded AMRreviews on tuberculosis, viruses, parasites, andfungi. We (BA, GM) conducted the dataextraction using a standardized and pretestedformat in Microsoft Excel. The data extractedincluded: the the first author of the reviews,number of original studies in each review, studyyears in the original studies, review type (SR,MA), aims of the reviews, and the main/pooled/resistance level of different antibacterial drugsmainly used for the treatment of the includedinfectious syndromes.

Quality of AMR Reviews (SRs and MAs)

Two reviewers (BA and GM) independentlyassessed the methodologic quality of each AMRreview included in the overview using theAMSTAR 2 tool. This tool consists of 16 ques-tions (Q), each being answered with ‘‘yes,’’‘‘partial yes,’’ ‘‘no,’’ or ‘‘no meta-analysis done.’’Of the 16 items or domains in the AMSTAR 2framework, the creators of the tool specify 7critical domains (items Q2, Q4, Q7, Q9, Q11,Q13, and A15) that can substantively impactthe validity of a review. The other nine are non-critical domains (items Q1, Q2, Q5, Q6, Q8,Q10, 14Q, and Q16) (Supplement 1).

Following the evaluation of each of theseven critical domains as well as the nine non-critical domains, the overall result was rated ashigh quality, moderate quality, low quality, andcritically low quality review (Supplement 2)[18]. A one-to-one disagreement that occurredduring quality assessment between reviewers(BA and GM) was solved through by a thirdperson.

Compliance with Ethics Guidelines

This article is based on previously conductedstudies and does not contain any studies withhuman participants or animals performed byany of the authors.

RESULTS

Standardize Treatment Guidelinesand Recommendations

From 2004–2019, nine STGs (three editions foreach health care level) were developed byEthiopia FMHACA. Of these, a total of six recenteditions (the recent edition in 2014), two foreach level of health care, were included forevaluation. The empirical antibacterial pre-scriptions in all STGs evaluated for the fiveclinical syndromes were described as first lineand alternative. As presented in Table 2, the firstline provided antibacterial drugs for similarclinical syndromes, and the STG editions arecomparable for all three levels of the health caresystem. For instance, macrolides (clar-ithromycin or azithromycin) are the primarychoice for CAP, a fluoroquinolone (cipro-floxacin or norfloxacin) for UTIs and bacterialdysentery, and penicillin (amoxicillin) for ton-sillopharyngitis and AOM in the last STG edi-tions for all the health care system levels.

Compliance with the desired criteria isshown in Table 1. Partially satisfactory recom-mendations related to antibacterial resistanceand alternative antibacterial agents chosen forpatients for recent antibacterial use were madefor mild CAP in the latest editions of all healthcare system STGs. For all other clinical syn-dromes, the empirical antibacterial recommen-dation did not describe any resistance or foundit was not justified by country-specific resistancepatterns (unsatisfactory) in all STGs.

Characteristics of AMR Reviews (SRsand MAs)

Of the 146 AMR reviews identified, 21 were fullyscreened. Only 12 of them with a total of 312

454 Infect Dis Ther (2020) 9:451–465

Table

2Guidelin

esforem

piricalantibioticchoice

forfiveinfectious

synd

romes

andantimicrobialresistance

recommendation

sin

Ethiopia

Typeof

infectious

synd

rome

STGsforthehealth

care

level

STGsforhealth

centers

STGsforprim

aryho

spitals

STGsforgeneralho

spitals

2ndedition(2010)

[10]

3rdedition(2014)

[12]

2ndedition(2010)

[13]

3rdedition(2014)

[14]

2ndedition(2010)

[15]

3rdedition(2014)

[16]

Mild

CAP

First-lin

eantibiotics

AMO

CLR,E

RY,D

OX

AMO

CLR,A

ZM

AMP

CLR/A

ZM/D

OX

Second

-line

antibiotics

ERY,D

OX

PPEP

CLR/ERY?

AMO/

AMC

ERY,D

OX,P

PEP

CLR/A

ZM

?AMP/

AMC

ERY,D

OX,P

PEP

CLR/A

ZM

?

AMO/A

MC

AMRrecommendation

Satisfactory

Partialsatisfactory

44

4

Unsatisfactory

44

4

Severe

CAP

First-lin

eantibiotics

Referred

Referred

BPE

P?

GN

CRO/BPE

P?

AZM/C

LR

BPE

P?

GN

CRO/BPE

P?

AZM/C

LR

Second

-line

antibiotics

CRO,E

RY,D

OX

–CRO,E

RY,D

OX

CRO,E

RY,D

OX

AMRrecommendation

Satisfactory

Partialsatisfactory

Unsatisfactory

44

44

UTI(uncom

plicated)

First-lin

eantibiotics

SXT

CIP,N

OR

SXT

CIP

orNOF

SXT

CIP,N

OF

Second

-line

antibiotics

NOF,

AMO

SXT

NOR,A

MO

NIT,S

XT,C

RO*

NOR,A

MO

NIT,S

XT,C

RO

AMRrecommendation

Satisfactory

Infect Dis Ther (2020) 9:451–465 455

Table

2continued

Typeof

infectious

synd

rome

STGsforthehealth

care

level

STGsforhealth

centers

STGsforprim

aryho

spitals

STGsforgeneralho

spitals

2ndedition(2010)

[10]

3rdedition(2014)

[12]

2ndedition(2010)

[13]

3rdedition(2014)

[14]

2ndedition(2010)

[15]

3rdedition(2014)

[16]

Partialsatisfactory

4

Unsatisfactory

44

44

4

Ton

sillitisandph

aryngitis

First-lin

eantibiotics

AMO

AMO

AMO

AMO

AMO

AMO

Second

-line

antibiotics

AMP,

AMC

AMP,

AMC

AMP,

AMC,P

PEP,

BPE

P

AMP,

AMC

AMP,

AMC,P

PEP

AMP,

AMC

AMRrecommendation

Satisfactory

Partialsatisfactory

Unsatisfactory

44

44

44

AOM

First-lin

eantibiotics

AMO

AMO

AMO

AMO

AMO

AMO

Second

-line

antibiotics

AMP,

SXT,A

MC,

ERY

AMP,

SXT,A

MC/

CIP

AMP,

SXT,A

MC,

ERY

AMP,

AMC

AMP,

SXT,A

MC,

ERY

AMP,

AMC,C

IP,

CHL

AMRrecommendation

Satisfactory

Partialsatisfactory

Unsatisfactory

44

44

4

Bacterial

dysentery

First-lin

eantibiotics

CIP

CIP

CIP

CIP

CIP

Second

-line

antibiotics

SXT,C

RO

SXT,C

RO

SXT,C

RO

SXT,C

RO

SXT,C

RO

456 Infect Dis Ther (2020) 9:451–465

original studies fulfilled the inclusion criteriaand were included in this overview (Fig. 1). Theoriginal studies were conducted between 1967and 2018. Only two of the included AMRreviews [28–30] were SRs, and the remaining tenwere both SRs and MAs [19–27]. Four AMRreviews reviewed the AMR profile of gram-pos-itive bacteria (GPB) [20, 21, 26, 28], fourreviewed the AMR of gram-negative bacteria(GNB) [19, 24, 25, 30], and four reviewed theAMR of both GPB and GNB [23, 24, 27, 29].

The pooled resistance level of different bac-teria species for commonly used antibacterialdrugs is shown in Table 3. The resistance level ofampicillin (AMP) was 64–98.1% [19–24, 26],amoxicillin (AMO) 40–97% [20–22, 24–26],methicillin-resistant S. aureus (MRSA) 10.94-49% [20, 21, 23–26], ceftriaxone (CRO) 9.3–45%[22–27], erythromycin (ERY) 30–97.1%[20–22, 24–26], ciprofloxacin (CIP) 3.6–35.1%[19–24], and trimethoprim-sulfamethoxazole(SXT) 35–68% [19– 21, 24–26].

Quality of AMR Reviews (SRs and MAs)Using AMSTAR 2 Items

Table 4 presents the results of each AMR reviewaccording to the AMSTAR 2 questions and theirquality scores. All of the AMR reviews imple-mented components of PICO for observationalstudies (Q1) and the absence of a statement ofconflict of interest (Q16). In [ 50% of thereviews, the original studies provided adequatedetail (Q8), provided a satisfactory explanationfor any heterogeneity observed (Q14), usedappropriate methods for a statistical combina-tion of results (Q11), assessed the potentialimpact of risk of bias on the overall results(Q12), performed data extraction in duplicate(Q6), explained their selection of study designin the included studies (Q3), used a compre-hensive literature search strategy(Q4), and per-formed study selection in duplicate (Q5). Noneof the reviews reported on the sources of fund-ing for the studies included in their review(supplement 3).

Table

2continued

Typeof

infectious

synd

rome

STGsforthehealth

care

level

STGsforhealth

centers

STGsforprim

aryho

spitals

STGsforgeneralho

spitals

2ndedition(2010)

[10]

3rdedition(2014)

[12]

2ndedition(2010)

[13]

3rdedition(2014)

[14]

2ndedition(2010)

[15]

3rdedition(2014)

[16]

AMRrecommendation

Satisfactory

Partialsatisfactory

Unsatisfactory

44

44

4

CAP

commun

ity-acquired

pneumonia,UTIurinarytractinfection,

AOM

acuteotitis

media,AMO

AMO

amoxicillin,AMP

ampicillin,

AMC

amoxicillin-

clavulanate,ERYerythrom

ycin,AZM

azithrom

ycin,CLR

clarithrom

ycin,PP

EPprocainepenicillin,

BPE

Pbenzathine

penicillin,

CIP

ciprofloxacin,NOR

nor-

floxacin,

SXTtrim

ethoprim

sulfamethoxazole,C

RO

ceftriaxone,DOXdoxycycline,CHLchloramphenicol,N

ITnitrofurantoin,ededition

Infect Dis Ther (2020) 9:451–465 457

Overall Quality Score of AMR ReviewsUsing the AMSTAR 2 Tool

Overall, three-quarters of the included AMRreviews had low quality and less overall, one-fourth had moderate quality, and none had ahigh-quality score using the AMSTAR 2 score(Supplement 4).

DISCUSSION

In the era of the global burden of AMR, thetreatment for a growing number of infectiousdiseases mandates recommendations aboutAMR in the clinical treatment guidelines to helpreduce resistance [31]. To the best of ourknowledge, this is the first study to review theSTGs and evaluate the quality of AMR reviews inEthiopia, trying to link the evidence to clinicalapplication. The overview found two importantgaps: (1) The STGs did not routinely consider

local resistance patterns for empirical antibac-terial prescription for the five highly prevalentinfectious syndromes. (2) The AMR reviewspublished in the country produced evidencewith poor methodologic quality for clinicalapplication.

In this study, the majority of clinical treat-ment guidelines (health center, primary andgeneral hospital) have unsatisfactory AMR rec-ommendations for empirical antibacterial ther-apy; these findings are similar to others inwhich current guidelines elsewhere do not haveresistance recommendations. In this regard, arecent survey revealed that only 16 of 135(6.4%) guidelines discussed empirical antibiotictreatment considering specific microbiologyresistance [5]. In contrast, the WHO clinicaltreatment guideline development protocolsstrongly recommend consideration of the localresistance pattern for empirical therapy, asresistance often differs across geographicalregions or even within different health settings

Fig. 1 PRISMA flow chart of review selection

458 Infect Dis Ther (2020) 9:451–465

Table

3Characteristics

ofsystem

icreview

swithor

without

meta-analysisincluded

intheoverview

Autho

rsYears

Type

of stud

y

Num

ber

of prim

ary

stud

ies

Aim

ofthereview

Typeof

antimicrobial

agent

Penicillin

(%R)

Cephalosporin

(%R)

Macrolid

es(%

R)

Fluo

roqu

inolon

e(%

R)

Others(%

R)

Tadesse

[19]

1974—

2009

MA

14AMRof

Salmonella

speciesisolated

from

ahuman

specim

en

AMP(86.01)

CRO

(78.8)

Not reported

CIP

(3.61),

NOR(15.5)

SXT

(68),

MDR(79.6)

Deyno

[20]

2006–2

015

MA

21AMRof

CoN

Sclinically

significant

isolates

AMP(64),P

EP(57),

AMO

(40),M

RSA

(37)

CRO

(27)

ERY(30)

CIP

(17),N

OF

(39)

SXT

(50)

Eshetie

[21]

2004–2

014

MA

20MET

resistance

S.aureus

isolates

PEP(99),A

MP

(98.1),A

MO

(97),

MRSA

(35)

CRO

(30.8)

ERY(97.2)

CIP

(37.3)

SXT

(46.4)

Hussen

[22]

1999–2

018

MA

29AMRof

the

Shigella

species

AMP(83.1),A

MO

(84.1)

CRO

(9.3)

ERY(86.5)

CIP

(8.9)

MDR(83.2)

Sisay

[23]

2015–2

018

MA

37Clin

icallyrelevant

bacterialagainst

fluoroquinolone

Not

reported

Not

reported

Not reported

CIP

(18.6)

forS.

aureus

CIP

(7.4)forGBS

CIP

(24.3)

for

E.coli

Not

reported

Sisay

[24]

2000–2

018

MA

21AMRof

bacteria

from

wound

infection

MRSA

(49),A

MC

(27),A

MO

(69),

AMP(76)

forS.

aureus

AMP(84),A

MO

(73)

AMC

(57)

for

E.coli

CRO(36)

forS.

aureus

CRO

(45)

forE.coli

ERY(34)

forS.

aureus

CIP

(12)

forS.

aureus

CIP

(27)

forE.coli

SXT

(35)

for

S.aureus

SXT(53)

for

E.coli

Infect Dis Ther (2020) 9:451–465 459

Table

3continued

Autho

rsYears

Type

of stud

y

Num

ber

of prim

ary

stud

ies

Aim

ofthereview

Typeof

antimicrobial

agent

Penicillin

(%R)

Cephalosporin

(%R)

Macrolid

es(%

R)

Fluo

roqu

inolon

e(%

R)

Others(%

R)

Beshir

[25]

2007–2

017

MA

35DrugResistance

Patterns

of

E.coli

AMP(83.8),A

MO

(75.8),A

MC

(49)

DOX

(48.6)

CRO

(30.2)

ERY(52.9)

CIP

(27.6),N

OR

(32.8)

SXT,(57.5)

Deyno

[26]

1969–2

015

MA

45Resistanceof

S.aureus

toantimicrobials

AMO

(77),P

EN

(76),A

MP(75),

DOX

(43),M

ET

(47),A

MC

(30)

CRO

(34)

ERY(41)

CIP

(19),N

OR

(25)

SXT

(47)

Muhie

[27]

–MA

33Antibioticuseand

resistance

patternin

Ethiopia

Not

reported

CRO

(43.4)

to

E.coliCRO

(35.3)

toS.

aureus

Not reported

CIP

(33.2)

to

E.coli

CIP

(20.5)

toS.

aureus

MDR(68.1)

toE.coli

MDR(59.7)

toS.

aureus

Reta [28]

2003–2

018

MA

10Nasalcolonization

ofMET

resistantS.

aureus

MRSA

(10.94%)

Not

reported

Not reported

Not

reported

Not

reported

Reta [29]

–SR

39Bacterial

infections

and

theirAMR

AMP(10–

20),S.

pneumonia

AMP(47–

100%

),S.

aureus

AMP(66.7–1

00),

E.coli

CRO

(0–2

5)to

S.pneumonia

CRO

(6.7–1

00)to

S.aureus

CRO

(0–6

6.7)

to

E.coli

Not reported

CIP

(2–2

0),toS.

pneumonia

CIP

(0–6

6.7)

S.aureus

Cip

(4–1

00)to

E.coli

SXT

(20–

65)

toS.

pneumonia

SXT

(6–1

00)

toS.

aureus

SXT

(15–

70.4)to

E.coli

460 Infect Dis Ther (2020) 9:451–465

Table

4Methodologicqualityassessmentof

each

AMRreview

usingAMST

AR2item

s

Q1

Q2

Q3

Q4

Q5

Q6

Q7

Q8

Q9

Q10

Q11

Q12

Q13

Q14

Q15

Q16

Leve

l of q

ualit

y

Tade

sse [1

7]Lo

w q

ualit

y

Dey

no[1

8]M

oder

ate

qual

ity

Eshe

tie [1

9]C

ritic

ally

low

qua

lity

Hus

sen

[20]

Crit

ical

ly lo

w q

ualit

y

Sisa

y[2

1]C

ritic

ally

low

qua

lity

Sisa

y [2

2]Lo

w q

ualit

y

Besh

ir[2

3]M

oder

ate

qual

ity

Dey

no[2

4]M

oder

ate

qual

ity

Muh

ie[2

5]Lo

w q

ualit

y

Reta

[26]

Low

qua

lity

Reta

[27]

Low

qua

lity

Beye

ne [2

8]C

ritic

ally

low

qua

lity

Green

=yes;yellow=partialyes;red=No;

white—NMA:ameta-analysiswas

notcond

ucted

Table

3continued

Autho

rsYears

Type

of stud

y

Num

ber

of prim

ary

stud

ies

Aim

ofthereview

Typeof

antimicrobial

agent

Penicillin

(%R)

Cephalosporin

(%R)

Macrolid

es(%

R)

Fluo

roqu

inolon

e(%

R)

Others(%

R)

Beyene

[30]

1975–2

000

SR8

Typhoid

feverin

Ethiopia

AMP(46–

99)

SXT

(57–

100)

SxT (47.9–

100)

AMRantimicrobialresistance,M

Ameta-analysis,S

Rsystem

icreview

,AMO

amoxicillin,A

MPam

picillin,

AMCam

oxicillin-clavulanate,E

RYerythrom

ycin,P

EP

penicillin,

CIP

ciprofloxacin,NORnorfloxacin,SX

Ttrim

ethoprim

sulfamethoxazole,CRO

ceftriaxone,DOX

doxycycline,MET

methicillin,

MRSA

methicillin

resistantStaphylococcus

aureus,G

BSGroup

Bstreptococcus,R

resistance,C

oNScoagulasenegative

Staphylococcus

aureus

Infect Dis Ther (2020) 9:451–465 461

in the same country [32]. Recommendationsrelated to AMR in the STGs are especiallyessential in low-income countries includingEthiopia where the antibiogram profile of eachisolated bacteria to prevent the spread anddevelopment of resistance in the country hasnot been established for several reasons. Ifresistance is not described in a guideline, it isunlikely to consider modification of theantibacterial agent prescribed. Therefore, thefindings of these reviews imply that significantchanges in the components of STGs are neededto improve the way resistance data recommen-dations are made.

In Ethiopia, patients are referred to the nexthigher health care system, from a health centerto primary hospital to general hospital, whenthe cases are complicated or patients do notrespond to the available empirical antibacterialtherapy [32]. Therefore, patients in the higherhealth care level could have a history of takingmultiple antibiotics, one of the major risk fac-tors for AMR [33]. Antibiotic prescriptions inthe higher care levels logically should either bebased on specific microbiologic data, which isnot routinely practiced in our settings [34], orthe guideline should have a strong recommen-dation about AMR. Interestingly, however, ourinvestigation into the guidelines revealed thatthe types of antibacterial agent recommendedfor a similar syndrome (in similar editions)across the levels of health care systems werealmost equivalent. This indicates that guideli-nes could be potential sources of antimicrobialresistance as they permit individuals to takemultiple similar antibiotics at the differenthealth care levels.

The quality of clinical treatment guidelines,on the other hand, depends on the quality ofevidence used for the development of theguidelines or recommendations, includingspecific cases within the content of the guide-lines [35]. Therefore, a well-designed systematicreview or meta-analysis should be performed toprovide healthcare providers with the bestavailable evidence. In Ethiopia, even thoughthe rate of publication of reviews has risen inthe last decade, their methodologic quality hasbeen doubtful. This could be related to con-ducting reviews without registering a study

protocol, publishing in low impact journals,absence of funding, introducing bias risks in thestudies, or using weak study designs [36]. In thisoverview, three-quarters of AMR reviews hadlow quality or less, and none had highmethodologic quality. Consistent with ourfinding, an overview in the country, not limitedto AMR, found that three-quarters of thereviews had poor methodologic quality [37].Similarly, another overview assessing reviews ofantibiotic use among livestock showed that85% had critically low quality [38]. The majorweaknesses of reviews included in our overviewwere related to the assessment of sources offunding for the original studies, discussing therisk of bias in the individual studies, studyselection in duplicate, data extraction in dupli-cate, and adequate investigation of publicationbias. The aforementioned overviews also repor-ted that none of the reviews reported sources offunding for the primary studies [37, 38].Therefore, we might conclude that the pooledestimates reported in the reviews and in ourcountry were susceptible to methodologic flaws,which seriously weaken the confidence in usingthe finding for clinical decisions. Good-qualityresearch effects positive changes in patient careand decreases the ongoing threat of antimicro-bial resistance in the country [39]. Therefore,Ethiopian researchers should focus on usingstrong study designs such as a randomizedclinical trials, cohort-based studies, and largemultidisciplinary registered reviews rather thanmainly observational studies [39]. Constraintson Ethiopia’s researchers include the limitednational budget and small number of opengrant opportunities, which could be minimizedby involvement and winning medium and largeinternational and regional grants [36].

While the quality of the SRs and MAs washeterogeneous with most having a low qualityscore or less, the pooled resistance patterns ofthe common antimicrobial agents used for thetreatment of the five infectious syndromes inour health settings showed a high resistancelevel. In this case, the resistance of AMO, AMP,and SXT ranged from 35 to 98.1%. Similarly,high levels of AMR (50–100%) to commonlyused antibiotics (including AMP and SXT) werealso reported in a systemic review in east Africa

462 Infect Dis Ther (2020) 9:451–465

[40]. In addition, macrolides generally and ery-thromycin/clarithromycin in particular are therecommended first-line empirical antibioticsprescribed for community-acquired pneumoniain Ethiopian STGs. Six of the reviewed SRs andMAs (three moderate [20, 25, 26] and threecritically low quality [21, 22, 24]) revealed sig-nificantly high resistance of erythromycin(30–97.6%). According to the FMHACAnational baseline, AMR surveillance reportedthat the resistance of Streptococcus pneumonia,the most common cause of CAP, to ery-thromycin increased from 0% in 1996 to 19.2%in 2000 [9]. In Ethiopia, where clinicians are notsupported by basic bacteriology services toensure that treatment is tailored to the specificcondition, information deduced from suchpooled data is crucial in principle. However, thepoor quality of the majority of reviews hindersthe opportunity to develop and use evidence-based guidelines for recommendations relatedto AMR for empirical therapy, which wouldhelp control AMR, improve patient outcomes,and allow early recognition of treatment failure.

The study had certain limitations: Althoughthe AMSTAR 2 tool is designed to criticallyappraise systematic reviews, both randomizedand non-randomized systemic reviews, ofhealthcare interventions, there are some weak-nesses of the criteria. First, two of the items inthe AMSTAR 2 tool (items 12 and 13) suggestthat the authors should only include random-ized controlled trials with a low risk of bias in ameta-analysis or need to discuss the potentialimpact of including studies with a higher risk ofbias. The tool does not include an elaboration ofthe appropriate features that such a discussionshould include so objective evaluation of thesecriteria is difficult. Second, some items on theAMSTAR 2 checklist are more relevant to thecompleteness of reporting as opposed tomethodologic quality. Specifically, items 1, 8,10, and 16 concern the reporting of the reviewresearch question, details of the studies inclu-ded in the review, funding sources for theincluded studies, and sources of funding for thereview, respectively. Although being clear andcomprehensive is essential, research quality andresearch reporting are separate issues. Third, theeighth AMSTAR 2 checklist is more relevant to

the completeness of reporting as opposed tomethodologic quality. Third, it is also impor-tant to acknowledge that it has yet to be vali-dated [18]. Lastly, given the limited number ofreviews, including only 12 SRs and MAs, a topic-based in-depth analysis was not carried out, andthe results should be interpreted with caution.

CONCLUSION

Our study revealed that standard clinical treat-ment guidelines did not routinely consider thelocal antimicrobial resistance of the five infec-tious syndromes. The antimicrobial resistancereviews published in Ethiopia had poormethodologic quality scores using the AMSTAR2 tool. Therefore, to improve clinical decision-making, researchers in Ethiopia should focus onimproving the quality of AMR reviews ratherthan continuing to publish in quantity. Futureresearch is needed to investigate the strategiesto improve the methodologic and reportingquality of systematic reviews or meta-analyseson AMR in Ethiopia.

ACKNOWLEDGEMENTS

We thank the Food, Medicine, and Health CareAdministration and Control Authority ofEthiopia for providing and allowing the use ofstandardized treatment guidelines for thisstudy.

Funding. No funding or sponsorship wasreceived for this study or publication of thisarticle.

Authorship. All named authors meet theInternational Committee of Medical JournalEditors (ICMJE) criteria for authorship for thisarticle, take responsibility for the integrity ofthe work as a whole, and have given theirapproval for this version to be published.

Disclosures. Balew Arega, Asnake Agunie,Abraham Minda, Amdemeskel Mersha, AlazarSitotaw, Getachew Weldeyohhans, AyeleTeshome, Kelemework Adane, and Getachew

Infect Dis Ther (2020) 9:451–465 463

Mengistu declare that they have no conflict ofinterest.

Compliance with Ethics Guidelines. Thisarticle is based on previously conducted studiesand does not contain any studies with humanparticipants or animals performed by any of theauthors.

Data Availability. All data generated oranalyzed during this study are included in thispublished article/assupplementary informationfiles.

Open Access. This article is licensed under aCreative Commons Attribution-NonCommer-cial 4.0 International License, which permitsany non-commercial use, sharing, adaptation,distribution and reproduction in any mediumor format, as long as you give appropriate creditto the original author(s) and the source, providea link to the Creative Commons licence, andindicate if changes were made. The images orother third party material in this article areincluded in the article’s Creative Commonslicence, unless indicated otherwise in a creditline to the material. If material is not includedin the article’s Creative Commons licence andyour intended use is not permitted by statutoryregulation or exceeds the permitted use, youwill need to obtain permission directly from thecopyright holder. To view a copy of this licence,visit http://creativecommons.org/licenses/by-nc/4.0/.

REFERENCES

1. Cosgrove SE. The relationship between antimicro-bial resistance and patient outcomes: mortality,length of hospital stay, and health care costs. ClinInfect Dis. 2006;42(2):S82.

2. Ayukekbong JA, Ntemgwa M, Atabe AN. The threatof antimicrobial resistance in developing countries:causes and control strategies. Antimicrob ResistInfect Control. 2017;6:47.

3. Aseffa A, Abate E, Seyoum B, et al. Investigations onantimicrobial resistance: contributions of a clinicalresearch network. Ethiop Med J. 2019;1:3–7.

4. Yusuf E, Ong DS, Martin-Quiros A, et al. A largesurvey among European trainees in clinical micro-biology and infectious disease on training systemsand training adequacy: identifying the gaps andsuggesting improvements. Eur J Clin MicrobiolInfect Dis. 2017;36:233–42.

5. Elias C, Moja L, Mertz D, et al. Guideline recom-mendations and antimicrobial resistance: the needfor a change. BMJ Open. 2017;7:e016264.

6. WHO. WHO handbook for guideline development.2012.

7. Prorok JC, Iserman EC, Wilczynski NL, et al. Thequality, breadth, and timeliness of content updat-ing vary substantially for 10 online medical texts:an analytic survey. J Clin Epidemiol. 2012;65:1289–95.

8. Hazim C, Abubeker Ibrahim R, Westercamp M,et al. Establishment of a sentinel laboratory-basedantimicrobial resistance surveillance network inEthiopia. Health Secur. 2018;16(S1):S30–6. https://doi.org/10.1089/hs.2018.0052.

9. Ibrahim RA, Amete M. Antimicrobial resistancesurveillance in Ethiopia: implementation experi-ences and lessons learned and Wubshet M. Kefi-choAfr. J Lab Med. 2019;06;8(1):1109.

10. Food, Medicine, and Health Care Administrationand Control Authority. Standard TreatmentGuideline for health centers. Third Edition, 2010

11. Ethiopian Food and Drug Authority. Available:http://www.fmhaca.gov.et/covid19 publications.Accessed 15 Oct 2019.

12. Food, Medicine, and Health Care Administrationand Control Authority. Standard treatment guide-line for health cenetrs. Addis Ababa: Food, Medi-cine, and Health Care Administration and ControlAuthority; 2014.

13. Food, Medicine, and Health Care Administrationand Control Authority. Standard treatment guide-lines for primary hospitals. Addis Ababa: Food,Medicine, and Health Care Administration andControl Authority; 2010.

14. Food, Medicine, and Health Care Administrationand Control Authority. Standard treatment guide-lines for primary hospitals. Addis Ababa: Food,Medicine, and Health Care Administration andControl Authority; 2014.

15. Food, Medicine, and Health care Administrationand control authority. Standard treatment guideli-nes for general hospitals. Third Edition, 2010.

464 Infect Dis Ther (2020) 9:451–465

16. Food, Medicine, and Health care Administrationand control authority. Standard treatment guideli-nes for general hospitals. Third Edition, 2014.

17. Moher D, Liberati A, Tetzlaff J, Altman DG, thePRISMA Group. Preferred reporting items for sys-tematic reviews and meta-analyses: the Prismastatement. PLoS Med. 2009;6(6):e1000097.

18. Shea BJ, Reeves BC, Wells G, et al. AMSTAR 2: acritical appraisal tool for systematic reviews thatinclude randomized or non-randomized studies ofhealthcare interventions, or both. BMJ. 2017;21:358:j4008.

19. Tadesse G. A meta-analysis of the proportion ofantimicrobial resistant human Salmonella isolatesin Ethiopia. BMC Pharmacol Toxicol. 2014;15:51.

20. Deyno S, SintayehuFekadu S, Seyfe S. Prevalenceand antimicrobial resistance of coagulase-negativestaphylococci clinical isolates from Ethiopia: ameta-analysis. BMC Microbiol. 2018;18:43.

21. Eshetie S, Tarekegn F, Moges F. Methicillin-resistantStaphylococcus aureus in Ethiopia: a meta-analysis.BMC Infect Dis. 2016;16:689.

22. Hussen S, Mulatu G, Kassa ZY. Prevalence of Shi-gella species and its drug resistance pattern inEthiopia: a systematic review and meta-analysis.Ann Clin Microbiol Antimicrob. 2019;18:22.

23. Sisay M, Weldegebrea F, Tesfa T. Resistance profileof clinically relevant bacterial isolates against fluo-roquinolone in Ethiopia: a systematic review andmeta-analysis. BMC Pharmacol Toxicol. 2018;19:86.

24. Sisay M, Worku T, Edessa D. Microbial epidemiol-ogy and antimicrobial resistance patterns of woundinfection in Ethiopia: a meta-analysis of laboratory-based cross-sectional studies. BMC Pharmacol Tox-icol. 2019;20:35.

25. Tuem KB, Gebre AK, Atey TM. Drug resistance pat-terns of Escherichia coli in Ethiopia: a meta-analysis.BioMed Research International 2018. https://doi.org/10.1155/2018/4536905.

26. Deyno S, Fekadu S, Astatkie A. Resistance of Sta-phylococcus aureus to antimicrobial agents inEthiopia: a meta-analysis. Antimicrob Resist InfectControl. 2017;6:85.

27. Muhie OA. Antibiotic use and resistance pattern inEthiopia: systematic review and meta-analysis. Int JMicrobiol. 2019. https://doi.org/10.1155/2019/2489063.

28. Reta A, Mengist A, Tesfahun A. Nasal colonizationof methicillin-resistant Staphylococcus aureus inEthiopia: a systematic review and meta-analysis.Ann Clin Microbiol Antimicrob. 2019;18:25.

29. Reta A ,Kifilie AB, Mengist A. Bacterial infectionsand their antibiotic resistance pattern in Ethiopia: asystematic review Hindawi. Adv Prevent Med. 2019.https://doi.org/10.1155/2019/4380309.

30. Beyene G, Asrat D, Mengistu Y. Typhoid fever inEthiopia. J Infect Dev Ctries. 2008;2(6):448–53.

31. WHO. Report on surveillance of antibiotic con-sumption. 2016–2018.

32. Federal democratic republic of Ethiopia ministry ofhealth. Guideline for implementation of a patientreferral system. May 2010.

33. WHO. Implementation of the global action plan onantimicrobial resistance. WHO gap AMR;2017.

34. Gebretekle GB, Haile Mariam D, Abebe W, et al.Opportunities and barriers to implementingantibiotic stewardship in low and middle-incomecountries: lessons from a mixed-methods study in atertiary care hospital in Ethiopia. PLoS ONE.2018;13(12):e0208447.

35. American a cademy of pediatrics.classifying rec-ommendations for clinical practice guidelines.Pediatrics 2004;114(3):874–6.

36. Tesfa B. Challenges of Ethiopian researchers andpotential mitigation with Ethiopian science, tech-nology, and innovation policy 2015; Corpus ID:14798318:1–17.

37. Habtewold TD, Alemu SM, Mohammed SH, et al.Reviews and meta-analyses in Ethiopia had poormethodological quality: an overview of evidence1970 to 2018. J Clin Epidemiol. 2019;109:90–8.

38. Vriezen R, Sargeant JM, Vriezen E, Winder CB,O’Connor AM. Quality assessment of systematicreviews and meta-analyses that examine preventiveantibiotic uses and management practices designedto prevent disease in livestock. Anim Health ResRev. 2019;20:305–18.

39. Gaym A. Health research in Ethiopia—past, present,and suggestions on the way forward. Ethiop Med J.2008;46(3):287–308.

40. Ampaire L, Muhindo A, Orikiriza P, et al. A reviewof antimicrobial resistance in East Africa. Afr J LabMed. 2016;5(1):1–6.

Infect Dis Ther (2020) 9:451–465 465