Worku et al.

Ann Clin Microbiol Antimicrob (2023) 22:96 Annals of Clinical Microbiology

https://doi.org/10.1186/s12941-023-00643-6
and Antimicrobials

RESEARCH Open Access

Bacterial profile of surgical site infection

and antimicrobial resistance patterns
in Ethiopia: a multicentre prospective
cross‑sectional study
Seble Worku1,2,3* , Tamrat Abebe3, Ashenafi Alemu2, Berhanu Seyoum2, Göte Swedberg4,
Alemseged Abdissa2, Adane Mihret2,3 and Getachew Tesfaye Beyene2

Abstract
Background Globally, surgical site infections (SSI) are the most commonly reported healthcare-associated infections.
Methods A multicentre study was conducted among patients who underwent surgical procedures at four hospi-
tals located in Northern (Debre Tabor), Southern (Hawassa), Southwest (Jimma), and Central (Tikur Anbessa) parts
of Ethiopia. A total of 752 patients clinically studied for surgical site infection were enrolled. The number of patients
from Debre Tabor, Hawassa, Jimma, and Tikur Anbessa, hospitals was 172, 184, 193, and 203, respectively. At each
study site, SSI discharge culture was performed from all patients, and positive cultures were characterized by colony
characteristics, Gram stain, and conventional biochemical tests. Each bacterial species was confirmed using Matrix-
Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry (MALDI TOF). An antimicrobial susceptibility
test (AST) was done on Mueller–Hinton agar using the disk diffusion method. Logistic regression analysis was used
to assess associations of dependent and independent variables. A p-value < 0.05 was considered statistically signifi-
cant. Data were analysed using STATA 16 software.
Results Among 752 wound discharge cultures performed, 65.5% yielded growth. Among these, 57.9% and 42.1%
were Gram-negative and Gram-positive isolates, respectively. In this study, a total of 494 bacteria were isolated; Staph-
ylococcus aureus (31%), Escherichia coli (20.7%), and Klebsiella pneumoniae (9.8%) were the most common. Rare isolates
(0.8% each) included Raoultella ornithinolytica, Stenotrophomonas maltophilia, Alcalignes faecalis, Pantoea ecurina, Bacil-
lus flexus, and Paenibacillus tylopili. Enterobacteriaceae showed high levels of resistance to most of the tested antibiot-
ics but lower levels of ertapenem (32.9%), amikacin (24.3%), imipenem (20.3%), and meropenem (17.6%) resistance.
Multidrug-resistant (MDR) frequency of Enterobacteriaceae at Debre Tabor, Hawassa, Jimma, and Tikur Anbessa hospi-
tals was 84.5%, 96.5%, 97.3%, and 94%, respectively. Ages ≥ 61 years (AOR = 2.83, 95% CI: 1.02–7.99; P 0.046), prolonged
duration of hospital stay (AOR = 4.15, 95% CI: 2.87–6.01; P 0.000), history of previous antibiotics use (AOR = 2.83, 95%
CI: 1.06–2.80; P 0.028), history of smoking (AOR = 2.35, 95% CI: 1.44–3.83; P 0.001), emergency surgery (AOR = 2.65, 95%
CI: 1.92–3.66; P 0.000), and duration of operation (AOR = 0.27, 95% CI: 0.181–0.392; P 0.000) were significant risk factors.
Conclusion The most prevalent isolates from Gram-positive and Gram-negative bacteria across all hospitals were
S. aureus and E. coli, respectively. Many newly emerging Gram-negative and Gram-positive bacteria were identified.

*Correspondence:
Seble Worku
[email protected]
Full list of author information is available at the end of the article

© The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​
mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 2 of 18

Variation between hospitals was found for both SSI etiology type and MDR frequencies. Hence, to prevent the emer-
gence and spread of MDR bacteria, standard bacteriological tests and their AST are indispensable for effective antimi-
crobial stewardship.
Keywords Surgical site infections etiologies, Emerging pathogens, Multidrug resistance, Risk factors, Ethiopia

Introduction (AST) or solid epidemiological data from ongoing noso-

Surgical site infection (SSI) is the major costliest health- comial infection surveillance is needed to minimize the
care-associated infection and a substantial cause of mor- problem [20]. In developing countries, including Ethiopia,
bidity and mortality throughout the world [1, 2]. It occurs published reports on bacterial pathogens and their anti-
near or at the incision site and/or deeper underlying tissue biotics resistance patterns of frequently causing SSIs are
spaces and organs within 30 days of a surgical procedure relatively scarce [21] compared to the developed parts of
performed (or up to 90 days for implanted prosthetics) [3]. the world. Besides, virtually all earlier reports depend on
In low and middle-income countries SSI ranked the most phenotypic laboratory methods to characterize pathogenic
frequently reported case of nosocomial infections [4], bacteria and studies were done at single sites with small
and in some settings, up to one-third of patients who are sample sizes [9, 22, 23]. A recent systematic review and
operated on [5] can catch SSI, despite standard protocols meta-analyses study by Birhanu Y et al. [24] focused on
of preoperative preparation and antibiotic prophylaxis are the pooled prevalence of SSI and its aetiology in Ethiopia.
practiced [6]. The SSI rate in Ethiopia has been reported Also, the study had limitations, the papers included used
to be between 14.8 and 20% [5, 7–9], and surgical patients only phenotypic laboratory methods and the result did not
account for 38% of general surgical wards at various teach- display AMR data. Thus, there have always been ambigui-
ing hospitals [10]. It results from mostly bacterial contami- ties in the interpretation of the findings when using phe-
nation during or after the surgical procedure but only a notypic bacterial identification methods. In this study, to
small portion progresses to clinical infection due to innate avoid the ambiguity in the interpretation of strain identi-
host defences removing contaminants. The contamination fication, we employed Matrix-Assisted Laser Desorption/
that will lead to surgical site infection depends on the dose Ionization Time-of-Flight Mass Spectrometry (MALDI
of bacterial contamination, the virulence, and drug resist- TOF) technique for the confirmation of bacterial isolates.
ance of the bacteria [11]. Most SSI infections are prevent- Furthermore, this is comprehensive study with a large
able [11], however probable development of an infection sample size conducted to determine the bacterial profile of
depends on the age, immunocompromising conditions of surgical site infection and antimicrobial resistance patterns
the host, or the antimicrobial-resistance (AMR) nature of at four major hospitals in Ethiopia.
the infecting microorganisms [12]. The frequency varies
from one hospital to the other and is related to complica-
tions [13]. Patients with SSI are twice as likely to die, 60% Methods
more likely to spend time in an intensive care unit (ICU), Study site and design
and more than five times more likely to be readmitted to A multicentre cross-sectional study was conducted
the hospital after discharge [14]. The most common patho- between July 2020 and August 2021 at four selected
gens associated with surgical wound infections are Staphy- hospitals in Northern, Central, Southern and South-
lococcus aureus, Escherichia coli, Klebsiella spp., Proteus west Ethiopia. The study was conducted in purposively
spp., Citrobacter spp., Acinetobacter spp., Coagulase nega- selected University Teaching Hospitals in Ethiopia,
tive Staphylococcus aureus and pseudomonas aeruginosa namely, Debre Tabor Comprehensive Specialized Hos-
[7, 15]. Beta-lactam antibiotics are the most widely used pital (DTCSH), Hawassa University Teaching Hospital
antibiotics for SSI prophylaxis and therapy; however, 30% (HUTH), Jimma University Teaching Hospital (JUTH),
to 90% of antibiotics are misused or overused [16, 17]. This and Tikur Anbessa Specialized Hospital (TASH) (Fig. 1).
inappropriate overuse increases selection pressure, favour- DTCSH comprehensive specialized hospital provides
ing the emergence of drug-resistant bacteria, making the health service to over 5 million people located in Debre
choice of empirical therapy more difficult and expensive, Tabor town of South Gondar Administrative Zone,
and poses a serious threat to public health, thus increas- Amhara Regional State. Debre Tabor town is about 98 km
ing the global risk of SSI [18, 19]. The condition is more away to the East of Bahir-Dar (the capital of Amhara
serious due to irrational antimicrobial prescriptions and regional state) and about 666 km North of Addis Ababa
un-updated empirical therapy. Hence, the use of data (the capital of Ethiopia). It is the only specialized hospital
from clinical laboratories’ antibiotics susceptibility testing in the south Gondar Zone having over 400 beds.
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 3 of 18

Fig. 1 The map of the geographic locations of the four referral hospitals selected for this study

The hospital provides surgery, pediatrics, emergency, laboratories that perform culture and antimicrobial sen-
maternity, gynecologic/obstetric, and psychiatric, includ- sitivity testing.
ing other departments. In addition, the hospital serves as While three hospitals had established microbiology
a teaching centre for the region. laboratories the DTCSH had started performing bacte-
HUCSH is located in Hawassa city in Southern Ethio- riological culture and antimicrobial susceptibility test-
pia, 280 km from Addis Ababa. HUCSH is one of the ing at the time of this study. Therefore, with the help of
largest health facilities in the Southern part of the coun- Armauer Hansen Research Institute (AHRI), DTCSH
try and provides teaching, public health services and and my home institutions Debre Tabor University we
research activities. It serves more than 20 million people established a bacteriology laboratory, which was used for
locally and in the neighbouring regions. Currently, the wound culture processing and antimicrobial susceptibil-
hospital has over 400 beds and provides patient care to ity testing.
90,200 outpatients, 18,100 hospitalized patients and 1100
emergency cases annually. Patient recruitment and sample size calculation
TASH is the teaching hospital of Addis Ababa Univer- The source population the study participants drawn
sity located in Addis Ababa, the capital of Ethiopia and were all patients with suspected cases of SSI who were
the largest specialized hospital in Ethiopia, with over 700 admitted for elective and emergency surgery. Those who
beds. It is also an institution where specialized clinical developed signs and symptoms of SSI within 30 or 90
services that are not available in other public or private (received implant) days and gave consent and/or assent
institutions are rendered to the whole nation. to participate in the study were enrolled and decision
The TASH has 200 doctors, 379 nurses and 115 other to identify eligible patients as SSI cases were done by
health professionals dedicated to providing health care attending physicians. All age groups were included, but
services. The various departments, faculties and residents patients who had been on antibiotic treatment within
under specialty training in the School of Medicine pro- the preceding ten days, SSI later than 30 days after the
vide patient care in the hospital. operation, or refused to give assent or consent (partici-
In their outpatient and inpatient units, the hospital pate in the study) as well as patients with infected burn
offers a variety of services. They also have microbiology wounds, were excluded from the study. A total of 752
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 4 of 18

clinically diagnosed cases of SSI from different wards Research Institute (AHRI) and Sweden for further char-
were enrolled in the study. The sample size was calcu- acterization. The isolates were transported to Sweden
lated based on a single proportion sample size estima- utilizing a triple packaging method with dry ice, specifi-
tion formula (n = ­Z2 P (1—P) /d2) using a proportion of cally engineered for the safe carriage of Category A and B
20% [25]. As this was a multicenter study, to increasing for Infectious and non-infectious substances, in accord-
the sample size a precision (d) of 0.03 was used, where ance with UN regulation UN3373 for DNA samples and
Z stands for Z statistic with a level of confidence of 95%, UN2814 for bacterial isolates.
and the Z value of 1.96. With a 10% non-response rate,
the total sample size came to 752. A convenient sampling Sample collection and transportation
technique was used to recruit study participants until the Trained personnel collected SSI discharge aseptically
required sample size was achieved, and proportional allo- (pus, pus aspirates, and wound swabs) with sterile syringe
cation was made among different hospitals based on the with needle pus aspirate or sterile cotton-tipped swabs
patient flow across the four study sites. from inside to outside. After cleaning the infected area
with 10% povidone-iodine, a sterile cotton-tipped swab
Operational definitions was placed in the center, and the rolling technique was
Surgical site infection occurs near or at the incision site used. All wound swabs were dipped in modified Stuart’s
and/or deeper underlying tissue spaces and organs within Transport Medium and immediately transported to the
30 days of a surgical procedure performed (or up to bacteriology laboratory for culture and drug susceptibil-
90 days for implanted prosthetics) [3]. ity testing within 1 h. Following that, all collected speci-
Clean wound where no inflammation is encountered mens were processed for the identification of bacteria
and the respiratory, alimentary or genitourinary tracts implicated in SSIs.
were not entered.
Clean contaminated wound is where the respiratory, Biochemical tests
alimentary or genitourinary tracts were entered but with- For the final identification of the isolates, the following
out significant spillage. biochemical tests were performed using colonies from
Contaminated when acute inflammation is encoun- pure cultures. For Gram-negative rods, gas production,
tered, or there is visible contamination of the wound. sugar fermentation, H2S production, indole production,
Dirty wound wound in the presence of pus, where there citrate utilization, lysine decarboxylase production, urea
is a previously perforated hollow viscous or compound/ hydrolysis, and motility tests were used. Gram-positive
open injury more than four hours old [26]. cocci were identified using the results of the Gram reac-
Antibiotic a drug, which is products of fungi or bacte- tion, catalase, coagulase, bacitracin, and optochin tests
ria that kills bacteria or inhibits their growth. Antibiot- [27].
ics are not effective against viruses (also referred to as an
antimicrobial). Bacterial strain confirmation using matrix‑assisted laser
Multidrug resistance (MDR) refers to resistance at least desorption ionization‑time of flight mass spectrometry
one antimicrobial agent in three or more antimicrobial (MALDI‑TOF MS)
classes. The isolates were transported to Sweden utilizing a triple
packaging method with dry ice, specifically engineered
Data collection for the safe carriage of Category A and B Infectious and
Professional nurses who had experience of wound swabs non-infectious substances. All bacteria were re-identi-
sample collection and microbiologists who were working fied and confirmed using MALDI-TOF MS at the Clini-
in the bacteriology laboratory were recruited as data col- cal Microbiology Department of Uppsala University
lectors. Training on socio-demographic and clinical data Hospital Uppsala, Sweden. From fresh cultures, a sin-
collection using structured questionnaires, wound swab gle colony of bacteria was smeared onto a MALDI-TOF
sample collection and sample transportation to bacteriol- plate, and the sample was air-dried. Next, 1 µl of formic
ogy laboratories and culture were given to all data collec- acid was added to each cell and air-dried, and then 1 µl
tors. Wound swab cultures, bacterial identification and of MALDI matrix solution was applied to the cells and
drug susceptibility testing were performed in accordance air-dried before reading. MALDI-TOF identification was
with a standardized laboratory protocol that was uni- automatically scored by the system software between 1
formly applied in all study sites (Fig. 2). The findings of and 3 points. All isolates with scores two and above were
each culture were communicated to attending physicians accepted, and all results below 1.7 and flagged red were
for patient management. All bacterial strains were stored rejected. Samples with scores 1.7–2 and flagged yellow
at − 80 °C and transported to the Armauer Hansen were re-analysed.
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 5 of 18

Fig. 2 Laboratory workflow illustrating through patient recruitment to confirmation of the species of bacterial isolates

Antimicrobial susceptibility testing (AST) Trimethoprim-sulfamethoxazole (1.25/23.75 µg), ampi-

The antibiotics susceptibility tests were performed on cillin-sulbactam (10/10 µg), aztreonam (30 µg), mero-
Muller-Hinton agar (Oxoid) by using the Kirby-Bauer penem (10 µg), Imipenem (10 µg), ertapenem (30 µg).
disk diffusion technique. Using a sterile wire loop, 3–5 Gram-positive isolates were tested against penicillin
pure colonies were transferred to a tube containing 5 mL (10units), ampicillin (10 µg), vancomycin (30 µg), eryth-
of sterile normal saline (0.85% NaCl) and gently mixed romycin (15 µg), ciprofloxacin (5 µg), cefoxitin (30 µg),
until a uniform suspension formed. Standard inoculum clindamycin (30 µg), erythromycin (15 µg), doxycycline
density was adjusted to 0.5 McFarland units. The excess (30 µg), chloramphenicol (30 µg), gentamicin (10 µg), and
broth suspension was removed by tapping against the oxacillin (5 µg), tetracycline (30 µg), [28]. Following that,
tube wall. The bacterial suspension was swabbed on the plates were incubated at 37 °C for 18–24 h. Each zone
MHA surface by using sterile swab then a set of antibiotic of inhibition was measured to the nearest millimeter, and
discs placed with sterile forceps at least 24 mm apart from classified as sensitive, intermediate, or resistant using
one another [28]. All antibiotics discs were OXOID prod- the standard technique [28]. MDR was a bacterium that
ucts (Oxoid Ltd, UK), and susceptibility of Gram-negative was simultaneously resistant at least one drug in three or
isolates was tested against: ampicillin (10 µg), gentamicin more categories.
(10 µg), amikacin (30 µg), ciprofloxacin (5 µg), chloram-
phenicol (30 µg), ceftazidime (30 µg), cefotaxime (30 µg), Quality control
ceftriaxone (30 μg), cefuroxime (30 µg), cefepime (30 µg), All specimens were collected according to the standard
tetracycline (30 µg), amoxycillin + Clavulanate (20/10 μg), operating procedure (SOP) [21]. A double data entry
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 6 of 18

method was used to ensure the accuracy of the data. Table 1 Socio-demographic characteristics and clinical data
The performance of all prepared media was checked by Characteristic Frequency (%)
inoculating control strains, E. coli (ATCC 25922) and
S. aureus (ATCC 25923), for each new batch of agar Hospitals
plates [22]. In addition, the sterility of culture media was DTCSH (n = 172) 172 (22.9)
checked by incubating 5% of the prepared media at 37 °C Growth 135 (78.5)
for 24–48 h. In addition, reagents for Gram-stain and No growth 37 (21.5)
biochemical tests were checked against control strains HUCSH (n = 184) 184 (24.5)
of S. aureus and E. coli. The 0.5 McFarland standard was Growth 120 (65.2)
used to standardize the inoculum density of the bacterial No growth 64 (34.8)
suspension for the susceptibility test. Each MALDI-TOF JUSTH (n = 193) 193 (25.7)
run also included quality control strains using E. coli Growth 126 (65.3)
(ATCC 25922) and S. aureus (ATCC 25923). No growth 67 (34.7)
TASH (n = 203) 203 (27)
Data analysis Growth 113 (55.7)
The data were checked for completeness, missing val- No growth 90 (44.3)
ues, and coding of questionnaires entered into Research Sex
Electronic Data Capture (RED-Cap) and exported to Male 418 (55.6)
STATA version 16.0. Frequencies and cross-tabulations Female 334 (44.4)
were used to summarize descriptive statistics (median, Age in (year)
percentages or frequency). Associations of possible risk ≤ 18 159 (21.1)
factors with SSIs was assessed using bivariate and multi- 19–40 419 (55.7)
variate logistic regression to study the effect of independ- 41–60 130 (17.3)
ent variables on the dependent variables. P-value less ≥ 61 44 (5.9)
than 0.05 were considered statistically significant. Surgical site infection
Superficial 265 (35.2)
Ethical considerations Deep 487 (64.8)
The Department of Medical Microbiology, Immunology, Preoperative hospital stay
and Parasitology (DMIP) and the AHRI/ALERT Research ≤7 298 (39.6)
Ethics Committee (AAREC) reviewed and approved >7 454 (60.4)
the study, and institutional review board (IRB) approval History of hospital admission
was obtained from Addis Ababa University’s College Yes 133 (17.6)
of Health Sciences and AAREC, AAUMF03-008/2020. No 619 (82.4)
Written permission letter was obtained from each study Previous use of antibiotics
site before starting the data collection. The purpose Yes 388 (52.6)
and procedures of the study was explained to the study No 364 (47.4)
participants, participants’ parents or guardians before Alcoholic
recruitment to the study. Those study participants who Yes 67 (8.9)
gave written informed consent and those children whose No 685 (91.1)
parents or guardians gave informed assent were selected Smoking
and enrolled in this study. Results obtained from all Yes 444 (59)
patients were communicated to attending physicians and No 308 (41)
all patient’s information was kept confidentially. Nature of surgery
Elective 246 (32.7)
Results Emergency 506 (67.3)
In the present study, a total of 752 patients from four dif- Type of surgery
ferent hospitals were investigated for SSIs. The number Clean/clean contaminated surgery 724 (96.2)
of patients from DTCSH was172, and the numbers from Contaminated surgery 28 (3.8)
HUCSH, JUSTH, and TASH, were184, 193, and 203, Timing of surgical antimicrobial prophylaxis
respectively (Table 1). Of the 752 study participants whose Before the operation 548 (72.8)
SSI discharge was inoculated onto growth media, 65.5% During the operation 204 (27.2)
(493 /752) showed bacterial growth (Table 1). DTCSH Duration of operation
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 7 of 18

Table 1 (continued) followed by E. coli (20.7%) and Klebsiella pneumo-

Characteristic Frequency (%) nia (9.8%) among SSIs (Fig. 1). Other less frequently
detected species were Acinetobacter baumannii (7.6%),
≤1 h 336 (44.7) Enterobacter cloacae (5.1%), Pseudomonas aeruginosa
>1 h 416 (55.3) (3.7%), Klebsiella variicola, and Enterobacter hormaeche
(1% each). Diverse species of Acinetobacter, Enterobac-
ter, Enterococcus, Staphylococcus, Aerococcus, Bacillus,
had the highest percentage of positive cultures (78.5%), Citrobacter, and Pseudomonas were identified. While
followed by JUTSH (65.3%), and HUCSH and TASH, Gram-positives was found at all four hospitals (42.1%),
respectively, had 65.2% and 55.7% of SSI bacterial growth it was mainly detected at DTCSH (40.4%), with 21.6%,
(Table 1). The study participants age ranged from 3 days 21.6%, and 16.3% isolated at TASH, HUCSH, and JUTSH,
to 85 years with median of 28 years and 418 (55.6%) were respectively (Fig. 2B). In addition, Raoultella ornithino-
males. Approximately 487 (64.8%) of patients had deep SSI, lytica, Stenotrophomonas maltophilia, Pantoea ecurina,
454 (60.4%) preoperative hospital stay > 7 days, 619 (82.4%) Providencia rettgeri, Alcalignes faecalis, and Morganella
history of hospital admission, 388 (52.9%) had previous morganii were detected as rare bacterial pathogens. Fig-
use of antibiotics, 448 (59.6%) had smoking history, 506 ure 2A shows the frequency and distribution of Gram-
(67.2%) of surgical procedures were emergency surgery, negative bacterial isolates at the four hospitals.
724 (96.2%) of patients with clean or clean contaminated
wounds dominated the wound class, 548 (72.8%) required Antibiotic resistance pattern of SSI bacterial isolates
antimicrobial prophylaxis before the procedure, and 55.3% The predominant isolate from Gram-positives, S. aureus,
underwent surgeries lasting greater than an hour (Table 1). revealed a high level of resistance toward penicillin 90.1%,
Bivariate and multivariable logistic regression analyses and ampicillin 76.5%, while 7.8%, 10.6%, and 12.4% of the
were used to see the relationship between the independ- isolates were resistant to clindamycin, chloramphenicol, and
ent variables over the dependent variable. On bivariate gentamicin respectively but 100% of S. aureus were sensitive
regression analysis, male sex, age ≥ 61, SSI type, preopera- to vancomycin (Table 3). All isolates of S. aureus showed
tive hospital stays, history of hospital admission, previous multiple drug resistance (resistance to two or more drugs).
use of antibiotics, smoking, emergency surgery, and dura- Cefoxitin, which is a surrogate marker of methicillin,
tion of operation ≥ 1 h had a statistically significant asso- showed 22.7% resistance against S. aureus. Enterococ-
ciation with the occurrence of SSI. The type of surgery cus species showed 70.4% resistance to ampicillin and
(wound), alcohol history and the timing of prophylactic 66.7% to erythromycin. Table 3 shows the AMR pattern
antibiotics ≥ 1 h had no statistically significant association of Gram-positive bacteria.
(Table 2). The result of the multivariate regression showed The Enterobacteriaceae showed high resistance
that ages ≥ 61 years (AOR = 2.83, 95% CI: 1.02–7.99; P toward ampicillin (93.2%), ceftriaxone (90.5%), cefuro-
0.046), prolonged duration of hospital stay (AOR = 4.15, xime (88.7%), aztreonam (82.9%), ceftazidime (80.6%),
95% CI: 2.87–6.01; P 0.000), history of previous antibiot- cefepime (77%), ampicillin-sulbactam (76.1%), trimeth-
ics use (AOR = 2.83, 95% CI: 1.06–2.80; P 0.028), history of oprim-sulfamethoxazole (77.5%), tetracycline (72.5%),
smoking (AOR = 2.35, 95% CI:1.44–3.83; P 0.001), emer- amoxicillin-clavulanic acid (71.2%) (Table 4).
gency surgery (AOR = 3.24, 95% CI: 2.29–4.77; P 0.000), Low resistance frequency of Enterobacteriaceae was
and duration of operation (AOR = 0.27, 95% CI: 0.181– detected for amikacin (24.3%), imipenem (20.3%), mero-
0.39; P 0.000) were significant risk factors (Table 2). penem (17.6%), and ertapenem (32.9%) (Table 4).
The resistance of Enterobacteriaceae to merope-
Frequency and distribution of identified bacterial isolates nem and imipenem was (11.6%,18.6%), (18%, 22.9%),
The total number of pathogenic bacterial isolates (12.2%,13.5%) and (26.9%,25%) at DTCSH, HUCSH,
were 65.7% (494/752) from all SSI culture (Figs. 3, 4A). JUTSH, and TASH, respectively (Fig. 5A).
Gram-negative were 57.9% (286/752), and Gram posi- The predominant isolate, E. coli (n = 102) revealed a
tive 42.1% (208/752) according to Fig. 2B, C). Of these, high level of resistance to ampicillin (94.6%), ceftriaxone
2.6% (13/493) of cultures were a mixture of two colony (99%), cefotaxime (93.8), ceftazidime (79.4%), cefepime
types, while 2.4% (12/493) were commensals or contami- (77%), cefuroxime (73.5%), ampicillin-sulbactam (72%),
nants and 97.4% showed single bacterial growth. Species trimethoprim-sulfamethoxazole (71.5%), tetracycline
of the mixed cultures were Raoultella ornithinolytica, (70.6%), and low-level resistance to gentamicin (57.8%),
Paenibacillus tylopili, S. aureus and coagulase negative chloramphenicol (41.2%), ertapenem (24.5), imipenem
staphylococci. Among the identified types of bacteria, (11.6%), amikacin (10.8%), meropenem (9.8).
Staphylococcus aureus was the predominant one (31%),
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 8 of 18

Table 2 Bivariate and multivariate analysis to identify factors associated with patient demographic and clinical characteristics with
surgical site infection culture showed growth
Characteristics Bacterial growth n (%) P-value Crude-OR (95%CI) Adjusted- OR (95%CI) P-value
Growth No growth

Sex
Male 309 (41.1) 109 (14.5) 0.194 0.81 (0.6293– 1.112) .94 (0.6420–1.373) 0.75
Female 233 (31) 101 (13.4) 1
Age in (year)
≤ 18 119 (15.8) 40 (5.3) 0.181 1.15 (0.937–1.421) 1.263 (0.7855–0.031) 0.33
19–40 287 (38.2) 132 (17.6) 1
41–60 98 (13) 32 (4.3) 1.260 (0.7489–2.122) 0.38
≥ 61 38 (5.1) 6 (0.8) 2.83 (1.017–7.898) 0.046
SSI
Superficial 183 (24.3) 82 (10.9) 0.149 1.271 (0.918–1.762) 1
Deep 359 (47.7) 128 (17) 0.992 (0.6783–1.451) 0.97
Preoperative hospital stay
≤7 145 (19.3) 124 (16.5) 0.000 3.908 (2.807–5.442) 1
>7 397 (52.8) 86 (11.4) 4.15 (2.87–6.01) 0.000
Previous use of antibiotics
Yes 326 (43.3) 72 (9.6) 2.83 (1.06–2.80) 0.028
No 216 (28.7) 138 (18.4) 0.000 2.641 (1.897–3.673) 1
Alcoholic history
Yes 54 (7.2) 13 (1.7) 0.093 1.711 (0.915–3.201)
No 488 (64.9) 197 (26.2)
Smoking history
Yes 359 (47.7) 89 (11.8) 2.35 (1.44–3.83) 0.001
No 186 (24.7) 118 (15.7) 0.000 2.646 (1.915–3.656) 1
Nature of surgery
Elective 147 (19.5) 99 (13.2) 0.000 2.395 (1.726–3.323) 1
Emergency 395 (52.5) 111 (5.5) 3.24 (2.298–4.77) 0.000
Type of surgery
Clean/Clean contaminated surgery 518 (68.9) 206 (27.4) 0.09 1.578 (0.925–2.691)
Contaminated surgery 24 (3.2) 4 (0.5)
Timing of surgical antimicrobial prophylaxis
Before the operation 150 (19.9) 58 (7.7) 0.984 0.996 (0.698–1.421)
During the operation 392 (52.1) 152 (20.2)
Duration of operation
≤1 h 198 (27.7) 138 (17) 1
>1 h 344 (45.7) 72 (9.6) 0.332 (0.239–0.461) 0.266 (0.181–0.392) 0.000

K. pneumoniae (n = 48) were resistant to ampicillin gentamicin (86.8%), cefepime (84.2%), meropenem
(100%), ceftriaxone (100%), cefotaxime (93.8%), amoxicil- (84.2%), and SXT (81.4%). In addition, A. baumannii has
lin-clavulanic acid (91.7%), ceftazidime (88.5%), cefepime lower-level resistance to imipenem (65.9%) and ampi-
(81.2%), cefuroxime (77.1%), tetracycline (66.7%), ertap- cillin-sulbactam (63.1%) (Table 4). The resistance fre-
enem (43.8%), meropenem (41.7%), amikacin (33.3%), quency of Acinetobacter species to meropenem at DRH,
imipenem (29.2%). Amikacin and meropenem were 100% HUCSH, JUSTH, and TASH was 75%, 83.3%, 42.8%, and
effective against all of the isolates of Klebsiella variicola 46.2%, respectively (Fig. 3B). P. aeruginosa showed mini-
and Proteus mirabilis. In the non-fermenter group, A. mal resistance to ceftazidime (66.7%), cefepime (55.5%),
baumannii showed the highest resistance to cefotax- gentamicin (47.8%), ciprofloxacin (22.2%), and amika-
ime (95.3%), ertapenem (92.1%), ceftazidime (89.5%), cin (10.5%) (Table 4). In addition, 100% and 94.5% of
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 9 of 18

(Pathogenic bactrial isolates (n=494)

GNB=57.9%,GPB=42.1%

Stapyloccous aureus (n=153)

Other Gram negative and positve

(n=128)
Escherichia coli (n=102)

Klebsiella pneumoniae (n=48)

Acinetobacter baumanni (n=38)

Shigella dysenteriae (n=25)

Fig. 3 Frequency and distribution of bacteria isolated from patients investigated for surgical site infection at four different hospitals in Ethiopia.
GNB Gram-negative bacteria, GPB Gram-positive bacteria, Other GN and GP (n = 128): Raoultella ornithinolytica (n = 1), Stenotrophomonas maltophilia
(n = 1), Acinetobacter soli (n = 2), Acinetobacter pitti (n = 2), Acinetobacter lactucae (n = 1), Pseudomonas plecoglossicida (n = 1), Pantoea ecurina
(n = 1), Citrobacter freundii (n = 1), Citrobacter sedlakii (n = 2), Providencia rettgeri (n = 2), Alcalignes faecalis (n = 2), Proteus mirabilis (n = 4), Morganella
morganii (n = 1), Aerococcus viridans (n = 3), Bacillus flexus(n = 1) Paenibacillus tylopili (n = 1), Enterobacter cloacae (n = 19), Enterobacter asburiae (n = 1),
Enterobacter bugandensis (n = 4), Enterobacter hormaeche (n = 5), Enterococcus faecium (n = 8): Enterococcus gallinarum (n = 3), Enterococcus hirae (n = 2),
Enterococcus durans (n = 2), Staphylococcus hominis (n = 4), Staphylococcus haemolyticus (n = 3), Staphylococcus warneri (n = 2), Staphylococcus sciuri
(n = 6), S. epidermidis (n = 8)

Pseudomonas species were sensitive to meropenem and other hand, MDR frequency for Pseudomonas species
imipenem, respectively (Table 4, Fig. 5C). was 66.7%, 66.7%, 83.3% and 50% at DTCSH, HUCSH,
and JUSTH and TASH, respectively (Fig. 6C).
Multidrug resistance
The overall Multidrug resistance (MDR) to three or more Discussion
antibiotics was observed in 100% of S. aureus (Table 3) In surgically treated patients, post-operative SSI is still
and 93.3% Enterobacteriaceae (Table 4). Enterobacte- one of the leading causes of morbidity and mortality,
riaceae that showed MDR to eight (R-9), nine (R-10) and it increases the cost of health care due to repeated
and ten (R- ≥ 11) antibiotics from different groups had readmission. In order to effectively manage bacterial
a frequency of 6.3%, 6.7%, and 64.9%, respectively. Only infection, it is crucial to identify the bacterial patho-
0.5% Enterobacteriaceae showed zero resistance (R-0) to gens and choose an antibiotic that is efficient against
all antibiotic classes tested, whereas 3.1% Enterobacte- the organism [1, 2].
riaceae showed resistance to one antibiotic (R-1) class. As far as risk factors associated with the occurrence
For Enterobacteriaceae, the MDR frequency at DTCSH, of SSI are concerned, in the present study, the likeli-
HUCSH, JUSTH, and TASH was 84.5%, 96.5%, 97.3%, hood of SSI occurrences among patients aged ≥ 61 years
and 94%, respectively (Fig. 6A). E. coli, K. pneumoniae, increased by a factor of 2.8. Similar findings have been
E. cloacae, S. dysenteriae, K. variicola, and P. mirabi- conducted in Ethiopia [15, 16, 29] and elsewhere [30].
lis showed an overall MDR frequency of 96.1%, 95.9%, This might be due to a weakened immune response to
79.3%, 82%, 100%, and 100%, respectively. The overall infectious agents and poor nutritional status [31].
MDR frequency of A. baumannii and P. aeruginosa was Patients who had a longer duration of hospital stay
95% and 77.8%, respectively. The MDR frequency for Aci- developed SSIs 4.1 times more frequently (P = 0.000)
netobacter species was 73% at DTCSH, 83.3% at HUCSH, than those who had shorter time. This finding was in
100% at JUTSH, and 100% at TASH (Fig. 4B). On the agreement with many studies in Ethiopia [29, 32, 33]
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 10 of 18

Percentage at each Hospital

Hospitals
A DTCSH(n=135)
Stapyloccous aureus (n=153)
Escherichia coli (n=102) HUCSH (n=120)
Klebsiella pneumonia ( n=48) JUTSH(n=126)
Acinetobacter baumammi(n=38)
Shigella dysenteriae(n=25) TASH (n=113)
Enterobacter cloacae(n=19)
Pseudomonas aeruginosa (n=18)
Entrococcus faecalis(n=12)
Enterococcus faecium(n=8)
Staphylococcus epidermidis(n=8)
Staphylococcus sciuri(n=6)
Enterobacter hormaechei(n=5)
Klebsiella variicola (n=5)
Enterobacter bugandensis(n=4)
Proteus mirabilis(n=4)
Staphylococcus hominis(n=4)
Aerococcus viridans(n=3)
bactrial Isolates

Enterococcus gallinarum(n=3)
Staphylococcus haemlytics(n=3)
Enterococcus hirae(n=2)
Alcalignes faecalis(n=2)
Citrobacter sedlakii(n=2)
Providencia rettgeri(n=2)
Stapyloccous warneri(n=2)
Enterococcus durans(n=2)
Acinetobacter soli(n=2)
Acinetobacter pitti(n=2)
Acinetobacter lactucae(n=1)
Stenotrophomonas maltophilia(n=1)
Raoultella ornithinolytica(n=1)
Pseudomonas plecoglossicida (n=1)
Pantoea ecurina(n=1)
Citrobacter freundii(n=1)
Morganella morganii(n=1)
Bacillus flexus(n=1)
Paenibacillus tylopili(n=1)
Enterobacter asburiae(n=1)
Total bactrial growth at each hospital%

Hospitals
DTCSH(n=51)
B Escherichia coli (n=102)
HUCSH (n=75)

Klebsiella pneumonia ( n=48) JUTSH(n=92)

Acinetobacter baumammi(n=38) TASH (n=68)
Shigella dysenteriae(n=25)
Enterobacter cloacae(n=19)
Pseudomonas aeruginosa (n=18)
Enterobacter hormaechei(n=5)
Klebsiella variicola (n=5)
Gram-negative isolates

Enterobacter bugandensis(n=4)
Proteus mirabilis(n=4)
Alcalignes faecalis(n=2)
Citrobacter sedlakii(n=2)
Providencia rettgeri(n=2)
Acinetobacter soli(n=2)
Acinetobacter pitti(n=2)
Acinetobacter lactucae(n=1)
Stenotrophomonas maltophilia(n=1)
Raoultella ornithinolytica(n=1)
Pseudomonas plecoglossicida (n=1)
Pantoea ecurina(n=1)
Citrobacter freundii(n=1)
Morganella morganii(n=1)
Enterobacter asburiae(n=1)
Percentage
Fig. 4 Frequency and distribution of bacterial isolates from the total number of bacteria isolated at each hospital A total identified bacteria at each
site, B Gram-negative isolates and C Gram-positive isolates. DTCSH Debre Tabor Comprehensive Specialized Hospital, HUCSH Hawassa University
Comprehensive Specialized Hospital, JUTSH Jimma University Teaching Specialized Hospital, TASH Tikur Anbessa Specialized Hospital, n number
of bacterial isolates
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 11 of 18

C Hospitals
DTCSH(n=84)
Stapyloccous aureus (n=153) HUCSH (n=75)
Entrococcus faecalis(n=12) JUTSH(n=92)
Enterococcus faecium(n=8) TASH (n=68)
Staphylococcus epidermidis(n=8)
Gram-positive isolates

Staphylococcus sciuri(n=6)
Staphylococcus hominis(n=4)
Aerococcus viridans(n=3)
Enterococcus gallinarum(n=3)
Staphylococcus haemlytics(n=3)
Enterococcus hirae(n=2)
Stapyloccous warneri(n=2)
Enterococcus durans(n=2)
Acinetobacter pitti(n=2)
Bacillus flexus(n=1)
Paenibacillus tylopili(n=1)
Percentage
Fig. 4 continued

Table 3 Antimicrobial resistance pattern of Gram-positive bacteria isolated from patients diagnosed with surgical site infection in
Ethiopia: a multicenter prospective cross-sectional study 2022
Gram-positive isolates (%) of resistance to antimicrobial agents
P AMP E TE FOX OXA DOX SXT CPR CN CHL DC V MDR%

S. aureus (n = 153) 90.2 76.5 43.7 43 20 20 23.8 24.2 20 12.4 10.6 7.8 0 100
Other staphylococcus spp.
S. epidermidis (n = 8) 75 75 37.5 62.5 12.5 12.5 25 25 25 25 12.5 12.5 0 73.9
Staphylococcus sciuri (n = 6) 100 100 50 50 100 100 50 50 33 33 33 25 0
Staphylococcus hominins (n = 4) 75 75 75 50 25 25 50 50 0 75 25 25 0
Staphylococcus haemolyticus (n = 3) 100 67 67 67 33 33 67 33 0 67 67 67 33
Staphylococcus warneri (n = 2) 100 100 100 50 50 50 50 50 50 50 50 NA 0
Total (n = 176) 88.6 77.3 45.5 44.9 22.7 22.7 26.7 24.4 20.5 16.5 17.6 9.7 0 96.6
Entrococcus faecalis (n = 12) NA 66.7 66.7 75 NA NA NA NA NA NA 66.7 NA NA
Enterococcus faecium (n = 8): NA 75 100 75 NA NA NA NA NA NA 75 NA NA
Enterococcus gallinarum (n = 3) NA 100 NA 67 NA NA NA 33 NA NA 67 NA NA
Enterococcus hirae (n = 2) NA 50 50 50 NA NA NA NA NA NA 50 NA NA
Enterococcus durans (n = 2) NA 100 50 50 NA NA NA NA NA NA 50 NA NA
Total (n = 27) NA 70.4 66.7 59.3 NA NA NA NA NA NA 66.7 NA NA
Other gram positives
Aerococcus viridans (n = 3) 100 100 67 33 NA NA NA 67 100 67 67 NA NA
Bacillus flexus (n = 1) NA 100 100 100 NA NA NA 0 100 0 0 NA NA
Paenibacillus tylopili (n = 1) NA 100 100 100 NA NA NA 100 100 0 0 NA NA
Total (n = 5) 20 60 80 60 NA NA NA 80 100 40 40 NA NA
P Pencillin, AMP Ampicillin, E Erytromycin, TE Tetracycline, FOX: Cefoxitin, OXA Oxacillin, DOX Doxycycline, SXT Trimethoprim-Sulfamethoxazole, CPR Ciprofloxacin, CN
Gentamicin, CHL Chloramphenicol, DC Clindamycin, V Vancomycin, NA not applicable, MDR Multidrug resistance

and elsewhere [30]. This is a notable finding because it higher chance of developing SSI than with non-pre-
is associated with additional costs in a country with a vious use of antibiotics [35]. This could be because
staggering economy and healthcare system [34]. broad-spectrum antibiotics have a high risk of causing
Similarly, the present study demonstrated that with superinfection of resistant strains due to selective pres-
previous use of antibiotics, patients had a 2.8 times sure [18, 19].
Table 4 Antibiotics resistance patterns of Gram-negative bacteria isolates among patients with surgical site infection in Ethiopia: a multicenter prospective cross-sectional study
2022
Worku et al. Ann Clin Microbiol Antimicrob

Gram-negative isolates Resistance to antimicrobial agents (%)

AMP AMC CHL CRO SXT CN AK CXM CTX CPR CAZ FEP IMP MEM ET TE SAM ATM MDR%

E. coli (n = 102) 94.1 68.6 41.2 99 71.5 57.8 10.8 73.5 89.2 73.5 79.4 76.5 11.8 9.8 24.5 70.6 72 64 96.1
K. pneumonia (n = 48) 100 91.7 45.8 93.8 64.6 70.8 33.3 77.1 93.8 62.5 85.5 81.25 29.2 41.7 43.8 66.7 85.2 96.8 95..9
(2023) 22:96

E. cloacae complex (n = 29) 96.6 79.3 62.1 89.7 75.7 58.6 24.1 72.4 86.2 55.2 79.3 79.3 37.9 17.2 31 62.1 71.4 63.1 79.3
S. disentriae (n = 25) 100 28 32 92 64 28 16 84 44 76 76 20 20 20 48 66.9 66.5 82
K. variicola (n = 5) 100 40 40 100 60 60 0 60 0 40 100 80 40 0 20 60 71.9 62.2 100
P. mirabilis (n = 4) 100 25 25 75 50 50 0 75 25 50 50 75 0 0 0 50 66.1 65.2 100
Rare Enterobacteriaceae isolates (n = 9) 88.8 88.8 44.4 77.8 44.4 66.7 22.2 77.8 77.8 66.7 77.8 66.7 33.3 11.1 22.2 55.5 45.1 38.7 100
Total Enterobacteriaceae isolates (n = 222) 93.2 71.2 62.2 90.5 77.5 60.1 24.3 88.7 90.5 81.1 80.6 77 20.3 17.6 32.9 72.5 76.1 82.9 93.3
Acinetobacter species
A. baumammi (n = 38) – – – 94.7 81.4 86.8 42.1 – 95.3 73.7 89.5 84.2 65.9 84.2 92.1 – 63.1 – 95
Other Acinetobacter species (n = 5) – – – 95.3 20 60 0 – 95.3 40 60 60 20 40 100 – 40 – 60
Total Acinetobacter species (n = 43) – – 95.3 86 13.5 39.5 – 95.3 69.7 81.4 86 67.4 14.9 90.7 – 67.4 – 93
Pseudomonas species
P. aeruginosa (n = 18) – – – – – 47.8 27.8 – – 22.2 66.7 55.5 5.5 0 38.9 – – 66.7 77.8
P. plecoglossicida (n = 1) – – – – – 0 0 – – 100 100 100 0 0 100 – – 100 100
Total pseudomonas species (n = 19) – – – – – 47.3 21.1 – – 26.3 68.4 57.9 5.3 0 42.1 – – 68.4 73.8
AMP Ampicillin, AK Amikacin, AC Amoxicillin-Clavulanic Acid, ATM Aztreonam, CHL Chloramphenicol, CN Gentamicin, CRO Ceftriaxone, FEP Cefepime SXT Trimethoprim-Sulfamethoxazole, CPR Ciprofloxacin, CXM
Cefuroxime, CTX Cefotaxime, ET Ertapenem, IPM Imipenem, MEM Meropenem, TE Tetracycline
Page 12 of 18
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 13 of 18

The type of surgery was also statistically associated Rare surgical site infections isolate including Raoultella
with SSI in the present study. Undergoing emergency ornithinolytica, Stenotrophomonas maltophilia, Alcal-
surgery showed approximately 3.24 times higher chances ignes faecalis, and Paenibacillus tylopili, were identified.
of acquiring SSIs when compared to elective surgery Such newly emerging bacteria-causing infections in SSI
(P = 0.000), which complies with related studies [35]. patients may result in future challenges [46–48]. These
The risk of developing SSI with smoking histories was species, along with K. variicola and Pantoea ecurina,
found to be 2.35 times more than in those who did not have never before been identified in patients in Ethiopia
have smoking history. There was a significant association who were being evaluated for SSI. The isolation of these
between smoking patients’ history and SSI (P = 0.001). new SSI etiologies emphasizes the necessity of institu-
This finding was in agreement with smoking history were tion-based diagnostic and intervention practices.
independent predictors of SSIs in multivariate logistic In our finding, S. aureus revealed a high level of resist-
regression analysis [36]. Smoking weakens immunity and ance to penicillin (88.6%) and ampicillin (77.3%), which
increases the risk of SSI [37]. was comparable to study in Turkey [46]. On the other
The overall culture positivity rate from patients with SSI hand, all isolated S. aureus were susceptible to vanco-
in the current study was 65.5%. Similar results reported mycin, and our finding was the same as earlier studies in
from India (68%) [38]. Which was slightly smaller than Jimma [36] and Turkey [46].
results previously reported from Jimma (71.7%) [9]. In Gram-negative bacteria showed higher resistance
contrast, the current study was lower than reports from to ampicillin (93.2%), ceftriaxone (90.5%), cefuro-
Tikur Anbessa (75.6%) [39], Gondar (83.9%) [40], and xime (88.7%), aztreonam (82.9%), ceftazidime (80.6%),
elsewhere (82%) [41]. Lower rates of positive culture were cefepime (77%), ampicillin-sulbactam (76.1%), trimeth-
reported from India Bangladesh (61.8%) [42]. oprim-sulfamethoxazole (77.5%), tetracycline (72.5%),
In this study, the Gram-negative bacterial isolation and amoxicillin-clavulanic acid (71.2%). Since β-lactam
rate was greater (57.9%) than the Gram-positive isolates antibiotics are the most widely used antibiotics, many
(42.1%), which is comparable with the study done in Addis studies in Ethiopia [11, 34] and around the world have
Ababa [43]. On the other hand, the current study was lower found similar resistant patterns [40]. This shows antibiot-
in Gram- negative isolates than reports from Mizan-Tepi ics require a periodic evaluation and the establishment of
(73.2%) and higher in Gram-positive isolates (24.8%) [22]. antibiotic policies for prophylaxis and treatment guide-
This could be due to a difference in the study population. lines in the Ethiopian setting.
In the current study, the profiles of bacterial isolates In this study, multidrug resistance (MDR) was observed
highly associated with SSI were S. aureus (31%), fol- (100%) in S. aureus and (93.4%) in Gram-negative bacte-
lowed by E. coli (20.7%), and K. pneumoniae (8.9%). ria, similar to findings from Bahir Dar [18].
Studies from Gondar [44], Addis Ababa [45], and India In the current study, A. baumannii showed the high-
[40] reported similar results. The high prevalence of S. est to resistance cefotaxime (95.3%), ceftazidime (89.5%),
aureus infection could be due to an endogenous source gentamicin (86.8%), cefepime (84.2%), and trimethoprim-
as well as environmental contamination. sulfamethoxazole (81.4%). Many studies have found
In contrast to previous reports, a study from Addis that these organisms have a high resistance to the most
Ababa found E. coli 23.1%, followed by multidrug-resist- commonly used antibiotics [34, 47, 48]. In addition, A.
ant Acinetobacter species 22.1% [34]. Similarly, in a study baumannii also showed remarkably high resistance to
done by Jimma, E. coli was frequently identified and fol- ertapenem (92.1%), meropenem (84.2%), and imipenem
lowed by Klebsiella spp. [41]. This variation in the dis- (65.9%).
tribution pattern of bacterial isolates may be due to the Amikacin and meropenem were 100% effective against
diversity of the study population, setting, and local anti- all of the isolates of P. mirabilis and K. variicola in our
microbial usage pattern, which leads to the introduc- study. However, studies done in Mekelle [37], and
tion of pathogens that may be resistant to currently used elsewhere [29] showed that ciprofloxacin were effec-
antibiotics. tive against Proteus and Pseudomonas isolates. The

(See figure on next page.)

Fig. 5 Frequency of antibiotic resistance at four hospitals; A Enterobacteriaceae B Acinetobacter species C Pseudomonas species. The percentage
represents the numbers of resistant isolates, out of the total number of isolates at all hospitals. AMP ampicillin, AMC amoxicillin/clavulanate,
AK amikacin, SXT trimethoprim-sulfamethoxazole, C chloramphenicol, CAZ ceftazidime, CTX cefotaxime, CRO ceftriaxone, CXM cefuroxime, CIP
ciprofloxacin, CN gentamicin, TE tetracycline, ATM aztreonam, SAM ampicillin-sulbactam, FEP cefepime, IMP Impemene, MEM meropeneme, ET
Ertapeneme, DTCSH Debre Tabor Comprehensive Specialized Hospital, HUCSH Hawassa University Comprehensive Specialized Hospital, JUTSH
Jimma University Teaching Specialized Hospital, TASH Tikur Anbessa Specialized Hospital
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 14 of 18

A)

B)

C)

Fig. 5 (See legend on previous page.)

Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 15 of 18

A)

B)

C)

Fig. 6 Frequency of multidrug resistance at four hospitals. A Enterobacteriaceae B Acinetobacter species C Pseudomonas species. Percentages
represent the number of resistant isolates out of the total number of isolates at each hospital. Debre Tabor Comprehensive Specialized Hospital
(DTCSH), HUCSH Hawassa University Comprehensive Specialized Hospital, and Jimma University Teaching Specialized Hospital (JUTH) and Tikur
Anbessa Specialized Hospital (TASH), MDR multidrug resistance

differences maybe the rational use of antibiotics and the ertapenem, respectively. Effective treatment options for
fact that the cost of the drugs may be higher relative to Enterobacteriaceae were limited to amikacin, merope-
others, so people do not take these drugs for self-medica- nem, and imipenem. The most frequent isolate is E. coli,
tion in the study area [49]. which showed the highest resistance to ampicillin (cef-
Carbapenem resistance among Enterobacteriaceae was triaxone, cefotaxime, ceftazidime, cefepime, cefuroxime,
17.6%, 20.3%, and 32.9% to meropenem, imipenem, and ampicillin-sulbactam, trimethoprim-sulfamethoxazole,
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 16 of 18

and tetracycline. Similar studies were conducted in Ethi- CLSI Clinical Laboratory Standards Institute
DMIP Department of Microbiology Immunology and Parasitology
opia [33] and Iraq [50]. This might be due to the indis- DTTRH Debre Tabor Teaching, and Referral Hospital
criminate use of antibiotics in both hospitals [18, 19]. HUTH Hawassa University Teaching Hospital
In our study, an alarming level of carbapenem-resist- JUTH Jimma University Teaching Hospital
MALDI-TOF MS 
Matrix-assisted laser desorption ionization-time of flight
ant E. coli to ertapenem (24.5%), imipenem (11.8%), and mass spectrometry
meropenem (9.8%) was detected; however, it was lower MDR Multi drug resistance
than in another study [51]. The cause of the higher rates MHA Mueller Hinton Agar
SOP Standard operation procedure
compared to other settings may be irrational use or mis- SSI Surgical site infections
uses of antibiotics. The discrepancy of antibiotics resist- STATA​ South Texas Art Therapy Association
ance across sites includes differences in the rational use TASH Tikur Anbessa Specialized Hospital
ESBL Extended-spectrum beta lactamases
of antibiotics [52].
Acknowledgements
The authors would like to thank Addis Ababa University, Armauer Hansen
Strengths and limitation
Research Institute, Debre Tabor University, Uppsala University, and The Centre
The strength of this study was multicentred, enrollment of for Innovative Drug Development and Therapeutic Trials for Africa (CDT-Africa)
all age groups, a reasonably large sample size and re-char- for supporting this study. We would like to extend our gratitude to the Debre
Tabor compressive specialized hospital, Tikur Anbessa Specialized Hospital,
acterizing bacteria using MALDI TOF–MS an advanced
Hawassa University Teaching Hospital, and Jimma University Teaching Special-
bacterial identification method, and the limitation were ized Hospital for allowing us to conduct the study. We are thankful to the
unable to investigate anaerobic bacterial and fungal agents physicians, nurses, and microbiologists from all the study sites who helped us
due to limited laboratory resources at the hospitals. undertake this study. Lastly, our gratitude goes to all the study participants.

Author contributions
Conclusions SW: was the primary researcher, conceived the study, data collection, analysis,
interpretation of the findings, drafting the manuscript, and write-up. AAl
This multicenter study identified frequent and diverse substantially participated in laboratory work. TA, BS, GS, AAb, AM and GTB:
Gram-negative and Gram-positive SSI etiologies with- substantially participated in the design of the study, reviewed the manuscript,
out significant variation in primary etiologies between and provided critical intellectual content. All authors read and approved the
final manuscript.
hospitals. Isolation of various newly emerging bacte-
rial strains in all sites showed the growing epidemiology Funding
and diversity of SSI etiologies. E. coli and S. aureus were This study was sponsored by Addis Ababa University, Armauer Hansen
Research Institute, Uppsala University, SIDA/SAREC bilateral research coopera-
the leading Gram-negative and Gram-positive isolates, tion with Ethiopia and The Centre for Innovative Drug Development and
respectively. High antimicrobial resistance was detected Therapeutic Trials for Africa.
with varying frequency between hospitals. Gram-positive
Availability of data and materials
isolates revealed maximum sensitivity to vancomycin and The data sets generated during and/or analysed during the current study are
clindamycin, whereas, among Gram-negative isolates, available from the corresponding authors on reasonable request.
amikacin, imipenem, and meropenem were the most
effective antibiotics Furthermore, the overall ceftriaxone Declarations
resistance is about 90.5%. Among the study participants,
Ethics approval and consent to participate
72.1% took prophylaxis and developed SSIs. The find- Ethical clearance and approval were obtained from Addis Ababa University’s
ing of high levels of carbapenem resistance, especially College of Health Sciences and AAREC, AAUMF03-008/2020. Above all the
towards ertapenem, is alarming. data were collected after full informed and written consent or assent was
obtained from each participant.
Hence, to prevent the emergence and spread of MDR
SSI, we recommend effective antimicrobial steward- Consent for publication
ship and antibiotic treatment to based on AST of the Not applicable.

pathogens. At the national level, regular surveillance Competing interests

and monitoring of antimicrobial resistance patterns are We authors declare that no competing interest.
indispensable. This includes the careful monitoring of the
Author details
antibiotics used as prophylaxis and empiric treatment by 1
Department of Medical Laboratory Science, College of Medicine and Health
the concerned authorities. Sciences, Debre Tabor University, Debre Tabor, Ethiopia. 2 Bacterial and Viral
Diseases Research Directorate, Armauer Hansen Research Institute, Addis
Ababa, Ethiopia. 3 Department of Microbiology, Immunology and Parasitology,
Abbreviations College of Health Sciences, School of Medicine, Addis Ababa University, Addis
AAERC AHRI/ALERT Ethics Review Committee Ababa, Ethiopia. 4 Department of Medical Biochemistry and Microbiology,
AHRI Armauer Hansen Research Institute Uppsala University, Uppsala, Sweden.
AMR Antimicrobial resistance
AST Antimicrobial susceptibility testing Received: 4 July 2023 Accepted: 10 October 2023
ATCC​ American Type of Culture Collection
CDC Center for Disease Control and Prevention
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 17 of 18

References of Gondar Teaching Hospital, Northwest Ethiopia. Int J Microbiol.

1. Forrester JD, Maggio PM, Tennakoon L. Cost of health care-associated 2018;2018:3165391.
infections in the United States. J Patient Saf. 2022;18(2):e477–9. 24. Birhanu Y, Endalamaw A. Surgical site infection and pathogens in Ethio-
2. Berríos-Torres SI, Umscheid CA, Bratzler DW, Leas B, Stone EC, Kelz RR, pia: a systematic review and meta-analysis. Patient Saf Surg. 2020;14:7.
et al. Centers for disease control and prevention guideline for the preven- 25. Mengesha RE, Kasa BG, Saravanan M, Berhe DF, Wasihun AG. Aerobic bac-
tion of surgical site infection, 2017. JAMA Surg. 2017;152(8):784–91. teria in post surgical wound infections and pattern of their antimicrobial
3. Borchardt RA, Tzizik D. Update on surgical site infections: the new CDC susceptibility in Ayder Teaching and Referral Hospital, Mekelle, Ethiopia.
guidelines. Jaapa. 2018;31(4):52–4. BMC Res Notes. 2014;7:575.
4. Lewis SS, Moehring RW, Chen LF, Sexton DJ, Anderson DJ. Assessing the 26. Garner JS. Hospital Infection Control Practice Advisory Committee. Guide-
relative burden of hospital-acquired infections in a network of commu- line for isolation precaution in hospitals. Infection control and hospital
nity hospitals. Infect Control Hosp Epidemiol. 2013;34(11):1229–30. epidemiology. Infect Control Hosp Epidemiol. 1996;17:53–80.
5. Mengesha RE, Kasa BG-S, Saravanan M, Berhe DF, Wasihun AG. Aerobic 27. Cheesbrough M. District laboratory practice in tropical countries part II.
bacteria in post surgical wound infections and pattern of their antimi- 2nd ed. New York: Cambridge University Press; 2006. p. 45–58.
crobial susceptibility in Ayder Teaching and Referral Hospital, Mekelle, 28. Clinical Laboratory Standard Institute. Performance standards for
Ethiopia. BMC Res Notes. 2014;7(1):575. antimicrobial susceptibility testing; 30th ed. CLSI document M100. CLSI.
6. Raza MS, Chander A, Ranabhat A. Antimicrobial susceptibility patterns of 2021;M100.
the bacterial isolates in post-operative wound infections in a tertiary care 29. Awoke N, Arba A, Girma A. Magnitude of surgical site infection and its
hospital, Kathmandu, Nepal. Open J Med Microbiol. 2013;3(3):159. associated factors among patients who underwent a surgical procedure
7. Godebo G, Kibru G, Tassew H. Multidrug-resistant bacterial isolates in at Wolaita Sodo University Teaching and Referral Hospital, South Ethiopia.
infected wounds at Jimma University Specialized Hospital, Ethiopia. Ann PLoS ONE. 2019;14(12): e0226140.
Clin Microbiol Antimicrob. 2013;12:17. 30. Narula H, Chikara G, Gupta P. A prospective study on bacteriological pro-
8. Halawi E, Assefa T, Hussen S. Pattern of antibiotics use, incidence and file and antibiogram of postoperative wound infections in a tertiary care
predictors of surgical site infections in a Tertiary Care Teaching Hospital. hospital in Western Rajasthan. J Fam Med Prim Care. 2020;9(4):1927–34.
BMC Res Notes. 2018;11(1):538. 31. Ayala D, Tolossa T, Markos J, Yilma MT. Magnitude and factors associ-
9. Misha G, Chelkeba L, Melaku T. Bacterial profile and antimicrobial suscep- ated with surgical site infection among mothers underwent cesarean
tibility patterns of isolates among patients diagnosed with surgical site delivery in Nekemte town public hospitals, western Ethiopia. PLoS ONE.
infection at a tertiary teaching hospital in Ethiopia: a prospective cohort 2021;16(4): e0250736.
study. Ann Clin Microbiol Antimicrob. 2021;20(1):33. 32. Lubega A, Joel B, Justina LN. Incidence and etiology of surgical site
10. Spagnolo A, Ottria G, Amicizia D, Perdelli F, Cristina ML. Operating infections among emergency postoperative patients in Mbarara regional
theatre quality and prevention of surgical site infections. J Prev Med Hyg. referral hospital, South Western Uganda. Surg Res Pract. 2017. https://​doi.​
2013;54(3):131. org/​10.​1155/​2017/​63651​72.
11. National Institute for Health and Care Excellence: Clinical Guidelines. Sur- 33. Shiferaw WS, Aynalem YA, Akalu TY, Petrucka PM. Surgical site infection
gical site infections: prevention and treatment. London: National Institute and its associated factors in Ethiopia: a systematic review and meta-
for Health and Care Excellence (NICE). Copyright © NICE 2020. 2020. analysis. BMC Surg. 2020;20(1):1–15.
12. Moradali MF, Ghods S, Rehm BHA. Pseudomonas aeruginosa lifestyle: 34. Badia JM, Casey AL, Petrosillo N, Hudson PM, Mitchell SA, Crosby
a paradigm for adaptation, survival, and persistence. Front Cell Infect C. Impact of surgical site infection on healthcare costs and patient
Microbiol. 2017;7:39. outcomes: a systematic review in six European countries. J Hosp Infect.
13. Isibor JO, Oseni A, Eyaufe A, Turay A. Incidence of aerobic bacteria and 2017;96(1):1–15.
Candida albicans in post-operative wound infections. Afr J Microbiol Res. 35. Misha G, Chelkeba L, Melaku T. Incidence, risk factors and outcomes
2008;2(11):288–91. of surgical site infections among patients admitted to Jimma Medical
14. Organization WH. Global guidelines for the prevention of surgical site Center, South West Ethiopia: prospective cohort study. Ann Med Surg
infection. Geneva: WHO; 2016. (Lond). 2021;65: 102247.
15. Biadglegne F, Abera B, Alem A, Anagaw B. Bacterial isolates from wound 36. Costa T, Medeiros P, Salles M. Smoking increases the risk of surgical site
infection and their antimicrobial susceptibility pattern in Felege Hiwot infection after hydrocelectomy in adults: a retrospective cohort study in
referral Hospital North West Ethiopia. Ethiop J Health Sci. 2009;19(3). Brazil. J Infect Dev Ctries. 2018;11:950.
16. Mulu W, Kibru G, Beyene G, Damtie M. Postoperative nosocomial infec- 37. Jiang C, Chen Q, Xie M. Smoking increases the risk of infectious diseases:
tions and antimicrobial resistance pattern of bacteria isolates among a narrative review. Tob Induc Dis. 2020;18:60.
patients admitted at Felege Hiwot Referral Hospital, Bahirdar, Ethiopia. 38. Vasundhara Devi P, Sreenivasulu Reddy P, Shabnum M. Microbial profile
Ethiop J Health Sci. 2012;22(1):7–18. and antibiotic susceptibility pattern of orthopedic infections in a tertiary
17. Munckhof W. Antibiotics for surgical prophylaxis. Aust Prescr. 2005;28(2). care hospital: a study from South India. Int J Med Sci Public Health.
18. Ranjan KP, Ranjan N, Bansal SK, Arora D. Prevalence of Pseudomonas 2017;6(5):838–41.
aeruginosa in post-operative wound infection in a referral hospital in 39. Asres GS, Legese MH, Woldearegay GM. Prevalence of multidrug resistant
Haryana, India. J Lab Phys. 2011;2(2):129–32. bacteria in postoperative wound infections at Tikur Anbessa specialized
19. Li B, Webster TJ. Bacteria antibiotic resistance: new challenges and hospital, Addis Ababa, Ethiopia. Arch Med. 2017;9 (4):0.
opportunities for implant-associated orthopedic infections. J Orthop Res. 40. Asres GS, Legese MH, Woldearegay GM. Prevalence of multidrug resistant
2018;36(1):22–32. Bacteria in postoperative wound infections at Tikur Anbessa specialized
20. Fadeyi A, Adigun I, Rahman G. Bacteriological pattern of wound swab hospital, Addis Ababa, Ethiopia. Archives of Medicine. 2017;9(4).
isolates in patients with chronic leg ulcer. Int J Health Res. 2008. https://​ 41. Mohammed A, Adeshina GO, Ibrahim YK. Incidence and antibiotic sus-
doi.​org/​10.​4314/​ijhr.​v1i4.​55375. ceptibility pattern of bacterial isolates from wound infections in a tertiary
21. Iskandar K, Sartelli M, Tabbal M, Ansaloni L, Baiocchi GL, Catena F, et al. hospital in Nigeria. Trop J Pharm Res. 2013;12(4):617–21.
Highlighting the gaps in quantifying the economic burden of surgical 42. Khanam RA, Islam MR, Sharif A, Parveen R, Sharmin I, Yusuf MA. Bacterio-
site infections associated with antimicrobial-resistant bacteria. World J logical profiles of pus with antimicrobial sensitivity pattern at a teaching
Emerg Surg. 2019;14(1):50. hospital in Dhaka City. Bangladesh J. 2018. https://​doi.​org/​10.​3329/​bjid.​
22. Abayneh M, Asnake M, Muleta D, Simieneh A. Assessment of bacte- v5i1.​37710.
rial profiles and antimicrobial susceptibility pattern of isolates among 43. Dessie W, Mulugeta G, Fentaw S, Mihret A, Hassen M, Abebe E. Pattern
patients diagnosed with surgical site infections at Mizan-Tepi University of bacterial pathogens and their susceptibility isolated from surgical
Teaching Hospital, Southwest Ethiopia: a prospective observational site infections at selected referral hospitals, Addis Ababa, Ethiopia. Int J
cohort study. Infect Drug Resist. 2022;15:1807–19. Microbiol. 2016. https://​doi.​org/​10.​1155/​2016/​24189​02.
23. Bitew Kifilie A, Dagnew M, Tegenie B, Yeshitela B, Howe R, Abate E. 44. Bitew Kifilie A, Dagnew M, Tegenie B, Yeshitela B, Howe R, Abate E.
Bacterial profile, antibacterial resistance pattern, and associated fac- Bacterial profile, antibacterial resistance pattern, and associated factors
tors from women attending postnatal health service at University from women attending postnatal health service at University of Gondar
Worku et al. Ann Clin Microbiol Antimicrob (2023) 22:96 Page 18 of 18

Teaching Hospital, Northwest Ethiopia. Int J Microbiol. 2018. https://​doi.​

org/​10.​1155/​2018/​31653​91.
45. Kalayu AA, Diriba K, Girma C, Abdella E. Incidence and bacterial etiologies
of surgical site infections in a Public Hospital, Addis Ababa, Ethiopia.
Open Microbiol J. 2019. https://​doi.​org/​10.​2174/​18742​85801​91301​0301.
46. Szaniawski MA, Spivak AM. Recurrent Paenibacillus infection. Oxf Med
Case Rep. 2019;2019(5):omz034.
47. Tena D, Fernández C, Lago MR. Alcaligenes faecalis: an unusual cause of
skin and soft tissue infection. Jpn J Infect Dis. 2015;68(2):128–30.
48. Patterson SB, Mende K, Li P, Lu D, Carson ML, Murray CK, et al. Steno-
trophomonas maltophilia infections: clinical characteristics in a military
trauma population. Diagn Microbiol Infect Dis. 2020;96(2): 114953.
49. de With K, Allerberger F, Amann S, Apfalter P, Brodt HR, Eckmanns T,
et al. Strategies to enhance rational use of antibiotics in hospital: a
guideline by the German Society for Infectious Diseases. Infection.
2016;44(3):395–439.
50. Ali KM, Al-Jaff BM. Source and antibiotic susceptibility of Gram-negative
bacteria causing superficial incisional surgical site infections. Int J Surg
Open. 2021;30:100318.
51. Mora-Guzmán I, Rubio-Perez I, Maqueda González R, Domingo Garcia
D, Martín-Pérez E. Surgical site infection by carbapenemase-producing
Enterobacteriaceae. A challenge for today’s surgeons. Cir Esp (Engl Ed).
2020;98(6):342–9.
52. Ma F, Xu S, Tang Z, Li Z, Zhang L. Use of antimicrobials in food animals
and impact of transmission of antimicrobial resistance on humans. Biosaf
Health. 2021;3(1):32–8.

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations.

Ready to submit your research ? Choose BMC and benefit from:

• fast, convenient online submission

• thorough peer review by experienced researchers in your field
• rapid publication on acceptance
• support for research data, including large and complex data types
• gold Open Access which fosters wider collaboration and increased citations
• maximum visibility for your research: over 100M website views per year

At BMC, research is always in progress.

Learn more biomedcentral.com/submissions