Bacterial profile, antimicrobial susceptibility patterns, and associated factors among bloodstream infection suspected patients attending Arba Minch General Hospital, Ethiopia

Bacterial bloodstream infections are of great concern globally. Of late, the emergence of drug resistant bacteria worsen the related morbidity and mortality. This study was aimed to determine the bacterial profile, antimicrobial susceptibility patterns, and associated factors among the blood stream infection (BSI) suspected patients attending the Arba Minch General Hospital (AMGH), southern Ethiopia, from 01 June through 31st August, 2020. A cross-sectional study was conducted among 225 BSI suspected patients. Systematic random sampling method was used to select patients. Blood culture was done to isolate bacterial pathogens. Antimicrobial susceptibility test was performed by employing the Kirby-Bauer disc diffusion method. Descriptive statistics and multivariable logistic regression analysis were done by Statistical Package for Social Service (SPSS) version 22. The rate of prevalence of bacteriologically confirmed cases was 22/225 (9.8%). Majority of BSI were caused by Gram-positive cocci, 13/22 (59.1%), particularly the isolates of S. aureus, 7/22 (31.8%) followed by Enterococci species, 4/22 (18.2%) and coagulase-negative Staphylococci (CoNS), 2/22 (9.1%). Among the Gram-negative bacteria 9/22 (41.1%), Klebsiella species 4/22 (18.2%) was the prominent one followed by Escherichia coli 2/22 (9.1%), Pseudomonas aeruginosa 2/22 (9.1%), and Enterobacter species 1/22 (4.5%). All the isolates of Gram-negative bacteria were susceptible to meropenem whereas 69.2% of the isolates of Gram-positive counterparts were susceptible to erythromycin. Slightly above two third (68.2%) of the total isolates were multidrug resistant. Insertion of a peripheral intravenous line was significantly associated with BSI [p = 0.03; Adjusted Odds Ratio = 4.82; (Confidence Interval: 1.08–21.46)]. Overall results revealed that eventhough the prevalence of BSI in Arba Minch is comparatively lower (9.8%), multidrug resistance is alarmingly on the rise, which is to be addressed through effective surveillance and control strategies.


Materials and methods
Study area and period. The study was conducted at AMGH in Gamo Gofa Zone from 01st June through 31st August 2020. As per the annual report (2019) of Zonal Health department, the total population is 1, 544, 752. The hospital is situated in Arba Minch town of Southern Nations, Nationalities and Peoples' Region and located 505 km south of Addis Ababa. Arba Minch General Hospital is currently serving more than 1.5 million people per year and providing preventive, rehabilitative, and curative care through several departments. The in-patient wards are serving more than eleven thousand while the emergency outpatient department (OPD) is treating about seven thousand patients annually.
Study design and population. A cross-sectional study was conducted at five different wards of AMGH, comprising all patients clinically suspected of BSI (those who have body temperature greater than 38 °C, having tachycardia/tachypnea, leukocytosis/leukocytopenia) who fulfill the inclusion criteria. The inclusion criteria for the study subjects were (1) all patients ≥ 18 years who attended the OPD during the study period with suspected BSI, (2) those who were already admitted and who are suspected to have BSI. The exclusion criteria were: (1) those who took antibiotics 72 h prior to sample collection and (2) those who were in a state of coma. The study has been ethically approved by the Institutional Review Board of the College of Medicine and Health Sciences, Arba Minch University (Ref. IRB/176/12/17/03/2020).
Sample size determination and sampling technique. The required sample size was calculated using a single population proportion formula. A prevalence of 15.8% was chosen from a previous study conducted in adults having BSI, Jimma, Ethiopia 22 . After considering a confidence interval of 95% (z = 1.96) and 5% of marginal error (d = 0.05), the initial sample size was estimated to be 204 and by computing a 10% non-response rate, the final sample size was consolidated as 225 (n). The average values of 'n' obtained in the previous year (2019), during three consecutive months (June 1st through August 31st) were 29 (OPD), 143 (internal medicine), 60 (surgery), 20 (ICU), and 48 (obstetrics and gynaecology). The sample size for each ward was calculated according to the proportion formula and is equal to: The final sample size proportionately attained for each ward was 22 (OPD); 36 (obstetrics and gynaecology); 15 (ICU); 107 (internal medicine) and 45 (surgery). A systematic random sampling technique was used to select the study participants by calculating the K th value and it was inferred from the number of BSI suspected patients who attended the respective wards. The K th value was calculated as 300/225 = 1.3 and hence the patients were selected in a consecutive fashion.
Operational definition of BSI suspected patients. A patient whose age was > 18 and was showing atleast two or more clinical signs and symptoms, which include fever or hyperthermia, leukopenia or leukocytosis, tachycardia, and tachypnea with suspected or defined sites of infection, was considered a BSI suspect. Data collection. Written informed consents were obtained from each participant. All of them were examined by trained physicians. A pre-tested semi-structured questionnaire was used to collect the socio-demographic and socio-economic factors (age, sex, marital status, residence, educational status, occupation, family size and monthly income) through a face-to-face interview, and the clinical data of each patient were obtained from medical records.
Blood sample collection, culture and identification. Ten ml of blood was withdrawn twice aseptically by using a butterfly vacutainer from two separate peripheral veins maintaining a time gap of 30-60 min which was then directly inoculated into the blood culture bottle containing 100 ml Tryptic Soy Broth (TSB) in 1:10 ratio, from patients in the respective wards of AMGH. Samples were then transported to the nearby Medical Microbiology and Parasitology Laboratory, Department of Medical Laboratory Science, College of Medicine and Health Sciences, Arba Minch University and immediatelyincubated aerobically at 37 °C. Routine inspections were done twice a day, for a week for the presence of bacterial growth, like uniform/subsurface turbidity, hemolysis, coagulation of broth, surface floccular deposit, pellicle formation, and gas production, . They were further examined by Gram staining and then sub-cultured aseptically on blood agar, chocolate agar, MacConkey agar, and mannitol salt agar. The chocolate agar plates were incubated inside a candle jar to provide 5-10% CO 2 , whereas the other three agar plates (blood agar, MacConkey agar, and mannitol salt agar) were incubated aerobically for 18-24 h at 37 °C.
Growths if found in both the bottles were interpreted as positive whereas growth in any one of the blood culture bottles was declared as contamination (pseudobacteremia) 22 . Blood culture bottles with no signs of bacterial growth were similarly sub-cultured in the above mentioned pairs of agar media, after a week and were considered culture-negative if there were no growths 23 26 . For the assay, inoculums of respective bacteria were prepared in a sterile normal saline in such a way to maintain equivalent density as per the 0.5 McFarland standard. Test organisms were uniformly swabbed over the Mueller-Hinton agar and exposed to a concentration gradient of antibiotic diffusion, and then incubated at37°C for 24 h. Fifteen commercially available antibiotic discs (Oxoid, Basingstoke, Hampshire, UK) were used. For Grampositive bacteria, penicillin (P) (10 μg), cefoxitin (FOX) (30 μg), chloramphenicol (CHL) (30 μg), tetracycline (TC) (30 μg), doxycycline (DOX) (30 μg), vancomycin (VAN) (30 μg), erythromycin (ERY) (15 μg), gentamicin (CN) (10 μg), and ciprofloxacin (CIP) (5 μg) were used. In the case of Gram-negative bacteria, ampicillin (AMP) (10 μg), piperacillin (PIP) (100 μg), ceftriaxone (CRO) (30 μg), cefepime (CFP) (30 μg), amoxicillin-clavulanate (AUG) (20 μg), gentamicin (CN) (10 μg), tetracycline (TC) (30 μg), chloramphenicol (CHL) (30 μg), ciprofloxacin (CIP) (5 μg), and meropenem (MEM) (10 μg) were chosen. Antibiotics were selected based on the prescription policy followed in AMGH, which is the same as the national policy and also as per the recommendations of CLSI, 2019 26 . Diameters of the zones of inhibition were measured to the nearest millimeter and categorized as sensitive, intermediate, and resistant according to CLSI 26 . Isolates were classified as either susceptible or resistant to an antibiotic and all isolates with intermediate resistance were classified as resistant for a better fit in further statistical analysis.
Detection of multiple drug resistance and Methicillin-resistant S. aureus. The multi-drug resistance (MDR) in this study was extrapolated as resistance to three or more groups of antibiotics tested 27 . Isolates of S. aureus were further tested for methicillin resistance according to the CLSI guidelines by using cefoxitin disc 26 . Quality control. Prior to data collection, training was given to data collectors. A pretest was conducted in 5% (n = 11) of study participants at Chencha hospital to guarantee the quality. One day training was given for data collectors to minimize inter personal variations in the identification of clinical cases. The data were checked for completeness, accuracy, clarity, and consistency by the investigator on a daily basis. To maintain the quality, standard operating procedures (in-house SOP manual) were followed during collection (aseptic technique), transportation, and processing. Before sample processing, the prepared culture media were checked for sterility by incubating five percent of it for overnight and examining the presence of any growth. Fitness of the media were checked by inoculating control strains, before culture and sensitivity tests were also performed. Control slides were used to check the quality of Gram staining. Corresponding American Type Culture Collection (ATCC) strains were used as reference (standard) for quality control related to culture and susceptibility tests; S. aureus (ATCC 25923), E. coli (ATCC 25922), and P. aeruginosa (ATCC 27853). All the reference strains were procured from Ethiopian Public Health Institute. Data analysis. The data were analyzed using SPSS, Chicago, IL, the USA for Windows, version 22. Descriptive statistics, including frequency, mean and standard deviationswere done. Bivariate and multivariate logistic regression analyses were performed to evalaute the association between variables and the BSI. Variables with a p-value < 0.25 in the bivariable logistic regression model were subsequently analyzed in the multivariable logistic regression to control the confounding factors, and a p-value ≤ 0.05 from multivariable logistic regression was considered statistically significant 28 .

Prevalence of bloodstream infections.
Out of the 225 blood samples processed for culture, 22 (9.8%) culture sets which showed bacterial growth had bacteriologically confirmed BSI. No growth was observed in 198/225 (88%) blood culture sets. Five (2.2%) of the positive results were contaminanted and therefore exempted from further analysis. Amongst the culture-positives (n = 22), 13 (59.1%) were males. The age groups 18-40 and 41-59 had the highest number of positive cases, ie., 10 (45.5% in each group). In the present study, maximum number of culture-positive cases were observed among patients admitted in the departments of internal

Distribution of isolated bacteria.
In connection with the site of infections, Gram-positive cocci, S.
aureus was the predominant one (n = 7), with a distribution, respiratory tract (n = 2), genitourinary tarct (n = 2), surgical site (n = 1) and multiple source (n = 1). However, in the case of Gram-negative isolates, Klebsiella species (n = 4) was the most common, which distributed in respiratory tract (n = 3) and genitourinary tract (n = 1). With regard to the co-morbid conditions, Gram-positive isolate, S. aureus (n = 5) showed a preponderance, ie., n = 5. It was distributed in such a proportion that in patients with diabetic mellitus, n = 2, COPD, n = 2 and liver diseases, n = 1. However, in the case of Gram-negative isolates, Klebsiella species exhibited a predominance (n = 4), with a distribution pattern of two in HIV patients and one each in those suffering from diabetis mellitus and renal disease.
In culture-positive patients, in asscoiation with the various medical procedures, S. aureus (n = 4) was the dominant isolates, just like in the above discussed cases of clinical charateristics. Two isolates of it was found in te patients with peripheral intravenous lines and urinary catheterization. Mentioning the case of Gram-negatives, Klebsiella species was the pronounced isolate (n = 3), with two in patients with urinary catheterization and one each in the case of a patient under mechanical ventilation.
Only one of the isolates of CoNS, 1/2 (50%) was methicillin-resistant whereas all others showed susceptibility towards both gentamicin and erythromycin. However, all these isolates exhibited resistance to penicillin, and 1/2 (50%) of them were resistant against tetracycline, chloramphenicol, doxycycline, and ciprofloxacin.
All the isolates of P. aeruginosa were found to be resistant to piperacillin, 2/2 (100%), and on the contrary, all were susceptible to meropenem, 2/2 (100%). Similarly, isolate of Enterobacter species had shown resistance to ampicillin 1/1 (100%), but it issusceptible to ciprofloxacin, chloramphenicol, tetracycline, gentamicin, ceftriaxone, cefepime, amoxicillin-clavulanate and meropenem. Hundred percentage of E. coli were resistant to both ampicillin and ceftriaxone. Furhtermore, it was also observed that all these isolates were susceptible to cefepime, ciprofloxacin, and meropenem. Three fourth (75%) of the species of Klebsiella isolates were resistant to ampicillin, ceftriaxone, chloramphenicol, and amoxicillin-clavulanate. However, 75% of the isolates were susceptible to gentamicin and ciprofloxacin while all of them showed susceptibility to meropenem (Table 5).

Factors associated with bloodstream infections.
In bivariable logistic regression analysis, BSI was statistically significant in patients with age ≥ 60 (p = 0.03), and in those who underwent a medical procedure (incorporation of a peripheral intravenous device (p = 0.02) and or any surgery (p = 0.04)). However, in multi-

Discussion
We found that the prevalence of BSI is 9.8% (95% CI; 6.1-14.1). The cuture-positivity rate is comparable to the results reported from a couple of cities in Ethiopia such as Addis Ababa (13%) 18 and Jimma (8.8%) 29 , and also from Cambodia (8.8%) 30 . However, this prevalence was lower than that found in another study conducted in Jimma (15.8%) 22 itself and also Mekelle (28%), Ethiopia 16 , as well as in another African country, Zambia (24%) 31 . This fluctuation in the rates of prevalence could be attributed to differences in methods employed for blood culture (manual or automated), volume of blood used (5 or 10 ml), number of blood cultures taken (2 or 3 sets), and the lack of clinical indications (variations in clinical signs and symptoms of BSI), eventually increasing the proportion of negative results 32 . Besides, the sample size, nature of patients, design of study, geographical locations, blood culturing rates as well as infection control strategies followed in different countries might have also contributed 32,33 . Above all, infections caused by anaerobes and other etiological agents would had add to this disparity, with respect to the rate of isolation of cultures. The prevalence of BSI is found to maintain a male preponderance; however, there was no statistically significant association between the gender and BSI. A couple of studies also had documented that, majority of BSI isolates were obtained from males 34,35 . In contrast, studies reported from Mekelle and Jimma indicated that the prevalence of BSI is higher among females 16,22 .
All cases of BSI were found to be mono-bacterial and are consonant with some earlier studies reported from Addis Ababa, Ethiopia () and also from India 3,35,36 . The diversity of bacterial isolates (S. aureus, Enterococci species., CoNS, Klebsiella species , Enterobacter species , P. aeurogenosa, and E. coli) observed in the present study is in concordance with the information from some earlier results published from USA and the Gondar, Ethiopia 37,38 .
The infectious agents responsible for BSI vary from country to country with unique geographical peculiarities 39,40 . The most commonly isolated bacterial types belong to the Gram-positive group (59.1%) and their Gram-negative counterparts correspond to 40.9%. The same trend of predominance exhibited by the Grampositive group was also documented in previous studies reported from two cities of Ethiopia, Jimma (53%) 22 , and Mekelle (72.2%) 16 , and also from other countries, Zambia (61.7%) 31 and USA (59%) 37 . Besides, the estimated pooled prevalence of Gram-positive and Gram-negative bacterial isolates are 15.5 and 10.4% respectively, in diverse groups of BSI suspected patients from different parts of Ethiopia 17 . However, this is in contrast to the results of other studies done in Cambodia 30 , Côte d'Ivoire 40 , and Ethiopia itself (Addis Ababa) 36 , which reported Gram-negative as the dominant group. Epidemiological variations and diversity in etiological agents might have contributed to this disparity 32,33,41 . It has been envisaged that due to the emergence of modern medical care system, Gram-positive cocci have begun to be detected often as the predominant group of pathogens causing BSI in the early 1980s itself 42 .
Among the Gram-positive group, S. aureus was the most prevalent (31.8%) isolate in the blood culture. Similar patterns of prevalence, in accordance with the present study were already noted in two other cities of Ethiopia viz. Addis Ababa 36 and Jimma 22 , and also in another country, Cambodia 30 . A probable reason for the highest prevalence of S. aureus could be its widespread presence in hospital environments as a contaminant, which might have invaded the admitted patients, eventually establishing infections 43 . Moreover, it is also a commensal of skin and mucous membranes and may further invade the patient during surgery or instrumental manipulations 43 . However, further studies are required to arrive at a definite conclusion.
Enterococci species was the second most frequently isolated Gram-positive bacteria and the same trend was also observed in a previous work conducted in India too 35 . In contrast to our findings, a couple of studies reported from Ethiopia (Jimma and Addis Ababa) showed that CoNS are the second most predominant isolate among the commonly found pathogens 22,44 . These fluctuations could be correlated to the fact that during the last two decades, Enterococci species have been emerging as one of the major pathogens causing BSI 45 . However, in our study, CoNS was found to be a rarely isolated Gram-positive bacterium. Gram-negative bacilli, Klebsiella species was the most frequently isolated pathogens (18.2%) and this is in line with a previous study done in Addis Ababa, Ethiopia 44 . Bloodstream infections may arise as a primary condition or may be secondary to a focal infection at a defined body site, most commonly arising from the respiratory, gastrointestinal, and urogenital tracts 46 . In our study, a high prevalence of BSI was observed among patients having respiratory tract infections. An earlier study reported from Jimma, Ethiopia also had documented that respiratory tract is the primary site of infections 22 . These might be due to the fact that BSI are more common and possibly is the reflection of typical complications arising from community-acquired pneumonia 47 . The most prevalent co-morbid condition found in culture-positive patients was diabetes mellitus, 22.7%, and this resembles the results of some earlier studies done in Cambodia 30 and Côte d'Ivoire 40 .
It has also been revealed that 76.9 and 61.5% of Gram-positive cocci were resistant to penicillin (and doxycycline respectively. These observations are in line with the results of several studies reported from various cities of Ethiopia (Addis Ababa, Jimma and Mekelle) 16,22,44 . In our study, most of the Gram-positive bacteria showed lower levels of percentage resistance to gentamicin (33.3%), erythromycin (30.3%), tetracycline (44.4%), ciprofloxacin (44.4%), and vancomycin (25%). The same trend of resistance was observed in several studies reported from Ethiopia (Jimma and Mekelle) 16,22 . From these results, it may be envisaged that ciprofloxacin and tetracycline are not the suitable options for an empirical therapy for BSI.
Another important aspect is that 71% of isolates of S. aureus had shown resistance to penicillin whereas 57% of it was resistant to doxycycline curtailing their empirical usage in the study area and this is similar to an earlier trend reported from Mekelle and Addis Ababa, Ethiopia 16,36 . Twenty nine percentage of isolates of S. www.nature.com/scientificreports/ aureus exhibited resistance to erythromycin, which is the least whereas 43% of it was resistant against each of chloramphenicol, ciprofloxacin, gentamicin, and tetracycline. This is in accordance with some earlier findings reported from Mekelle and Jimma 16,29 . the trend observed in the current research indicated that 57.1% of S. aureus are MRSA strains. A similar pattern of resistance was observed in previous studies from Mekelle, Ethiopia, and USA 16,37 . The higher percentage of MRSA strains currently observed might be due to the frequent use of these drugs, especially the third-generation cephalosporin in hospital/clinics in Arba Minch as part of the emergency empirical therapy. Literature also indicated that the frequent use of cephalosporin in hospitals across the globe is correlated to the emergence and spread of MRSA 48 .
The second most predominant Gram-positive bacilli is Enterococcus species, 75% of which showed resistance to each of the two antibiotics, penicillin and doxycycline and these findings are similar to a previous research done in Nepal 49 . Our results also revealed that 75% of Enterococci species is susceptible to vancomycin. We found that all the isolates of CoNS are resistant to penicillin and this is in line with the results of a prior study conducted in Ethiopia (Addis Ababa) itself 44 . It was also observed that 50% of CoNS were methicillin resistant and this is also in agreement with a study reported from Addis Ababa 36 . On the other hand, all the isolates of CoNS were suscetible to gentamicin and erythromycin and are comparable to the results of studies done in Jimma, Ethiopia and also in India 22,50 . Besides, only 50% of the isolates showed suscetibilities towards tetracycline, doxycycline, ciprofloxacin and this is in accordance with a series of studies conducted in two cities of Ethiopia, viz. Jimma and Mekelle and also India 16,22,35 .
Gram-negative isolates showed varied patterns of resistance, ie., 85, 71 and 57% against ampicillin, ceftriaxone, chloramphenicol, and amoxicillin-clavulan respectively, ie., which are frequently used in the treatment of BSI in Arba Minch. The trend of resistance observed currently is similar to the patterns observed in a couple of previous studies conducted in Ethiopia 29,36 .
On the other hand, a lower percentage of resistance ie., 22.2, 33.3, 28.5 and 33.3% were shown against four antibiotics such as ciprofloxacin, gentamicin, tetracycline, and cefepime respectively. This resembles the results reported from Jimma and Addis Ababa 29,36,44 . It is to be specified that all the isolates of Gramnegative bacteria showed susceptibility to meropenem and a similar result was previously reported from Nepal 49 .
Three fourth (75%) of the isolates of Klebsiella species were resistant to ampicillin, ceftriaxone, and amoxicillin-clavulanate and on the contrary were susceptible to gentamicin and ciprofloxacin; however all the isolates were extremely susceptible to meropenem. These are comparable to the results of past studies done in Ethiopia 29,36 .
All the isolates of E. coli were susceptible to ciprofloxacin, cefepime, and meropenem which are by and large in line with the data obtained from a couple of studies reported from Ethiopia 29,36 . However, all of them exhibited resistance to ampicillin and ceftriaxone, and are found to be in consistency with earlier reports from Ethiopia and Nepal 22,44,49 .
Invariably all the isolates of P. aeruginosa were resistant to piperacillin and this is in accordance with the results of a study done in Addis Ababa 44 . On the other hand, all of them were susceptible to meropenem and is also mentioned in a work done in Nepal 49 . Isolate of Enterobacter species showed resistance to ampicillin and this is comparable to the results of earlier studies done in the capital city of Ethiopia 36 . Interestingly, it also showed susceptibility to all the other drugs tested.
Development of bacterial resistance against multiple drugs is a major crisis that restricts the drug of choice for the treatment of BSI. In our study, MDR was observed in the case of 68.2% of isolates, which is comparable to a study conducted in Mekelle, Ethiopia 16 . Among the MDR, 60% were Gram-positive and the rest were Gramnegative and these findings are in line with the results of a study conducted in Congo too 51 . Those authors had reported that 34.6% of the Gram-negative are MDR. Most frequently, resistance was displayed by Gram-negative bacteria against the antibiotic, ampicillin.
Among the various risk factors assessed related to BSI, insertion of a peripheral intravenous line was found to be significantly associated and it was seen that patients with this device are 4.8 times [AOR 4.82, (95 CI: 1.08-21.46)] more prone to infections. Studies so far conducted in different countries also substantiate having a peripheral intravenous line is very much a risk factor causing BSI 52,53 .
Peripheral intravenous line-associated BSI often come from either the tip of the catheter or from the skin around it, so that during insertion under emergency conditions, exogenous pathogens may enter into the bloodstream, eventually leading to BSI 54 . Another explanation is that BSI can also occur due to the lack of proper aseptic measures practiced during the process of catheterization,ie., non-compliance to catheter insertion protocols by hospital personnels.
Results of the bivariate analysis show that patients with age ≥ 60 (p = 0.03) and those who underwent a surgery (p = 0.04) were significantly associated with BSI; however, these were not identified as independent risk factors in the multivariate analysis. Likewise, no significant relationship was found among socio-demographic and socio-economic parameters, and clinical factors of patients such as the site of infections (respiratory, urinary, intra-abdominal, soft tissue and skin, and surgical), medical procedures (urinary catheterization, surgery, and mechanical ventilation), chronic diseases (diabetes mellitus, HIV, renal failure, COPD, liver disorder, and heart failure) and malnutrition.The lack of association among BSI and the aforementioned factors have also been described in studies conducted in Egypt and Japan 55,56 .

Limitations
Shortcomings of the present work include shorter duration and the type of design of the study (cross-sectional), smaller sample size,and the lack of some of the chemicals. In addition, extended spectrum beta lactamase production among isolates and minimum inhibitory concentration for vancomycin were not detected andalso only aerobic cultures were analysed, which limit the identification of anaerobic pathogens. Also, this single-institution www.nature.com/scientificreports/ based study did not includea wider community and or even other hospitals. Conventional methods of blood culture were employed and finally, only the suspected cases were selected and this would probablyhad resulted in the exclusion of cases of intermittent bacteremia. Overall, this study was a single institution-based cross-sectional one and therefore the results obtained do not represent the general population. However, a remarkable finding is that the isolates of blood culture were resistant to frequently used antibiotics in Arba Minch.

Conclusions
This study revealed that the overall prevalence rate of BSI in Arba Minch is 9.8% and it is comparatively lower than that earlier reported at the national level in Ethiopia. Gram-positive bacteria, especially the isolates of S. aureus were found to be the most prevalent causative agent of BSI. Based on the results of antimicrobial susceptibility tests, it might be inferred that antibiotics such as meropenem and erythromycin respectively are the effective drugs against Gram-negative and Gram-positive bacteria However, there is a notable growth in antibiotic resistance, against several clinically relevant antimicrobials (such as penicillin, ampicillin, doxycycline, amoxicillin-clavulanate, ceftriaxone, and chloramphenicol) in the study setting. A remarkably high rate (68.2%) of MDR was also observed. An important factor associated with BSI in Arba Minch is the usage of peripheral intravenous lines and it is the most mentionable finding, which is to be given due attention.