Prevalence of Bacterial Contamination and Antimicrobial Susceptibility Pattern among Blood and Blood Components collected with and without diverging method at Armed Forces Comprehensive Specialized Hospital, Addis Ababa, Ethiopia

Background: Transfusion of bacterial contaminated blood and blood components could be a cause of morbidity and mortality. Understanding the mechanism of blood contamination is important in developing infection control strategy. Methods: A comparative cross-sectional study was done 376 blood and blood components collected with and without diverging method in Armed Forces Comprehensive Specialized Hospital. Then, Culturing of collected blood and blood components were inoculated in broth then subcultured on agar plate. And then, the drugs sensitivity test was done for each isolate. Bivariate analysis and multivariate logistic regression were used to infer association. Results: The overall prevalence of bacterial contamination among blood and blood components were 17 (4.5%). The prevalence of bacterial contamination of blood and blood components collected with the non-diverging and diverging method was 14 (7.4%) and 3 (1.6 %) respectively with P value of 0.05. Staphylococcus epidermidis were the most dominate isolates. Gram positive isolates showed more than 74% sensitive for antibiotics and also became more than 9% resistant. Most gram negative isolates became sensitive but Pseudomonas aeruginosa showed resistant for Gentamicin. 29.4% (n= 5/17) isolated bacteria were multidrug resistant. Conclusion: There was a difference between bacterial contamination in blood and blood components collected between diverging and non-diverging blood collection methods. Comprehensive Antimicrobial Susceptibility Test ; BAP: Blood Agar Plate Brain-Heart ; ; Coagulase ; ENDFBBS: Ethiopia Blood service ; FFP: Fresh Frozen Plasma MAP: MacConkey Agar Plate ; PC: Platelet Concentrates ; PRBC: Packed Red Blood Cell; TSI: Triple sugar Iron test TTBI


Introduction
After the invention of blood circulation, the rst recorded thriving blood transfusion was from dogs to dogs and then, lamb to human being. In the 1800's, the blood transfusion was started among human beings then after the outcome became successful [1]. The stipulation of blood products in the US at the World War II and immediate to the postwar era were signi cantly increased. [2].
Blood transfusion is a medical mediation planned to offer protected blood products for those who need them [3]. Nevertheless, Transfusion transmittable infection can be transferred from blood products to recipients. Common TTI causative agents and bacteria can be screened from blood [4]. Bacterial contamination of blood products could be the causation for morbidity and mortality when transfused them to patients [5].
In the half of 1990s, both literature and organization's haemovigilance systems indicated case report about bacterial transmission and this interest depicted by the blood community [6]. Either skin commensals or gastrointestinal tract ora which accounts for 75% are the principal gram-positive aerobic pathogens [7].
Blood products may be infected during blood collection, inadequate disinfection of donor skin, processing and accident on blood bags. In addition to that, they also infected from asymptomatic bacteremia in the donor [8]. In developed countries, the incidence of bacterial contamination and its related transfusion transmittable bacterial infection (TTBI) has been greatly condensed as a result of proper donor screening [9], uses of enhanced donor skin decontamination method and diverting the rst blood volume [10]. As opposed to developed countries, developing countries are not put into practice the above measures [8].
In African, the threat of bacterial contamination; gained through collection and processing, is 2500 times higher than developed countries. Moreover, visual examination of blood bags for hemolysis is the most common means of detecting for bacterial contamination [11].
The studies done in Ethiopia showed signi cant and variable prevalence of bacterial contaminations and their antimicrobial susceptibility pattern among whole blood collected by the non-diverging method [11,12] were reported but not with diverging method. To this end, the aim of the current study was shown the prevalence of bacterial contamination of blood products collected by two different donor blood collection methods.

Study area and population
A comparative cross-sectional study was conducted from January 2017 to April 2018 in Armed Forces Comprehensive Specialized Hospital Addis Ababa, Ethiopia. The sample size was calculated based on two population proportion formula. The value of P 1 was taken as 15.33%( 0.1533) [11] from the previous study on whole blood and P 2 as 50%. Considered 95% con dence interval and commonly used values for C p , power; C 0.05, 95%, were 13.
A total of 376 units of blood products were selected using convenient sampling method for blood culturing. Each unit of whole blood and PRBC was mixed before sampling with handshaking and striper and then we detached 20-25 cm segment after sealed the tubing at 5-10 cm away from the end of each blood bag's tubing. Each detached segment was labeled with coded labeling paper. In the biological safety cabinet, each unit of FFP, platelet, and detached segment was decontaminated rst using packed swab saturated with 70% isopropyl alcohol and waited for one minute and then with 2% tincture iodine then waited for three minutes. After decontaminated the sterilized brain-heart infusion broth (Oxoid, Basingstoke, UK) blood culture bottle stopper in the same way like above, 3 ml of blood samples were drawn from each segment using sterile disposable syringe and then dispensed into 15 ml of BHI broth aseptically. The specimens were delivered to the AFCSH Microbiology laboratory incubation room then isolated, identi ed and tested for susceptibility [13].
The inoculated BHI was incubated at 37 o C aerobically then observed daily for any possible signs of bacterial growth such as pellicle formation, hemolysis, gas formation clotting and turbidity for 7 days. After overnight incubation, the BHI bottles were mixed then one ml blood sample was taken from the bottle with a sterile disposable syringe and subculture two drops of sample on sheep Blood agar plate (BAP), MacConkey agar plate (MAP) and Chocolate agar plate (CAP) blindly. Inoculated BHI bottles; that show bacterial growth at 2 to 7 days, were also subcultured on sheep BAP, MAP, and CAP and also resubcultured until we got the pure colony. The inoculated MAP was incubated aerobically but BAP and CAP were incubated with 5-10% CO 2 atmosphere. Whenever bacterial growth was observed on the incubated media, identi cation of bacterial species was done based on bacterial cultural characteristics, gram stain and biochemical tests. The biochemical tests were done depending on their gram reaction result. For gram-positive bacteria, we used catalase, coagulase, CAMP, bile esculin, novobiocin, TSI tests and so on. We also used for gram-negative bacteria TSI, citrate, indole, motility, urease, and oxidase. The result of biochemical referred to the chart of gram-negative and positive bacteria to identify the species level of the bacteria. In addition, differential media like Mannitol salt agar was used to differentiate staphylococci species [14].

Results
In this study, we included 376 blood samples from blood products which were collected by non-diverging and diverging blood donor collection methods and delivered to AFCSH with equal proportion. Of this, O blood group with Rh-positive was the highest to be examined 137, (36.4%). the minimum and maximum storage temperatures for whole blood were 3. The overall prevalence of bacterial contamination among blood products were 17(4.5%). Among the blood products, whole blood and PRBC took the highest bacterial contamination growth 6 (35.3%). In addition, the most contaminated blood group type was O Rh-positive 6 (35.3%). From the overall prevalence of The prevalence of bacterial contamination from those blood products which were collected with non-diverging 7.4% (n = 14/188) was higher than diverging method 1.6% (n = 3/188). There was a statistical signi cant association between blood donor collection method and bacterial contamination (AOR (CI): 7.8 (1-60.1), P value: 0.05). Refer Table 1.
The most isolated bacteria from blood products collected with non-diverging method were Staphylococcus epidermidis (n = 4, 28.57%) and Staphylococcus aureus (n = 4, 28.57%). Whereas identi ed isolates from blood and blood components collected with diverging method were only Staphylococcus epidermidis (n = 2, 66.7%).
Among gram positive bacteria such as Staphylococcus aureus, Staphylococcus epidermidis Listeria monocytogenes Listeria species which were isolated from those blood products collected with non-diverging method. The most sensitive antibiotics for the identi ed isolates were Ceftriaxone (100%), Ampicilin (100%), Tobramycin (100%) and Cefotaxime (100%). The highest resistant antibiotics were Trimethoprim-sulphamethoxazole (27%), Erythromycin (27%),. Similarly, gram-negative isolates were sensitive for the antibiotics such as Amoxicillin + clavulanic acid (100%), Piperacillin-tazobactam (100%), Cefepime (100%), Ceftriaxone (100%), Tobramycin (100%), Amikacin (100%), Cipro oxacin (100%) and Trimethoprim-sulphamethoxazole (100%). The only antibiotic that showed resistant to Pseudomonas aeruginosa 33.3% was Gentamicin. Refer Table 2 and 3 Staphylococci epidermidis and Listeria spp were isolated from blood and blood components collected with diverging blood collection method. Most antibiotics were sensitive for these organisms except Ceftriaxone (100%), Chloramphenicol (100%) were intermediate. In addition, Trimethoprimsulphamethoxazole (67%) was the most resistant antibiotic. Refer Table 4 Finally, among the isolated seventeen organisms, 5(29.4%) were shown to be Multidrug resistant (MDR ≥ 2 drugs) for one group of spectrum of antibiotics. Staphylococci epidermidis 3(17.6%) took the highest MDR bacteria. Refer Table 5  Table 1 Prevalence of overall bacterial contamination and association factors among collection method, blood and blood components in AFCSH Addis Ababa city, Ethiopia, January to April 2018.  Table 2 Antimicrobial susceptibility pattern of gram-positive bacteria isolated from blood culture among stored blood and blood components collected by non-divergin to April 2018.      Table 3 Antimicrobial susceptibility pattern of gram-negative bacteria isolated from blood culture among stored blood and blood components collected by non-diver AFCSH Addis Ababa city, Ethiopia, January to April 2018.     Table 4 Antimicrobial susceptibility pattern of gram-positive bacteria isolated from blood culture among stored blood and blood components collected by diverging January to April 2018.   Table 5 Multidrug resistant level of isolated bacteria stored blood and blood components collected by non-diverging and diverging method in AFCSH Addis Ababa city, Ethiopia, January to April 2018.

Discussion
The awareness of bacterial contamination among blood products which are used for blood transfusion as well as their sources of contamination is imperative for setting up the preventive measure at blood bank centers and blood transfusion centers. Moreover, it is signi cant to offer action and improve the blood collection practice, education and policy.
In our study, the overall prevalence of bacterial contamination among blood products was 4.5% (N = 17/376). Blood group type O + 35.3% (N = 6/17,) showed the most contaminated one. Studies from Debre Markos, Ethiopia carried out by Esmael A et al [12] also agreed with this study.
The prevalence of bacterial contamination among those blood products which were collected with non-diverging donor blood collection method was 7.4% (n = 14/188) and almost close to previous studies reported from Africa countries like Ghana by Adjei A A et al 9% (n = 28/303) [8] and Nigeria by Bolarinwa R A et al 8.8%, (n = 14/162) [16]. The bacterial contamination of blood products collected with diverging done in this study is more or less similar to another nding like in Malaysia carried out by Jumaah N et al was 1.7% (n = 12 /702) [17].
Among the blood products, the highest bacterial contamination was observed in whole blood and PRBC. Another study was done in Ghana by Adjei A A. et al also supported that whole blood became the rst to be contaminated in equal proportion except that PRBC is not included in their study [8].
The isolated bacteria in this study were both gram-positive bacteria (Staphylococcus aureus, Staphylococcus epidermidis, Listeria monocytogenes, and Listeria species) and gram-negative bacteria (Klebsiellae species, Proteus mirabilis and Pseudomonas aeruginosa). Our nding was agreed with different studies like in Nigeria by Bolarinwa R A et al [16] and India conducted by Barot T et al [18].  [20], this difference may be due to the difference in sample size, sample taken from only whole blood, the present study on the other hand higher in prevalence of bacterial contamination than other studies done in Zimbabwe 3.1% (n = 6/196) [21], Uganda 3.5% (n = 18 /510) [5]. The reason for increased prevalence may be in our countries the disinfection type is focused on application of 70% alcohol only.
The prevalence of bacterial contamination in the diverging method of the current study is higher than different countries nding such as New zealand done by Dickson M and Dinesh D was 0.04% (n = 2378/59461) [22]. The difference may be the implementation of a compressive activity like proper donor screening, double disinfection, closed processing system and the existence of national haemovigilance programme in New zealand.
From this study; when we compared the bacterial contamination between blood donor collection methods of blood and blood components, the non-diverging method was higher than the diverging method one. In addition, the calculated adjusted odd ration became 7.8 so that the non-diverging blood donor collection method was likely more exposed to bacterial contamination by 7.8 times than diverging one.
Those blood products which were collected in both method share same conditions like application of single disinfection method, the absence of bacterial contamination screening and active national haemovigilance programme except changing the direction of the rst 30-40 ml of blood into the diverging pouch. So, switching the rst ow of blood reduced the contamination rate by 5.8%. This concept is also supported by the reviewed study in Italy by Liumbruno G M et al. The purpose of diverting the rst 40-50 ml of donated blood to reduce the microbes or skin fragments especially comes from donor skin entering into the collection bag [23]. In addition, other study in Japan by Satake M et al also maintained that the positivity rate of bacterial contamination were 36/21786 (0.17%) and 11/21783 (0.05%) without and with diversion method in that order. Even if this study done in platelet concentrates only, used different methodology, sample volume and include anaerobic bacteria detection [24].
As opposed to the current study, a study done in Zimbabwe by Makuni N et al the nding of the highest contaminated among blood products was platelet 10.3% (n = 4/36) followed by PRBC 1.3% (n = 2/149). There was no unit of whole blood was contaminated by bacteria. The difference may be they took unequal amount of from each blood products by Makuni N et al [21].
Considering the drug resistance pattern, the nding s of this study was agreed with a study from Nigeria and they also reported that gram-positive bacteria were sensitive for Gentamicin and Ceftriaxone but the majority of the antibiotics were resistant [16]. The only antibiotic that showed resistant to the Pseudomonas aeruginosa was Gentamicin. Majority of the antibiotics were sensitive for Klebsiella spp and Proteus mirabilis. But a study from Debremarko indicated that all the gram-negative organisms isolated were resistant to Cotrimoxazole and susceptible to Cipro oxacin and Cefoxicitin [12]. The levels of Multi drug resistant for one grouped drug spectrum (MDR ≥ 2 drugs) were 5(29.4%). In our study gram-positive bacteria showed MDR for chemically different drugs. Our nding was lower than Study done here in Gondar (66.7%), Ethiopia by Wondimu H et al [11].
In conclusion, our study showed that the prevalence of bacterial contamination of stored blood products was higher in those collected with non-diverging method than diverging one. So, there was a difference between diverging and non-diverging blood collection method. Furthermore, Staphylococci epidermidis became the most commonly isolated bacteria. A considerable level of resistant bacteria and MDR organisms were observed. Therefore, blood bank should improve blood donor collection method from non-diverging to diverging method.

Limitations
This study couldn't identify anaerobic bacteria.