Antimicrobial susceptibility patterns of bacteria isolated from patients with ear discharge in Jimma Town, Southwest, Ethiopia

Background Otitis media is among the leading causes of childhood illnesses although it can also affect the adults resulting in frequent physician visits, drug prescription and a key contributor to antibiotic resistance. The aim of this study was to determine the risk factors, bacterial profile, and the antimicrobial susceptibility pattern of the isolates from patients with discharging ears which clinically equates to draining otitis media in developing countries with limited medical resources such as otoscope. Methods A prospective cross-sectional study was conducted on 173 patients with draining otitis media. The ear discharge specimens were collected and analyzed by standard microbial techniques. The antibiotic susceptibility profiles were determined for 19 different antibiotics by the standard disk diffusion method. Data was analyzed by SPSS version 22 and the P value of less than 0.05 was considered as statistically significant. Results Among 173 otitis media patients participated in the study; majority, 102(63%) were pediatrics, out of which 72 (41.61%) were in the age group of less than 4 years. Ear infection was bilateral in 39 (22.54%) and chronic in 100 (57.8%) of the patients. Pathogens were isolated from 160 (92.5%) of the patients with a total of 179 isolates. The predominant isolate was Staphylococcus aureus (30.72%) followed by Proteus spp. (17.89%). The result of this study showed that adult age (p = 0.031), rural residence (p = 0.005), previous history of health care visit and treatment (p = 0.000), upper respiratory tract infection (p = 0.018) and presence of cigarette smoker in the house (p = 0.022) had statistically significant association with chronic otitis media. Most of the isolated bacteria showed high level of resistance to ampicillin/amoxicillin (88.3%), penicillin G (79.5%) followed by trimethoprim /sulfamethoxazole (73.8%). Conversely, the majority of bacterial isolates showed moderate susceptibility to ciprofloxacin (72.9%), gentamicin (70.4%), and amikacin (69.3%). Bacterial isolates identified in this study showed trend of multiple drug resistance, majority (67%) being resistant to three or more antimicrobials. Conclusions Majority of the bacterial isolates were multidrug resistant, hence, efforts to isolate microorganisms and determine the susceptibility pattern should be strengthened to improve the treatment outcome of otitis media instead of the usual trend of empirical treatment.


Background
Ear infection which can be classified into otitis media (OM) and otitis externa, is a major public health concern in developing countries associated with high burden of disease and economic impact to patients, families and the health care system. It is one of the most frequently encountered illnesses in children leading to repeated outpatient department (OPD) visits in both developed and developing countries even if it can also affect the adults [1]. OM is a continuum of disorders ranging from a simple acute otitis media (AOM) to recurrent acute otitis media (RAOM), otitis media with effusion (OME) or chronic otitis media (COM) [2]. Untreated or inadequately treated OM due to either inaccurate diagnosis or inappropriate use of antibiotics leads to purulent otitis, often with perforation and further complications including RAOM, persistence of middle ear effusion which requires the insertion of drainage tube and often leads to hearing impairment, mastoiditis, meningitis, COM, brain abscess and sepsis [3,4]. It also leads to impaired speech and language development, poor school performance and impaired social interaction [5]. Moreover, otitis externa is an infection of the outer ear canal commoner in patients who have eczema and or diabetes.
Diagnosis of OM is complicated by a lack of correlation between clinical features and responsible pathogens and drug susceptibility pattern. Additionally, routine unavailability of an othoscope in many of the health facilities particularly in the developing countries which is necessary to differentiate the different spectrum of clinical findings necessary for the case definition of OM also limits the health workers' ability to make better diagnosis and classification of OM [6]. Ear infection can be classified in different ways depending on the duration of illness, and the type of clinical manifestation the patient is having or otoscopic findings [6]. In Ethiopia and other developing countries, one of the commonly used classification of ear infection is the one depending on the duration of illness, which classifies ear infection into acute ear infection (symptoms lasting for less than 14 days) and chronic ear infection (symptoms lasting for ≥14 days) [7,8]. Empiric treatment of ear infection is not always appropriate since drug susceptibility patterns change overtime and empiric antibiotic therapy may not be effective at times and could contribute to development of antimicrobial resistance in the long run [9,10]. Early, prompt and effective treatment of ear infection will significantly reduce both short and long term complications associated with ear infection and can also improve the quality of life of patients with ear infection. Thus, having up-to-date information on the etiologies responsible for ear infection and their antimicrobial susceptibility pattern is crucial. This study was done with an objective of determining the bacterial profile, antimicrobial susceptibility pattern and associated factors of ear infections among patients of all age groups visiting health institutions found in Jimma Town, Ethiopia during the study period with a complaint of ear discharge.

Study design, period and area
A prospective cross-sectional study was conducted at Jimma Town from February to September 2017 in five public health institutions: two hospitals (Jimma University Medical Center and Shanan Gibe Hospital) and three public health centers (Jimma, Higher-2, and Mendera Kochi Health Centers) that provide health services for the residents of Jimma Town and surrounding. All the health facilities give services to both children and adults. Majority of the patients coming to the health centers and Shanan Gibe Hospital are residents of Jimma Town whereas those coming to the Jimma University Medical Center (JUMC) could come from the town as well as the surrounding towns or rural areas.

Study participants and data collection
We used consecutive sampling technique where we approached every patient coming with ear discharge during the study period. We used the presence of an ear discharge as an entry point since routine otoscopic examination might not be carried out in many of the health facilities even if this approach undermines the magnitude of ear discharge as we might have missed some patients who didn't have an obvious ear discharge. Thus, the study participants were patients of all age category with a presumptive/clinical diagnosis of draining OM who were willing to give consent to participate in the study (parents'/care takers' consent was taken in case of children). The relevant patient information was collected by trained nurses, while the swabs were collected by nurses at the health centers and by ENT doctor and GPs at the two hospitals. Otoscope and headlight were used during specimen collection at one of the health facilities (JUMC) whereas naked eyes were used at the other health facilities since these equipment were not available at these health facilities. We classified ear infection as acute if the duration of discharge was < 14 days and chronic if the duration of discharge was ≥14 days based on the classifications we use locally [7,8]. All the relevant history and physical examinations were done by the treating health workers and the investigators just collected the necessary information for the study.

Isolation and identification of bacteria
Ear discharge was collected under strict aseptic technique using single use commercially available sterile cotton swabs with utmost care to avoid surface contamination (i.e. first cleanse the external ear canal with antiseptic solution and then the pinna was pulled outward and backward to make it more straight forward for taking the ear swab, then sterile cotton swabs were gently rotated and taken out) and then immediately immersed into Amies transport media with charcoal (Oxoid, England).Within 2 h of collection, swabs were transported to the microbiology laboratory of JUMC, inoculated on blood agar, chocolate agar, and MacConkey agar. The MacConkey and blood agar plates were incubated in aerobic condition, whereas chocolate agar plate was kept in a candle jar, which was able to generate about 5-10% CO 2 . All of the inoculated media were incubated at 37°C for 18-24 h. All organisms grown were identified according to the standard microbiological methods by using Gram staining reaction, culture characters, colony morphology, pigment production, and type of hemolysis on blood agar. Moreover, conventional biochemical tests like Catalase, Coagulase, Optochin and bacitracin test for gram-positive bacteria and Oxidase test, Triple Sugar Iron Agar (TSI) (OXOID, UK), Citrate utilization test (OXOID, UK), Urease test (OXOID, UK) and Motility Indole Lysine (MIL) [OXOID, UK]) for gram-negative bacteria were used for identification [11].

Statistical analysis
The data was checked for completeness, coded, cleaned and entered into Epi-Data version 3.1 and then exported to and analyzed using SPSS version 22.0. Descriptive statistics such as frequency, percentage and cross tabulation were used to present the findings. Chi-square test was performed to evaluate the presence of statistically significant association and P-value less than 0.05 was considered as statistically significant.

Results
A total of 173 patients with ear discharge participated in this study, out of which 109 (63%) were children < 14 years. Participants' age ranged from 2 months to 70 years with a mean age of 12.8 years and median of 7 years. Most study subjects were in the age group of < 4 years (41.61%). Majority had unilateral 134 (77.46%), chronic 100 (57.8%) and purulent 113 (65.3%) ear discharge. Acute ear infection was more common in those < 4 years (53.8%) whereas chronic ear infection was more commonly seen in the age group of 15-44 years (34.6%). History of previous visit to health care facilities related to ear infection within 2 month period was noted among 102 (58.96%) patients and these patients had reportedly been treated with one or more antimicrobial agents, the most commonly prescribed antibiotics being amoxicillin, followed by amoxicillin-clavulanic acid ( Table 1).
The chi-square test indicated that adult age (p = 0.031), rural residence (p = 0.005), previous history of health care visit and treatment (p = 0.000), upper respiratory tract infection (p = 0.018) and presence of cigarette smoker in the house (p = 0.022) had statistically significant association with chronic ear infection rather than acute ear infection ( Table 2).
Among 173 of the study participants, 179 bacterial agents were isolated from 160 (92.5%) patients. Citrobacter spp. and Coagulase negative Staphylococcus were the predominant organisms isolated from patients with acute ear infection whereas P. vulgaris, P. mirabilis and P. aeruginosa were the predominant organisms isolated from patients with chronic ear infection. Among the culture positive swabs, 89.02% demonstrated single organisms, while the remaining 10.98% were positive for two organisms The rates of poly-microbial infection among acute and chronic ear infection cases were 12 (7.5%) and 7 (4.4%), respectively. Out of the 179 bacterial isolates, 12 different pathogenic bacterial species were identified. Among the isolated pathogens, Gram-negative bacteria 101 (56.4%) were predominant isolates than Gram-positives 78 (43.6%), with the ratio of~1.3:1. The predominant bacterial isolate was S. aureus (30.72%) followed by Proteus spp. (17.89%), and P. aeruginosa (10.61%) ( Table 3).

Discussion
In the present study, the highest percentage of ear infection was found among pediatric patients (63%) and this agrees with reports from other parts of Ethiopia [3,4,[14][15][16][17]. In addition, the majority of bacterial isolates were identified in the age group < 4 years (41.61%) which is also in line with other previous studies [4,15,16,18,19], although a study done in Dessie (Ethiopia) reported higher frequency in the age group of 16-35 years (42.4%) [20].
In this study, the majority of the patients (57.8%) had ear discharge for ≥14 days, which is in agreement with previously done studies in Ethiopia, in which COM accounted for 60-83% of the OM cases [3,4,21,22].
In the present study, from the total bacterial isolates, gram-negative bacteria (56.4%) were slightly higher than gram-positive bacteria, which is in agreement with previous studies done in various parts of Ethiopia: Mekelle (56%), Gondar (56.4%), Bahir Dar (58.8%), Addis Ababa (60.5%), Dessie (74.2%) and (78.7%), Wollo (75.8%) and Hawassa (79.5%) [3, 4, 14-16, 18, 20, 23]. The leading isolated bacteria in this study was S. aureus (30.72%), followed by Proteus spp. (17.89%) and P. aeruginosa (10.61%), similar to reports of other investigators from Mekelle and Addis Ababa [4,18]. Unlike our findings, Proteus spp. followed by S. aureus and Pseudomonas spp. were the predominant isolates reported by other researchers from different parts of Ethiopia [3,15,16,20,22] and relatively different patterns were reported from elsewhere [25][26][27] with P. aeruginosa as the main isolate followed  Note: AMP ampicillin/amoxicillin, AMC amoxicillin clavulanic acid, CRO ceftriaxone, CIP ciprofloxacin, C chloramphenicol, FEP cefepime, CN gentamicin, CAZ ceftazidime, SXT sulfamethoxazole-trimethoprim, TOB tobromycin, AK amikacin, CXM cefuroxime, NA not applicable by S. aureus and Proteus spp. The possible reasons for such variation in the bacterial profile might be attributed to the difference in climatic and geographic variation of the study sites. In our study, the most prevalent organism responsible for acute ear infection was S. aureus (37.5%). Even though, the global reports show that H. influenzae, S. pneumoniae, and M. catarrhalis to be the most prevalent organisms responsible for AOM [3], our findings is in agreement with reports from other African countries that indicate S. aureus and S. pyogenes were the predominant isolates [28]. The reason for this might be the differences in geographic location, prevalence of respiratory infection, coverage of pneumococcal conjugate vaccine (PCV), and possible overuse of antimicrobials that might have killed the sensitive organism and favored the drug resistant ones to be predominant, biofilm phenotypes property of the S. aureus and potentially other local and regional factors.
In the present study, both Proteus spp. (P. vulgaris (p = 0.018) and P. mirabilis (p = 0.037)) (84.4%) were more common among the chronic than the acute ear infections. This finding is comparable to Seid et al. from Dessie [3] and Muleta et al. from Jimma [22] who reported rates of 85.4 and 74.5%, respectively and contradicts with Wasihun and Zemene's from Mekele [4] who reported that P. mirabilis was seen in 63% of COM and P. vulgaris in 57% of AOM. The possible reasons for this might be Proteus spp. were common isolates in patients presenting lately (2 months after onset of ear discharge), as a result patients with discharging ears may not notice immediately for early diagnosis or the antibiotic treatment was not effective [29].
In this study, those with history of previous health care visit and treatments (58.96%) showed a significant association with chronic ear infection (p = 0.000), which is similar to Wasihun's and Zemene's reports [4]. This could probably be due to failure of empiric treatment of AOM. Thus, it might be wise to take a swab if the patient has ear discharge, then begin treating the patient with topical +/− systemic antibiotics according to empirical guidance available. In the meantime, proper etiology based diagnosis should be in place and once the organism is known, appropriate management of ear infection becomes effective. This in turn facilitates the rational use of antibiotics based on recent and local data in our health facilities in order to reduce sequales associated with ear infection [1]. In this study, the isolated bacteria showed highest rate of resistance to the different antibiotics commonly used for OM treatment which is in line with earlier reports from different parts of Ethiopia [3,4,16,20] and a good overall antimicrobial susceptibility pattern (> 70%) was seen to gentamicin and ciprofloxacin which is also in line with other studies conducted in Ethiopia [3,4,16,20,23] and in other countries asuch as Ardebil [30], Iraq [31], Nepal [32], India [33], and Jordan [34]). In contrast to these reports, gentamicin and ciprofloxacin were reported as ineffective from a study conducted in Nigeria [24].
Overall, 67% of the bacterial isolates from this study were characterized as MDR pathogenic bacteria. The reason for this might be linked to a prescription of antibiotics without laboratory guidance, purchasing of drugs without proper prescription (self-medication) in the local pharmacies and drug stores, misuse of antibiotics, indiscriminate use of antibiotics including animal husbandry, inappropriate prescribing habits and an over-zealous desire to treat every infection using antibacterial agents. Moreover, biofilm bacterial properties of common isolates, unavailability of bacterial culture facilities and poor infection prevention and control practices may be some of the different factors that can contribute to the development of MDR among these isolates.
The limitations of this study include the fact that we did not include the OME and OM without discharge as well as the fact that we did not try to isolate strict anaerobic bacteria and fungi which might also be the possible causative agents for OM. Additionally, we didn't intend to differentiate OM from otitis externa.

Conclusion and recommendation
S. aureus, Proteus spp. and P. aeruginosa were the three predominant bacterial isolates from patients with ear discharge. The profile and predominant bacteria isolated highlights the need for continuous surveillance and reporting of the microbiology of ear infection in our local community in order to guide clinicians use the appropriate antimicrobials towards the incriminated etiologies. The pathogens which have been isolated from ear discharge have shown a high level of antibiotics resistance and MDR features in the study area. Almost all the isolated bacteria showed a considerable level of resistance particularly to the commonly used antibiotics like ampicillin, amoxicillin, amoxicillin/clavulanic acid, and trimethoprim-sulfamethoxazole. Ciprofloxacin, amikacin and gentamicin are effective against all the bacterial isolates. In general, the result of this study revealed that antibiotic-resistant bacteria recovered from patients with ear infection are alarmingly increasing in Jimma area and becoming a major public health problem in the management of patients with a middle ear infection. Due to the increase in resistance to antibiotics one can advocate to swab the ear first and wait for the result especially if the patient has already had treatment as shown by our data which is often the case. We strongly recommend nationwide antimicrobial surveillance to make the right recommendation of antibiotics along with strict adherence to antibiotic use policy to reduce the spread of drug-resistant microbes and associated complications in the country. Hence, parallel to empiric treatment of ear infection, execution of culture and antimicrobial susceptibility test shall be taken as routine and mandatory practice in order to appropriately manage ear infection, reduce associated complications (individual, household, and health system) and reduce emergence of drug resistance in the community. Moreover, in the new cases we recommend taking swab before starting a regime based treatment in case there is already resistance. Further researches are needed to identify high resistance strains and characterization of resistance strains using molecular techniques.