Antibiotic Susceptibility Patterns of Bacterial Isolates from Routine Clinical Specimens from Referral Hospitals in Tanzania: A Prospective Hospital-Based Observational Study

Background

Antimicrobial resistance (AMR) has become a severe global health threat, especially in sub-Saharan Africa where the burden of infectious diseases is high.¹ Common illnesses like pneumonia, diarrhea, sexually transmitted infections, postoperative infections, and tuberculosis are increasingly becoming untreatable.² Patients infected with drug-resistant pathogens are at increased risk of unfavorable clinical outcomes and consume more health-care resources than patients infected with non-resistant pathogen of the same type.³ Recent data shows that about 700,000 death per year are attributable to AMR infections and projected to increase to 10 million annually by 2050 if the present trends persist.⁴ Clinically, the emergence of AMR is primarily driven by non-compliance with prescribed antimicrobials therapy, which in turn promotes spontaneous mutations in chromosome or control genes leading to new mutant pathogens with selective pressure in the presence of antimicrobials.⁵ Further accumulation of such beneficial mutations catalyzes the evolution of multidrug-resistant strains,⁶ which necessitate the use of broad-spectrum antibiotics as the ultimate choice. Especially alarming is the emergence and spread of multi-drug resistant bacteria which are hard to treat with the available antibiotics.² This include the emergence and spread of extended-spectrum beta-lactamases (ESBL), AmpC beta-lactamases, and carbapenemase-producing Gram-negative bacteria (ie carbapenem-resistant Enterobacteriaceae-CRE) and Staphylococcus aureus (MRSA) which are rapidly increasing worldwide.^7–9

In sub-Saharan Africa, AMR is mostly exacerbated by limited access to appropriate therapies, irrational use of antibiotics, and lack of clinical microbiology laboratories for drug susceptibility testing.¹ Most infections are managed empirically using antibiotics which are mainly obtained over-the-counter without a proper prescription. At the level of public health systems, the situation is worsened by a lack of coordinated AMR surveillance and weak regulatory frameworks for the access and use of antibiotics.¹⁰ In addition, the emergence of resistant pathogens is aggravated by underlying illnesses, such as human immunodeficiency virus (HIV), which has promoted the utilization of antibiotics against opportunistic infections.¹¹

Although bacterial infections particularly Gram-negative bacteria are among major causes of illness and death in sub-Saharan Africa,^12,13 data on their antibiotic resistance remain scant due to limited disease detection and surveillance capacity. Prevalent bacterial infections include lower respiratory infections (pneumonia), diarrheal diseases, urinary tract infections, bloodstream infections (typhoid, sepsis, meningitis, and bacteremia), sexually transmitted infections (eg, gonorrhea), and healthcare-associated infections (eg, MRSA).^12,13 A recent situation analysis on antibiotic use and resistance in Tanzania revealed a lack of national data representativeness for antibiotic resistance for common bacterial infections.¹⁴ The current study aimed to provide additional evidence on the resistance pattern of bacterial isolates to widely used antibiotics in Tanzania.

Methodology

Study Design and Setting

We conducted a hospital-based observational study between October 2018 -September 2019 involving four cross-border tertiary hospitals from four different regions in Tanzania. Such hospitals included: Maweni Regional Hospital in Kigoma Region, Musoma Regional Hospital in Mara Region, Sumbawanga Regional Hospital in Rukwa Region, and St. Benedict Ndanda Hospital (Masasi) in Mtwara Region (Figure 1). These hospitals were among satellite study sites for the East African Public Health Laboratory Network (EAPHLN) project and health facilities that form a surveillance system for monitoring cross-border disease outbreak dynamics in the country. The majority of inhabitants (80%) reside in rural settings, practicing subsistence farming and informal trade.

Figure 1 Map of the Republic of Tanzania showing geographic locations of the four study sites.

Administrative boundary shapefile was obtained from the Tanzania National Bureau of Statistics.

Data Collection

Clinical specimens were collected from both outpatients and inpatients as part of each hospital’s routine clinical care. Thus, specimens were requested based on clinicians’ assessment and then submitted for microbiological tests.

Laboratory Procedure

Specimen processing, identification of organisms to the genus and/or species level, and in-vitro antibiotic susceptibility testing were performed in accordance with the standard microbiological procedures and the CLSI guidelines.¹⁵ Pathogenic bacteria were identified using standard microbiological methods such as morphology on culture media, gram staining, and conventional biochemical tests. In case of ambiguity, the analytical profile index (API™) biochemical test kit (BioMérieux, France) and/or serological tests (antisera for V. cholerae) were used to confirm identification. In-vitro antibiotic susceptibility testing was performed using the Kirby Bauer disc diffusion method on Mueller-Hinton agar medium. The following antibiotic agents were tested: Amoxicillin/Clavulanic acid (20/10µg), Ampicillin (10µg), Ceftriaxone (30µg), Chloramphenicol (30µg), Ciprofloxacin (5µg), Clindamycin (2µg), Erythromycin (15µg), and Trimethoprim/Sulfamethoxazole (1.25µg). Other antibiotic agents that were tested included: Gentamicin (10µg), Meropenem (10µg), Nitrofurantoin (300µg), and Tetracycline (30µg). Gram-negative bacteria E. coli and Klebsiella species were screened for ESBL production using Ceftazidime (30μg) and Cefotaxime (30μg). The double-disk synergy test for Ceftazidime (30μg), Cefotaxime (30μg), and Co-amoxicillin/Clavulanate (20/10μg) were used to confirm ESBL production. MRSA was detected using Cefoxitin (30μg) disc and isolates with a zone of inhibition of ≤ 21 mm were referred to as phenotypically confirmed MRSA. Reading and interpretation of zone sizes were as per CLSI guidelines. Isolates with intermediate or resistant results on antibiotic susceptibility were classified as resistant strains during data analysis. Laboratory activities were done at a respective hospital’s laboratory as per the routine procedure. Participating laboratories were supported with reagents by the project and regular supportive supervision from the study research team.

Quality Control

Quality control strains K. pneumoniae ATCC 700603 and E. coli ATCC 25922 were used as positive and negative controls (respectively) for the detection of ESBL. E. coli ATCC 25922, S. aureus ATCC 25923 and other quality control strains were used for all laboratory methods in accordance with the CLSI guidelines.

Data Management and Analysis

Demographic information (ie, age, sex, and place of residence) was collected by laboratory technicians from the request forms and entered in an Excel spreadsheet. Data analysis was performed using STATA™ version 14.1 (Stata Corp LP, College Station, Texas, USA). Variables (ie, bacterial isolates, antibiotics susceptibility and demographic characteristics) were summarized as frequencies, percentages, medians, and inter-quartile ranges as deemed appropriate.

Results

Patients Demographic Characteristics

We processed a total of 2, 620 specimens during the observation period, but only 388 (14.8%) were culture-positive and 19 (0.7%) were discarded due to contamination. The patients’ median (IQR) age was 28 (12–44) years and females constituted 203 (52.3%) of the participants. St. Benedict Ndanda Hospital contributed the majority of the specimens (see Table 1).

Table 1 Demographic Characteristics of Study Participants (N=388)

Culture Results

Most of the collected specimens were ear pus (28.6%), urine (24.0%), wound pus (20.6%), stool (15.0%), and blood (8.3%). S. aureus was highly recovered from wound pus, ear pus, and blood (Table 2).

Table 2 Distribution of Identified Bacterial Isolates by Clinical Specimen

The most frequent isolate recovered was S. aureus (28.3%), followed by E. coli (15.2%), P. aeruginosa (10.6%), P. mirabilis, V. cholerae 01 Ogawa, Klebsiella sp. and Streptococcus sp. S. aureus isolates were prevalent among males (54.5%) and common in individuals aged 16–49 (48.6%) while E. coli was predominant in females (79.7%). P. aeruginosa was common in specimens from female patients (63.4%). Both V. cholerae and Klebsiella sp. isolates were more prevalent among males (Table 3).

Table 3 Distribution of Identified Bacterial Pathogens from Different Clinical Samples

Antibiotic Resistance Pattern of the Bacterial Isolates Identified

Table 4 reports the antibiotic resistance patterns of bacterial isolates to commonly used antibiotics tested by the participating laboratories. Gram-negative bacteria were prevalent isolates in this study. Both gram-negative and-positive demonstrated a high level of resistance to widely used antibiotics tested. E. coli isolates were highly resistant to Ampicillin (100%), Amoxicillin-Clavulanic Acid (75.0%), Gentamicin (70.2%), Tetracycline (70.2%) and Ciprofloxacin (23 [42.6%]), but least resistant to Ceftriaxone, Meropenem, and Nalidixic Acid. Only two isolates were screened for ESBL and both isolates were susceptible to Ceftazidime, hence no evidence of ESBL in the present study. Klebsiella spp. demonstrated resistance to Ampicillin (100%), Amoxicillin-Clavulanic acid (90.9%), Gentamycin (64.7%), and Ceftriaxone (55.6%), Ciprofloxacin (52.6%). Among Gram-positive, the most prevalent S. aureus was highly resistant to Erythromycin (76.3%), Gentamycin (54.0%), Ciprofloxacin (40.0%) and Clindamycin (34.9%). In addition, 22/33 (66.7%) of S. aureus isolates were resistant to Cefoxitin hence presumed MRSA.

Table 4 Drug Susceptibility Pattern of Bacterial Isolates Tested in the Laboratories

Discussion

The emergence and pervasiveness of resistant bacteria represent a substantial global public health crisis.² In the present study, we evaluated the in-vitro susceptibility patterns of bacterial pathogens to widely prescribed antibiotics. Gram-negative bacteria constituted the majority of the isolates, this corroborates with previous studies in Tanzania.^16,17 Predominant bacteria were S. aureus, E. coli, P. aeruginosa, P. mirabilis, V. cholerae 01 Ogawaa, Klebsiella sp., Streptococcus sp., Salmonella sp., Citrobacter sp., N. gonorrhoeae, S. aureus, E. coli, P. aeruginosa, Klebsiella spp., Salmonella spp., and Citrobacter spp. Most of these isolates are in the WHO’s list of priority bacterial pathogens for research, discovery and development of new antibiotics.² The pathogens exhibited a substantial high level of resistance to widely prescribed antibiotics such as Ampicillin, Amoxicillin-Clavulanic Acid, Erythromycin, Trimethoprim, Ceftriaxone, Tetracycline, Nalidixic Acid, and Chloramphenicol. Previous studies have shown a similar pattern of antibiotic resistance in Tanzania^16–18 and other parts of Africa.¹⁹

Ear pus, urine, wound pus, stool, and blood specimens were the most frequent clinical specimens analyzed. The presence of a large number of ear pus specimen points toward suppurative otitis diseases which are common in children and adolescents below 15 years especially in sub-Saharan Africa,²⁰ and their transmission is linked to poor hygienic conditions. P. aeruginosa, S. aureus, Proteus spp., and Klebsiella spp. are the most cited causative agents for otitis in sub-Sharan Africa and are reported to exhibit a high level of resistance to multiple antibiotics.²⁰ On the other hand, the majority of the participants were women of reproductive age, which explains the high number of urine specimens and E. coli processed in this study.

Resistance was also observed to third-generation cephalosporins (Ceftriaxone and Ceftazidime) and Clindamycin and Meropenem are among essential antibiotics reserved for use in tertiary hospitals listed in the WHO antibiotic stewardship programmes,^21,22 suggesting that we are precariously nearing a point of treatment option failure. The use of third-generation cephalosporins is insufficiently controlled in Tanzania and a high resistance level has been described previously.¹⁶ The observed high rate of resistance is presumably because patients may have constantly pre-exposed to a wide variety of antibiotics including self-prescribing sub-optimal dosages before their referrals to tertiary hospitals.

As this study was conducted in hospitals located in borders, the spread of AMR across neighboring countries such as Kenya, Uganda, and Burundi are highly possible. This highlights the need for coordinated actions among stakeholders including the neighboring countries for effective measures in fighting infectious antibiotic resistance.

Gram-negative bacteria such as E. coli, P. aeruginosa, P. mirabilis, V. cholerae 01 Ogawa, Klebsiella spp., and Streptococcus spp. were prevalent, similar to earlier studies conducted in Tanzania.^16–18 E. coli, and Klebsiella spp. were frequently found in urine specimens; the two are cited as the leading etiologies of urinary tract infections (UTI) especially among pregnant women.^23,24 E. coli and Klebsiella spp. are the most important bacteria associated with ESBL-mediated resistance to multiple antibiotics, including carbapenems and cephalosporins – the most potent agents for treating multi-drug resistant bacteria.² In our study, both isolates E. coli and Klebsiella spp. were highly resistant to many antibiotics tested include Ciprofloxacin, an antibiotic commonly used to treat UTI. P. mirabilis, which is also known to cause UTI, was less frequently recovered in urine, but abundant in the ear and wound swabs. The high observation of UTI causing bacteria in the current study can be described as a result of poor sanitation and hygiene practices. For better patient management, clinicians should prescribe antibiotics for UTI depending on whether the infection is uncomplicated (lower-tract UTI) or complicated (associated with a condition, such as a structural/functional abnormality of the genitourinary tract or the presence of an underlying disease).²⁵

Cholera remains a major public health problem especially in sub-Saharan Africa,^26,27 and during the implementation of this study, Tanzania was experiencing numerous waves of cholera outbreaks. Like most other bacteria of clinical relevance, cholera is continuously becoming more resistant to a wide range of antibiotics.^26–28 In the present study, cholera isolates were highly resistant to Ampicillin, Trimethoprim-Sulfamethoxazole, Nalidixic Acid, Chloramphenicol, and least resistant to Ciprofloxacin. The isolates were, however, susceptible to widely used antibiotics for treatment of cholera (Tetracycline), Ceftriaxone, Meropenem, Gentamicin, Erythromycin, and Cefoxitin. As in the case of UTI, availability of clean and safe water sources, proper sanitation and hygiene, and public health education is critical in preventing/controlling the spread of V. cholerae in communities.

The majority of P. aeruginosa isolates were recovered from wound pus and ear pus and the pathogen was highly resistant to Ampicillin, Amoxicillin-Clavulanic Acid, Gentamicin, Ceftriaxone, Tetracycline, Nalidixic Acid, and Meropenem but more susceptibility to Ciprofloxacin. Resistance to Meropenem was 66.7%, relatively higher than the 55.2% reported by Mikomangwa et al (2020)¹⁸ and 19% in a study conducted at Bugando Medical Centre using isolates collected between 2012 and 2017.²⁹ This suggests that P. aeruginosa is increasingly becoming resistant to antibiotics. N. gonorrhoeae was susceptible to Ceftriaxone but exhibited high resistance to Chloramphenicol (100%) and Ciprofloxacin (87.5%). High level of resistance to Azithromycin, Tetracycline, Ciprofloxacin, Penicillin, and to the injectable extended-spectrum cephalosporin Ceftriaxone have been described especially in regions where Gonorrhea is most prevalent.³⁰ By 2017, treatment failure to third-generation cephalosporin antibiotics had been confirmed in at least 10 countries worldwide; South Africa, Australia, United Kingdom, Canada, Japan, Norway, France, Slovenia, Sweden, and Northern Ireland.³⁰

S. aureus, a Gram-positive bacterium was the most common isolate that constituted nearly one-third of all bacterial isolates, mainly recovered from the ear and wound pus and blood specimens. Isolation of S. aureus in stool and urine was more likely due to skin contamination picked up during specimen collection. Of the isolates tested, (66.7%) were resistant to Cefoxitin and classified as MRSA infection. This observation corroborates with that recently reported at Muhimbili National Hospital and Bugando Medical Centre (61%),¹⁸ and those from other sub-Saharan Africa countries; Kenya (53.5%),³¹ Rwanda (82%)³²) and Eretria (72%).³² S. aureus isolates also exhibited resistance to Erythromycin, Tetracycline, and Gentamicin and moderate susceptibility to Ciprofloxacin, and Ceftriaxone. S. aureus isolates were mainly susceptible to Nalidixic acid, Clindamycin, and Chloramphenicol.

Limitation

Despite the current findings can be used for operational discussion on possible comprehensive national control programs on antimicrobial resistance, genotypic characterization of antibiotic resistance and virulence genes could have provided more information on the genetic profile and complement phenotypic methods. In addition, in some instances, fewer isolates were tested against specific antibiotics (eg, only two E. coli isolates were assessed for the presence of ESBL), hence hindering their generalizability.

Conclusion

Clinical isolates exhibited high resistance to most of the commonly used antibiotics. This emphasizes the need to strengthen national control programs such as routine antibiotic susceptibility testing and surveillance.