Background: Diarrhoea is considered as the most widespread illness occurring in Bangladesh. In addition, the growing antimicrobial resistance of diarrhoeal pathogens in this developing country hardens physicians to supervise patients effectively.

Objectives: In our study, we retrospectively analysed the data of a diarrhoeal pathogens and their antimicrobial resistant patterns isolated from diarrhoeal patients attending Dhaka Hospital, International Centre for Diarrhoeal Disease Research, Bangladesh (icddr, b), Dhaka and domiciliary patients of Dhaka city of Bangladesh.

Patients and Methods: During a six-year period (2009 - 2014), a total of 90207 diarrhoeal specimens were collected from patients to screen bacterial etiology by standard culture methods. Antimicrobial susceptibility testing was performed for culture-positive selected diarrhoeal pathogens using disk diffusion method.

Results: One or more potential pathogens were identified in 20467 (23%) patients, more often from under-five children. The most predominant etiological agent was Vibrio spp. (33.23%), followed by Campylobacter spp. (26.04%), Shigella spp. (19.12%), Aeromonas spp. (12.21%), Salmonella spp. (6.74%) and Plesiomonas shigelloides (2.66%). Among children under the age of five, Campylobacter spp., Vibrio spp. and Shigella spp. were more common. V. cholerae O1 El Tor biotype was the predominant pathogen (90.75%) among all the Vibrio spp. isolated. High prevalence of multiple antibiotic resistances towards ampicillin (AM), ciprofloxacin (CIP) and co-trimoxazole (SXT) was common among Vibrio spp., Shigella spp. and Campylobacter spp. Beside, correlation with previous data from 2005 - 2008 showed that resistant percentages of Shigella spp., Campylobacter spp., Salmonella spp. to these antibiotics are increasing with time (P < 0.001).

Conclusions: These findings underscored the importance of monitoring and developing guidelines for better management of infectious diarrhoea in Bangladesh and elsewhere.

Infectious diarrhoea remains a very serious public health problem in low and middle-income countries (LMICs) including Bangladesh and is the leading cause of morbidity and mortality among young children less than five years of age (1). The emergence of resistance to different antimicrobial agents among bacterial pathogens associated with diarrhoea e.g. Shigella spp., Campylobacter spp., Salmonella spp., Aeromonas spp., pathogenic Escherichia coli and Vibrio spp. is a global concern. Although most diarrhoea episodes are self-limiting, the use of appropriate antibiotics can reduce the duration of diarrhoea, limit the progression of the disease, and lessen the severity of related symptoms such as fever, abdominal pain and vomiting. Another prime focus for antibiotic therapy is to reduce the chance of excretion of the causative agents and thereby the likelihood of transmission among child-care workers, medical professionals, and workers of the catering industry or services (2).

Access to clean water supply, promoting good personal hygiene, proper nutrition, and improvement of sanitary conditions would definitely have the greatest impact on diarrhoeal diseases management. However, the development of effective vaccines against the huge number of enteric pathogens is a serious challenge (3, 4). Anti-dehydration therapy associated with adequate nutritional support and intravenous fluid therapy are the best choices of treatment for diarrhoea, regardless of the etiology and the severity of the disease (5).

The gradual increase in antimicrobial resistance among enteric pathogens in LMICs is a critical area of concern and access to current antimicrobial susceptibility data is of importance to guide treatment decisions (5). Knowledge of long-term surveillance of bacterial etiology for diarrhoea and antimicrobial resistance is essential to make wise decisions for clinicians to prescribe effective antimicrobial therapies. Previously, we reported on bacterial etiology and their resistance patterns for the period of 2005 - 2008 (6).

The aim of the present study was to follow up the trends in potential bacterial pathogens isolated from diarrhoeal patients and the antimicrobial susceptibility patterns of the isolated organisms over the six-year period from 2009 - 2014.

From January 2009 to December 2014, a total of 90207 stool samples and rectal swabs were received from hospitalized (inpatients) and domiciliary (outpatients) diarrhoeal patients of all ages at Clinical Laboratory Services of International Centre for Diarrhoeal Disease Research, Bangladesh (icddr, b), Dhaka, Bangladesh. There were 14741, 16250, 13909, 14355, 14739 and 16213 samples in 2009, 2010, 2011, 2012, 2013 and 2014, respectively. All these samples were tested for the presence of Shigella, Salmonella, Vibrio, Aeromonas, and Plesiomonas. Antimicrobial susceptibility tests of the isolates were performed. Presence of Campylobacter species was screened in 71389 samples (9281, 10818, 10934, 11962, 12894 and 15500 samples in six consecutive years). All the data were analysed retrospectively for the bacterial pathogens present in the stool samples and their antibiotic resistant pattern over the study period from 2009 to 2014.

3.1. Bacteriological Isolation

The collected stool samples and rectal swabs were directly inoculated onto MacConkey agar, Salmonella-Shigella (SS) agar, Taurocholate-Tellurite-Gelatin agar (TTGA), and Brucella agar plate (BAP) with 5% sheep blood. All the plates were incubated at 37°C overnight, except for BAP which was incubated at 42°C in microaerophilic condition for at least 48 hours using Campygen gas pack (Oxoid, United Kingdom). For enrichment of Salmonella spp. Vibrio spp. and Aeromonas spp., the stool specimens were also inoculated into selenite cystine broth (Difco, BD) and bile peptone broth respectively. After overnight incubation at 37°C, the inoculated broths were subcultured onto SS agar and TTGA plates respectively. Shigella spp., Salmonella spp., Aeromonas spp., Vibrio spp., Campylobacter spp., Plesiomonas spp. were subsequently identified by standard bacteriological methods (7, 8) and API 20E biochemical profiles (bioMe´rieux, France) when necessary. The isolates were further confirmed serologically using commercially available antisera (Denka Seiken, Japan).

3.2. Susceptibility Testing

Antibiotic susceptibility tests of the isolates were performed using Kirby-Bauer disc diffusion method on Muller-Hinton agar plates following the guidelines of the Clinical and Laboratory Standards Institute (7). The antimicrobial agents tested for the isolates were: ampicillin (AM; 10 µg), ciprofloxacin (CIP; 5 µg), ceftriaxone (CRO; 30 µg), chloramphenicol (C; 30 μg), erythromycin (E; 10 μg), mecillinam (MEL; 25 µg), nalidixic acid (NA; 30 µg), cotrimoxazole (SXT; 25 µg), and tetracycline (TE; 30 µg). The interpretation of antimicrobial susceptibility for Vibrio spp. and Campylobacter spp. was performed as described previously (6). E. coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853 were used as control strains for the susceptibility tests. The disks were obtained from Oxoid (Basingstoke, UK).

3.3. Statistical Analysis

Statistical analysis was performed using Statistical Package for Social Sciences (SPSS) version 20.0 (IBM, USA). Crude odd ratios or relative risks were calculated using the χ² (chi-squared) method; two-tailed P values ≤ 0.05 were considered to be statistically significant. The χ² test for trend was considered to be statistically significant if P value was ≤ 0.05.

A total of 90207 stool samples or rectal swabs were received from diarrhoeal patients aged from 1 day to 102 years with a mean age of 15 years over a span of six years (January 2009-December 2014). Demographic analysis showed that 52372 (58%) patients were aged < 5 years and the male-to-female ratio was 1.46/1.0. The age distribution and the trends of gender distribution remained stable over the study period. Of the tested samples, 28.11% (4143), 29.41% (4784), 19.60% (2726), 17.59% (2525), 18.62% (2745), and 21.89% (3544) were found to be culture-positive in 2009, 2010, 2011, 2012, 2013, and 2014, respectively (P < 0.001) and 22.68% (20467) of all the received samples were culture-positive. The rates of isolation of different bacterial pathogens are summarized in Tables 1 and 2. Among the isolated pathogens, Vibrio spp. were the most predominant aetiological agents (33.23%), followed by Campylobacter spp. (26.04%), Shigella spp. (19.12%) and Aeromonas spp. (12.21%). Other frequently isolated pathogens included Salmonella spp. (6.74%) and Plesiomonas shigelloides (2.66%). Among the isolates from the under-five-year-old children, the incidence of Campylobacter spp. was 7.01%, followed by Shigella spp. (5.01%), Vibrio spp. (4.45%), Aeromonas spp. (2.56%), Salmonella spp. (1.31%), and P. shigelloides (0.46%) (Table 3).

Table 1. Bacterial Pathogens Isolated From Diarrhoeal Patients in Bangladesh From January 2009 to December 2014 and Comparison With the Earlier Study Period From 2005 to 2008a

Table 2. Bacterial Pathogens Isolated From Diarrhoeal Patients in Bangladesh From January 2009 to December 2014

Table 3. Age Group-Wise Isolation Rates of Bacterial Pathogens in Diarrhoeal Patients in Urban Dhaka, Bangladesh During 2009 – 2014

Among the total bacterial isolates, the frequency of Vibrio spp. was 36.79%, 35.97%, 30.55%, 36.08%, 35.22% and 23.36% in 2009, 2010, 2011, 2012, 2013 and 2014, respectively (χ² for trend from 2009 to 2014 = 219.212, P < 0.001). Of these Vibrio spp., V. cholerae O1 El Tor biotype was 90.75%, followed by V. cholerae non-O1, non-O139 (7.08%), V. parahaemolyticus (1.91%), V. fluvialis (0.13%). and V. alginolyticus (0.04%). V. cholerae O1 affected more patients aged ≥ 5 years than children aged < 5 years (4101 [66%]: 2076 [34%]; P < 0.001). V. cholerae O1 El Tor Ogawa serotype was predominant throughout the study period (Table 2) and had distinct seasonality with peaks during March-May and in October, whereas V. cholerae non-O1, non-O139 was prevalent throughout the year (Figure 1).

Figure 1. Distribution of V. cholerae O1 and V. cholerae Non-O1 and Non-O139 During the Study Period (January 2009 to December 2014)

Isolation variation was observed for Aeromonas spp. from 2009 to 2014. From 2009 to 2012, isolation showed a decreased pattern from 12.6% in 2009 to 6.8% in 2012 with a slight increase (13.8%) in 2010, but after 2012, isolation of Aeromonas spp. increased remarkably compared to the total isolates (9.7% in 2013 and 16.6% in 2014) (χ² for trend from 2009 to 2014 = 167.293, P < 0.001). A total of 5330 Campylobacter spp. were identified from the tested samples, and along the consecutive study years from 2009 to 2012, there was a sharp decrease in isolation of Campylobacter species with 9.8%, 9.7%, 7.4% and 5.2% of the tested samples. However, slight increased isolation rate of Campylobacter spp. was observed with 5.7% and 7.9% of the total tested samples in 2013 and 2014 (χ² for trend from 2009 to 2014 = 233.58, P < 0.001). Salmonella spp. showed a rising trend of isolation rate from 2009 to 2012, except in 2010 (6.9%, 5.8%, 7.6% and 9.2% of the total isolates in 2009, 2010, 2011 and 2012, respectively), and the isolation frequency decreased sharply in 2013 (6.5%) and 2014 (5.5%) (χ² for trend from 2009 to 2014 = 43.66, P < 0.001). Non-typhoidal Salmonella spp. were more frequently isolated (998 [72.3%]) than typhoidal Salmonella spp. (382 [27.7%]) (P < 0.002). There was a sharp decline in the isolation of P. shigelloides from 4.1% in 2009 to < 1.0% in 2014 (χ² for trend from 2009 to 2014 = 262.81, P < 0.001).

Among the V. cholerae O1 isolates, all the isolates showed reduced susceptibility to E, whereas resistance to SXT and TE was 99.0% and 61.0%, respectively. V. cholerae non-O1, non-O139 isolates showed reduced susceptibility to E (93.0%), but only 41.0%, 14.0% and 1.0% of the isolates were resistant to SXT, TE and CIP, respectively (Table 4). An increasing trend of resistance to CIP, NA and AM was observed in Shigella species. Overall, 78.0%, 64.0%, 58.0% and 31.0% of the S. flexneri isolates showed resistance to NA, SXT, AM, and MEL, respectively. Resistance to CIP in S. flexneri increased from 37.0% of the isolates in 2009 to 63.0% in 2014 (P < 0.001), whereas for S. sonnei, resistance increased up to 80.0% in 2014 from 18.0% in 2009 (P < 0.001). Overall, 96.0%, 58.0% and 57.0% of S. sonnei, S. boydii, and S. dysenteriae (not type 1), respectively, showed resistance to NA, while SXT resistance was 96.0%, 60.0%, and 56.0%, respectively (Table 5). Around 57.0% and 22.0% of the Aeromonas spp. showed resistance to SXT and TE, respectively, and resistance to CIP was relatively lower (< 7.0%) (Table 6). Among Salmonella spp., S. paratyphi showed increased susceptibility to AM (91%), CIP (99%), C (99%) and SXT (97%), but S. typhi showed 29%, 14%, 33% and 35% resistance to these antibiotics, respectively. Non-typhoidal Salmonella spp. were 93% susceptible to CIP and 96% to C (Table 7). Forty-six percent and 34.0% of the Campylobacter isolates were resistant to AM and TE, while resistance to CIP and E were above 94.0% and less than 15.0%, respectively (Table 8).

Table 4. Comparison of Percentage of Antimicrobial Resistance in Vibrio cholerae O1 and Non-O1, Non-O139 Isolates From Diarrhoeal Patients in Bangladesh Between the Recent (2009 - 2014) and Previous (2005 - 2008) Study Periods

Table 5. Comparison of Percentage of Antimicrobial Resistance in Shigella spp. Isolates From Diarrhoeal Patients in Bangladesh Between the Recent (2009 - 2014) and Previous (2005 - 2008) Study Periods

Table 6. Comparison of Percentage of Antimicrobial Resistance in Aeromonas spp. Isolates From Diarrhoeal Patients in Bangladesh Between the Recent (2009 - 2014) and Previous (2005 - 2008) Study Periods

Table 7. Comparison of Percentage of Antimicrobial Resistance in Salmonella spp. Isolated From Diarrhoeal Patients in Bangladesh Between the Recent (2009 - 2014) and Previous (2005 - 2008) Study Periods

Table 8. Comparison of Percentage of Antimicrobial Resistance in Campylobacter spp. Isolated From Diarrhoeal Patients in Bangladesh Between the Recent (2009 - 2014) and Previous (2005 - 2008) Study Periods

The overall follow-up finding of the study gave in depth information regarding the trends of bacterial etiology of diarrhoea and antimicrobial resistance profiles of major diarrheagenic bacteria in urban areas of Bangladesh and a comparison with our previous study during 2005 - 2008 (6).

Overall, the frequency of culture-positive samples remained nearly equal over the period, as in our earlier study where 25% of all the received samples were culture-positive for bacterial pathogens (6, 9-11). Gender distribution was slightly biased toward male patients (60%) and the gender biasness is congruent with other studies. This gender discrimination in health care seeking behaviour is in accordance with other studies (9, 12) and this is due to the existence of strong son-preference in Bangladesh. In addition, younger children are more prone to be infected by diarrheagenic pathogens due to their dietary habits and immaturity of their immune systems. Moreover, since diarrhoeal diseases are self-limiting, adult patients are more reluctant to seek medical help or visit a doctor in small clinics or collect medicines over the counter. As a result, the data of this study might be biased, because mostly, diarrhoea cases of children aged < 5 years were included.

Overall, most of the bacterial pathogens showed unpredictable natures in isolation during the study period. In the current study, Vibrio spp. showed a declining trend in isolation in first three consecutive years (2009, 2010, and 2011) and after an increase in 2012, a fall in isolation was observed in later years. Likewise our earlier report during 2005 - 2008 (6) and a report from India (13), our study showed predominant isolation of V. cholerae serogroup O1 El Tor biotype over the six-year period and patients aged ≥ 5 years continued to have the highest incidence of V. cholerae O1 infection. The average frequency of isolation of V. cholerae O1 than non-O1, non-O139 was higher during March - May and in October in each year.

Shigella spp. was the second most predominant bacterial agent with a fluctuating increase (2009 - 2012) and decrease (2013 - 2014) trend in isolation throughout the study period. Of the isolated Shigella spp., S. flexneri was the predominant serogroup, followed by S. sonnei, S. boydii and S. dysenteriae. It is interesting to note that S. sonnei emerged as the second predominant by overtaking S. boydii and S. dysenteriae since 2009. The present study supports the findings of other investigators from Bangladesh (9, 14) and also from our neighbour countries including India (15) and Pakistan (16). Just as the isolation of atypical Shigella spp. from diarrhoeal patients in Bangladesh, the frequency of nontypeable Shigella spp. had also been reported earlier from India (11). It has been hypothesized that altering trend in Shigella serogroups might be due to improvement of host nutritional status (17, 18), improvement of socioeconomic status, meteorological changes, sanitation condition, change in genetic makeup of bacteria, and reduction of the chances of cross immunity by P. shigelloides (14, 19, 20).

A steady decreasing trend of Campylobacter spp. isolation from 2009 to 2012 was observed in the present study which is analogous to our previous study (6). However, in the last two years (2013, 2014) of our study, the isolation rates of Campylobacter spp. showed increase. Development in laboratory techniques and improvement of isolation skills might have had an indirect influence on the increased isolation of Campylobacter spp. over the study period.

There was a raising trend in the isolation of Salmonella spp. except in 2010, from 2009 to 2012. After 2012, Salmonella spp. isolation rates showed decreasing trend, which is similar to our earlier study during 2005 - 2008 (6). The distribution of S. typhi showed a decreasing trend, whereas the frequency for nontyphoidal Salmonella spp. and S. paratyphi (A and B) increased in comparison with the earlier report (6). The isolation frequency of Aeromonas spp. decreased in the current study period in comparison to our previous report (6). Although the real contributing factor for the observed shifting trend is unclear, it might be due to an increased awareness in the urban population of infection risks as well as consequent improvements in hygiene and maintenance of good personal hygiene (6).

Antimicrobial resistance has emerged as a serious public health problem worldwide. In the present study, V. cholerae isolates were frequently resistant to SXT and TE; however, most of the isolates showed reduced susceptibility to E and were sensitive to CIP, which is in accordance with our earlier report (6). Similar resistance patterns to these antibiotics were also reported from India except for E, for which they found only 10% of the isolates resistant (11). Among all the Shigella isolates, 77% and 72% of S. flexneri were resistant to NA and SXT, respectively. The emergence of fluoroquinolone-resistant S. sonnei (62%) was followed by S. flexneri (47%), which restricted the choice of fluoroquinolone for treatment. These findings are in agreement with the data published from other parts of the world (21, 22). Susceptibility to TE, SXT and CIP in Aeromonas spp. was nearly stable during 2005 - 2014, whereas reduced susceptibility to E increased significantly during 2009 - 2014 (P < 0.001). A relatively lower level of resistance towards CIP, SXT and TE was reported from Australia (23). The resistance rates to NA, AM, SXT and C of Salmonella spp. generally decreased in the current period than that of 2005 - 2008, whereas the opposite trend was noted for CIP resistance (6) and our data is comparable with a report from India (24). A possible explanation for of the differences in susceptibility to different antibiotics might indicate the infrequent use of antibiotics.

CIP-resistant Campylobacter spp. was noted significantly higher in the present study in comparison to our previous report (6). Similar to our study, increase of CIP resistance among Campylobacter spp. has also been reported by other investigators (25, 26). These findings suggest the ineffectiveness of CIP as a drug of first choice for empirical treatment of campylobacteriosis. The significant increase of resistant to E among Campylobacter isolates limits the empirical choice of antibiotic for the treatment of campylobacteriosis in Bangladesh.

The emergence of antibiotic resistance is a continually evolving and dangerous concern which requires immediate attention and future planning to combat a global public health crisis. Multiple factors are responsible for the alarming rise in the occurrence of antimicrobial-resistant bacteria such as overuse and misuse of antibiotics, like erroneous antibiotic prescriptions for nonbacterial infections, and the incorporation of antibiotics to livestock feed and food sanitizing agents (6, 20). Prolonged survival of patients with chronic disease, nosocomial infection by multiply antibiotic-resistant (MAR) bacteria (27), increased number of immunosuppressed individuals, substandard hospital hygiene, and more international travel contribution for enhanced dissemination of resistance factors (28). The selective pressures applied by antibiotic drugs creates reservoir of antibiotic-resistant bacteria in the environment and has potentially lethal consequences for invasive diarrhoea endemic regions (20).

Versatile approaches including raising awareness about the judicious, prudent, or rational use of antibiotic are in need to prevent the spread of antibiotic resistance among enteric pathogens. Physicians should encourage patients to start antibiotic therapy after culture and sensitivity results have been obtained and patients should complete the full courses of antibiotics. The use of effective infection-control practices to prevent dissemination to healthy people, surveillance of antimicrobial resistance and antimicrobial use, and improved use of vaccination may circumvent the necessities of antimicrobial therapy. Development of alternative therapies like bacterial interference using nonpathogenic (commensal) bacteria, bacteriophage therapy, use of cationic peptides, and synthetic and amphipathic cyclic D, L-a-peptides, can also be alternatives to extensive antimicrobial use.

Updated follow-up study report is very much essential in regards with promoting specific and actual line of therapy, for the diarrhoeal disease management and to reduce the health care costs in LMICs like Bangladesh. Similarly, this study will help physicians to have the overview knowledge of antimicrobial resistance of different diarrhoeal pathogens to guide the patients successfully. Although having several advantages, this data may mislead physicians who prescribe antibiotics against unusual multiple antibiotic-resistant bacteria based on this data. After all, molecular characterization of bacterial isolates and a better antibiotic susceptibility test will help to overcome the limiting factors of such study in future.

This study was funded by International Centre for Diarrhoeal Disease Research, Bangladesh (icddr,b) and its donors which provide unrestricted support to icddr,b for its operations and research. Current donors providing unrestricted support include: Government of the People’s Republic of Bangladesh, Canadian International Development Agency (CIDA), Swedish International Development Cooperation Agency (Sida), and the Department for International Development, UK (DFID). The authors gratefully acknowledge these donors for their support and commitment to icddr,b’s research efforts. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.