Salmonella enteritidis diarrhoea in Harare, Zimbabwe
correspondence Dr Clifford Simango, Department of Medical Laboratory Sciences, University of Zimbabwe Medical School, PO Box A178, Avondale, Harare, Zimbabwe
Summary The study was conducted to determine the prevalence of Salmonella enteritidis diarrhoea in an urban area in Zimbabwe. Stool specimens from people of all ages presenting at primary level health centres in Harare were investigated for S. enteritidis and other bacterial and parasitic enteric pathogens. The first 46 S. enteritidis isolates were phage-typed, and all isolates were tested for susceptibility to ampicillin (10 μg), chloramphenicol (30 μg), cotrimoxazole (25 μg), tetracycline (30 μg), gentamicin (10 μg), nalidixic acid (30 μg), ciprofloxacin (5 μg) and ceftriaxone (30 μg). S. enteritidis was isolated in 74 (1.8%) of 4155 stool specimens which represented 30.7% of all Salmonella species isolated. The most common S. enteritidis phage type was 4 (78.3%) followed by 7, 9 and 23 (8.7%, 2.2%, 2.2%, respectively) All S. enteritidis isolates were sensitive to gentamicin, ciprofloxacin and ceftriaxone. Less than 10% of the isolates were resistant to the other antimicrobials except ampicillin, to which 13.5% were resistant. One isolate was resistant to ampicillin, chloramphenicol, cotrimoxazole and nalidixic acid.
Human infections with Salmonella enteritidis have been increasing wordwide since 1980. S. enteritidis is an important serotype of Salmonella comprising 37% of Salmonella isolates in Maryland since 1990 ( Morris et al. 1992 ) and 28% in the north-eastern United States in 1986 ( St Louis et al. 1988 ). It is normally found in the intestinal tracts of humans and of animals such as chickens and domestic livestock and remains an important cause of gastroenteritis in humans worldwide. It is usually acquired by ingestion of contaminated water and food, mainly poultry, eggs and egg products ( Humphrey et al. 1988, 1989 ; St Louis et al. 1988 ; Cowden et al. 1989 ; Hedberg et al. 1993 ). Eggs are the most probable primary source of contamination in many outbreaks of S. enteritidis infection in developed countries ( St Louis et al. 1988 ).
Phage-typing of S. enteritidis in the United States showed that type 8 strains are the most common ( Hickman-Brenner et al. 1991 ; Usera et al. 1994 ). In Europe the most common phage type is 4 ( Ward et al. 1987 ; Cowden et al. 1989 ; Stanely et al. 1992 ; Schroeter et al. 1994 ). Data on the prevalence of S. enteritidis as a causative agent of diarrhoea in developing countries are scarce. We aimed to determine the prevalence of Salmonella enteritidis as a causative agent of diarrhoea in an urban area in Zimbabwe.
Materials and methods
The study was conducted over a period of 8 months during the hot and wet season (October – May), when diarrhoea caused by bacteria peaks in Zimbabwe ( Cruickshank & Zilberg 1976). We investigated stool specimens from all people with diarrhoea presenting as outpatients at primary level health centres in the city of Harare. Specimens were collected in sterile containers and processed within 5 h.
The specimens were cultured for Salmonella, Shigella and Campylobacter species, enteropathogenic Escherichia coli and Yersinia enterocolitica. We also looked for Giardia lamblia, Cryptosporidium and Entamoeba histolytica parasites. Xylose-Lysine-Desoxycholate agar medium was used for culturing Salmonella species. Inoculated agar plates were incubated at 37 °C for 24 h. Colonies which appeared pink-red with a black centre were presumptively identified as Salmonella species and biochemically tested ( Cheesbrough 1985). Isolates which gave positive reactions for Salmonella were serotyped using polyvalent and monovalent antisera (Wellcome Diagnostics, Dartford, UK). The first 46 S. enteritidis isolates were sent to the Instituto Superiore della Sanità in Italy for phage-typing. Other bacterial enteric pathogens were identified by standard procedures ( Cheesbrough 1985). The formol-ether concentration technique ( Allen & Ridley 1970) was used for parasites and staining with modified Ziehl Neelsen stain ( Cheesbrough 1987) for Cryptosporidium oocysts.
The Kirby–Bauer disc diffusion method was used to test for susceptibility to ampicillin (10 μg) chloramphenicol (30 μg), cotrimoxazole (25 μg), tetracycline (30 μg), gentamicin (10 μg), nalidixic acid (30 μg), ciprofloxacin (5 μg), and ceftriaxone (30 μg) using Mueller Hinton agar medium.
We processed 4155 stool specimens for the detection of enteric pathogens. Their distribution is shown in Table 1. Shigella species were the most common bacterial enteric pathogens, followed by Salmonella and then enteropathogenic Escherichia coli. S. enteritidis comprised 30.7% of the Salmonella isolates. Giardia lamblia was more common than Cryptosporidium. Multiple infections were observed in 23 specimens but only one specimen had S. enteritidis isolated together with another pathogen.
Incidence of bacterial and parasitic enteric pathogens in isolates from patients with diarrhoea
The distribution of S. enteritidis by age group is shown in Table 2. The highest incidence was observed in people > 50 years (3.8%), the lowest in children < 10 years (1.2%). The most common S. enteritidis phage type was 4 (78.3%) followed by types 7, 9 and 23 ( Table 3). All S. enteritidis isolates were sensitive to gentamicin, ciprofloxacin and ceftriaxone as shown in Table 4. Less than 10% of the isolates were resistant to chloramphenicol, nalidixic acid, cotrimoxazole and tetracycline. The highest resistance was observed against ampicillin with 13.5% of the isolates being resistant to this drug. About 19% of the S. enteritidis isolates were resistant to one or more drugs ( Table 5). One isolate showed multidrug resistance to ampicillin, chloramphenicol, cotrimoxazole and nalidixic acid.
Age distribution of patients with diarrhoea due to Salmonella enteritidis
Distribution of phage types of Salmonella enteritidis
Antimicrobial susceptibilities of Salmonella enteritidis
Patterns of resistance in Salmonella enteritidis
Salmonella enteritidis infections are on the increase in many developed countries and have been associated with consumption of contaminated poultry, eggs and egg products. We found that S. enteritidis is the most common Salmonella serotype in diarrhoeal stool specimens, accounting for 30.7% and detected in 1.8% of all specimens cultured. Similar observations were noted by Morris et al. (1992) in the United States where S. enteritidis accounted for 37% in Maryland, and by St Louis et al. (1988) , who found that it comprised 28% of Salmonella isolates in the north-eastern United States.
We found S. enteritidis phage type 4 to be the most prevalent, which agrees with findings in Europe ( Ward et al. 1987 ; Cowden et al. 1989 ; Stanely et al. 1992 ; Schroeter et al. 1994 ) as well as Japan ( Suzuki et al. 1995 ) and Thailand ( Boonmar et al. 1998 ) since 1980. The predominance of this phage type in many countries suggests that S. enteritidis phage type 4 strains may possess unique characteristics and pose a great risk to public health in some regions of the world. Phage type 8, the most common in the United States ( Hickman-Brenner et al. 1991 ) and Canada ( Khakhria et al. 1991 ), was not found in our study. Phage type 7 was the second most common type among Zimbabwean isolates but was comparatively rare.
Some studies have implicated poultry and poultry products as the main source of human infection with S. enteritidis ( St Louis et al. 1988 ; Humphrey et al. 1988, 1989 ; Cowden et al. 1989 ; Hedberg et al. 1993 ). A study in Zimbabwe (unpublished data), which examined 113 S. enteritidis isolates in chickens and eggs from areas around Harare, identified phage type 4 as the most common (77.9%) followed by phage type 7 (14.2%), with a similar distribution pattern to that of human isolates observed in this study. The similarity of phage type distributions in humans and chickens suggests that chicken and egg consumption is an important cause of S. enteritidis, and that phage type 4 is the most important in humans and chickens in Zimbabwe. The other phage types were detected in comparatively low rates and appear to play a minor role in infections of both humans and chickens.
Although antimicrobial drugs are not recommended for the treatment of uncomplicated salmonellosis in humans, they are essential for the treatment of extraintestinal S. enteritidis infections such as septicaemia, endocarditis, empyema and meningitis. A few isolates of S. enteritidis in our study were resistant to one or more of the eight antimicrobial drugs tested, with one isolate displaying multidrug resistance to ampicillin, chloramphenicol, cotrimoxazole and nalidixic acid. Resistance to antimicrobial drugs in S. enteritidis isolated from humans has been shown to be low, as confirmed in our study. All S. enteritidis isolates were sensitive to ciprofloxacin, gentamicin and ceftriaxone; hence these are potentially useful in cases showing resistance to the antimicrobials. However, S. enteritidis resistance to some antibiotics is growing: a substantial increase in the incidence of resistance to ampicillin was observed in England and Wales ( Ward et al. 1990 ), and Tassios et al. (1997) found a high proportion of S. enteritidis to be resistant to ampicillin and doxycycline in Greece. The increase in resistance of S. enteritidis to ampicillin is due to a 34-MDa conjugative plasmid ( Vatopoulos et al. 1994 ) which is easily transferred to recipient cells ( Tassios et al. 1997 ). Multiple resistance has remained rare amongst human isolates of S. enteritidis, as was also observed in the present study, where only one isolate displayed resistance to four antimicrobial drugs.
We would like to thank the Istituto Superiore della Sanità in Italy for phage-typing the Salmonella enteritidis isolates and Dr Makaya of the Veterinary Laboratory Services in Zimbabwe, who made the necessary arrangements.