Epidemiological, genetic, serological and antimicrobial susceptibility characterization of non-typhi Salmonella isolates from hospitalized children in Nairobi, Kenya
Non-typhi Salmonella (NTS) are a major cause of septicaemia and bacteraemia in humans worldwide. About 1.4 million cases are observed in the United States of America annually, out of which 600 are fatal. In sub-Saharan Africa, NTS are among the most common causes of invasive bacterial childhood disease. Multi-drug resistant NTS species have been isolated in Kenya and pose a major public health concern. This study was conducted to investigate the NTS that cause bacteraemia in children admitted in hospitals in Nairobi, Kenya. The objectives of the study were to investigate the characteristics of the households of NTS cases of children admitted in three major Nairobi hospitals; to determine the antigenic and genetic characteristics of the NTS isolates from the admitted children; arid to determine the susceptibilities of the isolates to commonly used antimicrobial agents in Kenya. The NTS isolates were obtained from children hospitalized in three hospitals, in Nairobi Kenya namely, Kenyatta National Hospital, the Aga Khan Hospital and Gertrudes' Garden Children's Hospital between June and November 2002. The identity of the isolates was confirmed by biochemical tests using the Analytical Profile Index (API) 20E system (API System Motalieu Vercieu, France) and serotyping using specific agglutinating antisera (Murex Diagnostics, Dartford, UK). Homes of the cases were visited in order to administer a questionnaire designed to capture the characteristics of the households that could favour the transmission and maintenance of the NTS infections. Informed consent to visit the homes was always obtained fromthe guardians before the visits. In addition, drinking water samples, and stool samples from the household mates of the cases were collected for culture and isolation ofNTS. Xll Antimicrobial susceptibility tests to commonly available antimicrobials were performed for 83 isolates on Isosensitest agar using the disk diffusin technique. The antimicrobial agents included ampicillin l Oug, tetracycline 30l-lg, trimethoprim Sug, chloramphenicol 30l-lg, streptomycin lOl-lg,gentamicin l Oug, ceftazidime 30l-lg, ciprofloxacin Sug, cefuroxime 30l-lg, amoxicillinclavulanic acid 30l-lg, sulphamethoxazole l Ouug and nalidixic acid l Oug. Minimum inhibitory concentrations (MICs) for the commonly used antimicrobial agents were also determined for 50 isolates using the agar dilution technique. Plasmid DNA profiling as a method of strain differentiation was carried out. Plasmids were extracted using a commercial kit, Plasmid Mini Prep-Spin Kit (Qiagen, West Sussex, UK). The majority (77.4%) of the NTS cases were from households in slum areas with very poor sanitary conditions. Over 75% of the NTS isolates were from children aged below 5 years. Seventy-four isolates from the hospital index cases were confirmed as NTS. Nine NTS isolates were recovered from the stool samples from household mates of the index cases. S. enteritidis was the most common (37.8%) serotype among the NTS bacteraemia cases in hospitalized children followed by S. typhimurium (14.9%). Salmonella gallinarum-pullorum was isolated from one case (1.4%). The other 45.9% of the case isolates could not be serotyped conclusively due to lack of specific antisera. All the S. enteritidis and S. typhimurium isolates were resistant to one or more of the commonly used antimicrobials. A high prevalence of multi-drug resistance (MDR) among the 74 NTS from children was recorded. Only 2 (2.7%) of these were susceptible to all the 12 antimicrobials tested; twenty-nine (39.2%) were resistant to five or more of the antimicrobials; and 7 (9.5%) were resistant to eight or more antimicrobials tested. Plasmids of sizes ranging from 2 - 105 Kb were extracted from 16 of the total 83 isolates. Presence of Xlll plasmids was associated (p<O.05) with resistance to five or more antimicrobials. Plasmid 105Kb was significantly associated (p<O.05) with resistance to streptomycin while plasmid 8Kb was significantly associated (p<O.05) with resistance to both streptomycin and trimethoprim. Nontyphi Salmonella were not isolated from the 31 drinking water samples collected but were isolated from 9 of 123 (7.3%) stool samples from household mates of the NTS cases. Two of the stool sample isolates were different from those from index cases by their serotype and also by their antimicrobial resistance patterns while the rest were similar in both characteristics This study showed that NTS affects mostly children aged less than 5 years and that antimicrobial resistance in the NTS is very common. The study also found resistance to antimicrobial agents to be associated but not restricted, to presence of plasmids. Presence or absence of plasmids was not related to virulence. In addition, poor sanitary conditions appeared to favour the maintenance of NTS. Improvement of sanitary conditions may therefore reduce the number of infections. Better studies, preferably case-control study types, are required to confirm the association of the potential risk factors ofNTS in this high-risk age group of children. Antimicrobial abuse, overuse and misuse in both human and veterinary medical practices may be the cause of the antimicrobial resistance observed among the NTS. Enforcement of the law (Cap. 244, Pharmacy and Poisons act) that regulates the use of the antimicrobial agents is therefore highly recommended as a way of dealing with the problem of antimicrobial resistance. XlV 104 strain, which has been responsible for epidemics, particularly in Europe, USA and Canada, has reservoirs in cattle, and is transmitted mainly through consumption of contaminated meat, milk and milk products (Glynn et al., 1998; Fey et al., 2000; Threlfall, 2000; Murphy et al., 2001). Animal carriers are also an important source of infection (Wiktor and Van Oye, 1955; Plowright 1957; Stewart, 1957). In the USA, about 95% of NTS infections are related to food-borne transmission including contaminated poultry and poultry products, meat, milk and other dairy products (Cody et al., 1999; Mead et al., 1999; Ahmed et al., 2000; Threlfall, 2000). Between 1985 and 1991, 82% of S. enteritidis outbreaks in the USA that could be attributed to a specific food vehicle were associated with eggs (Mishu et al., 1994). As the incidence of S. enteritidis infection In chicken has increased in that country, so has the incidence of S. enteritidis infection In humans (St. Louis et al., 1988; Mishu et al., 1994). In most industrialized countries and also some less industrialized countries, S. enteritidis is transmitted through consumption of foods containing raw or incompletely cooked eggs and home cooked products containing eggs (Ahmed et al., 2000; Palmer et al., 2000; Indar-Harrinauth et al., 2001; Liebena et al., 2001). Laboratory studies have confirmed that tomatoes and other fresh fruits and vegetables can support the growth of salmonella (Asplund and Nurmi, 1991) and are therefore an important source of infection. Animal feeds have long been recognized as an important source of new Salmonella infections in livestock farms (Muller, 1957; Shapcott, 1984). Salmonella serotypes not previously detected in flocks of a broiler company in Australia were detected first in raw feed ingredients, and later appeared in live birds and processed carcasses (Mackenzie and Bains, 1976). Cox et al. (1983) collected poultry feed from commercial mills in the USA and found Salmonella in 92% of meat and bone meal samples and 58% offinished feed (mash) samples, but none in pelleted feed. A recent study in Nairobi, Kenya, conducted to investigate the likely sources of NTS in humans did not show animals as the reservoirs (Kariuki et al., 2002). Multi-drug resistant NTS have been isolated in Kenya, where resistance to two or more antibiotics including ampicillin, co-trimoxazole, streptomycin, tetracycline and chloramphenicol has been reported (Kariuki et al., 2000; Oundo et al., 2000). Sources ofNTS that cause bacteraemia in children in Kenya have not been investigated. This study was carried out to investigate NTS occurring In children admitted in major hospitals in Nairobi, Kenya. The general objective of the study was to determine the risk factors for invasive Salmonella infections in children and to determine the antigenic and genetic characteristics of the NTS involved. The specific objectives were: l.To investigate the characteristics of households of NTS cases of children admitted in three major Nairobi hospitals, that may be risk factors; 2. To determine the antigenic and genetic characteristics of isolates of NTS from cases of children admitted to hospitals in Nairobi, Kenya; 3.To determine the susceptibilities of the isolates from 2) to commonly used antimicrobial agents in Kenya.