The term nontyphoid salmonellosis refers disease states in which nontyphoid Salmonella is isolated from freshly passed stool or blood culture (1). The incidence of nontyphoid Salmonella infections has been increasing in Taiwan, despite a general improvement in sanitary conditions during the past decade (2). Nontyphoid Salmonella is the leading cause of enterocolitis in Taiwan, and antimicrobial resistance among nontyphoid Salmonella species has been increasing (3,4). Routine antimicrobial treatment of uncomplicated gastroenteritis caused by Salmonella species has been discouraged because it does not reduce the duration or severity of illness and may prolong asymptomatic carriage (5,6). Despite the increasing incidence of infections caused by antibiotic-resistant nontyphoid Salmonella, little is known about the impact of resistance on outcome. Furthermore, no published report is available indicating whether antibiotic treatment is mandatory for patients with antibiotic-resistant nontyphoid salmonellosis. Thus, this study was performed to investigate the epidemiologic, clinical, and microbiologic features of patients with nontyphoid salmonellosis in Taiwan and to elucidate the impact of antibiotic resistance on outcome. Moreover, we sought to identify the parameters to guide whether antibiotics are indicated in the era of antibiotic-resistant nontyphoid salmonellosis.
MATERIALS AND METHODS
Study Population and Case Definition
This is a prospective observational study. Between January 1996 and January 2002, all pediatric patients admitted to Veterans General Hospital, Kaohsiung, a tertiary referral center with 1,200 beds, whose stool (or rectal swab) or blood (or both) cultures were positive for nontyphoid Salmonella were enrolled in the study. A standard case report form was completed for each episode of nontyphoid salmonellosis. The following data were prospectively recorded: demographic, clinical, and micro-biologic features, underlying diseases, treatment regimen, complications, and response to treatment. The decision for antibiotic treatment was at the discretion of the attending physician with no input from the authors. The criteria for admission and discharge are customary. Patients were admitted when presenting with fever and any symptoms/signs of dehydration or bloody stool. Patients were discharged when afebrile for more than 24 hours and when symptoms/signs of dehydration had resolved.
All isolates were cultured and identified according to standard methods, with no major changes over time in the policy for the identification of Salmonella. All isolates were serotyped by the Wellcolex color Salmonella test (Murex, Dartford, United Kingdom), then confirmed by the slide agglutination test with the use of O antiserum to detect O antigen (Bacto, Liverpool, NSW, Australia). The antimicrobial susceptibility of Salmonella isolates was examined by the standard disc-diffusion method. Resistance to a specific antimicrobial was based on reference zone diameter interpretive standards (7). An organism was considered resistant if it was resistant to ≥ 1 of the antimicrobials tested and considered multiply resistant if resistant to ≥ 2 agents (8). The antimicrobial agents examined were ampicillin, chloramphenicol, gentamicin, trimethoprimsulfamethoxazole (TMP/SMZ), and ceftriaxone; ciprofloxacin was added after December 1998.
Diarrhea was defined as a decrease in consistency (i.e., soft or liquid) and an increase in frequency of bowel movements to ≥ 3 stools per day, as in other epidemiologic investigations (9). Bloody stool was defined as any stool reported by parent or guardian containing blood in one 24 hour period (10). Fever was defined as ≥ 37.5°C as measured by thermometer in the ear. No antibiotic therapy was defined when no antibiotics were used for at least 3 days after admission; however, other therapeutic modalities, such as oral or intravenous hydration, symptomatic therapy, and bowel decompression, usually were given. Outcome was measured by the response to treatment after admission defined as days of fever after admission and duration of hospital stay. Prolonged illness was defined as fever and/or length of hospital stay longer than the median for this cohort of patients.
A “severity score” was devised from factors associated with signs of systemic inflammation. The severity score is composed of three parameters:
- C-reactive protein (CRP) ≤ 3 mg/dL, 0 point; CRP > 3, ≤ 6 mg/dL, 1 point; and CRP > 6 mg/dL, 2 points.
- Band cells in peripheral blood ≤ 5%, 0 point; band cells > 5%, ≤ 10%, 1 point; band cells > 10%, 2 points.
- Days of fever before admission < 3 days, 0 point; ≥ 3 days, 1 point.
A score greater than 3 was considered severely ill. A score of 2 or 3 was moderately ill. A score of ≤ 1 was considered “mild.”
Concordant therapy was defined as receipt of an antibiotic(s) for the first 2 days after the stool or blood culture was obtained that was active against the organism in vitro. Discordant therapy was defined as receipt of an antibiotic(s) for the first 2 days after the stool or blood culture was obtained that was inactive in vitro against the nontyphoid Salmonella isolated.
Clinical and laboratory data were entered into a computer data base (Prophet Systems version 6.0, BBN/AbTech National Institutes of Health, Bethesda, Maryland, U.S.A.). Categorical data were analyzed using a χ2 or Fisher exact test. Continuous variables were compared by using the t test or the Mann-Whitney U test. A multiple linear model was used to evaluate the effects of multiple factors on duration of fever after admission and length of hospital stay.
Demographics and Clinical Features
Three hundred eleven consecutive patients were enrolled in the study. The characteristics and clinical manifestations of the patients are shown in Table 1. The median age was 15 months (1 month to 13 years old). Of the patients, 60.1% (187/311) were male. Bloody stool was noted in 47.6% (148/311) and occult blood in stool was positive in 75.6% (217/287). Among the patients, 67.5% (206/305) presented with mild to moderate dehydration, and 8.7% (27/311) had underlying diseases. Blood culture was positive in 47 patients. Eighteen patients had positive blood culture but negative stool culture for Salmonella. No patients experienced meningitis, septic arthritis, or osteomyelitis. To further elucidate the relationship of bacteremia and uncomplicated gastroenteritis, the patients were divided into three groups: patients with positive blood culture and negative stool culture; patients with positive blood and stool culture; and patients with negative blood culture and positive stool culture. The demographics and clinical features of these three groups are listed in Table 2. The mean age was significantly younger in the second group with positive blood and stool cultures (P = 0.04). The number of days of fever before admission was greater in the first group with positive blood culture and negative stool culture. Bloody stool, diarrhea, and dehydration were also encountered less frequently in this group (P = 0.005, 0.0001, and 0.004, respectively). No other significant difference was noted among the three groups of patients, including gender, white blood cell count, percent of polymorphonuclear neutrophils, platelet count, C reactive protein, HCO3−, and incidence of underlying diseases.
Salmonella enteritidis B caused 68.5% (213/311) of episodes, followed by S. enteritidis C1 (11.9%; 37/311), D (7.7%; 24/311), C2 (7.1%; 22/311), E (2.6%; 8/311), S. choleraesuis (1.6%; 5/311), and S. paratyphi (0.6%; 2/311). The percentage of Salmonella isolates in stool and blood are shown in Table 3. Salmonella enteritidis C1 and choleraesuis were significantly more likely to cause bacteremia (P = 0.01 and P = 0.002, respectively). Of the isolates, 60.5% (187/309) were resistant to ampicillin; 66.0% (196/297) were resistant to chloramphenicol; and 46.1% (142/308) were resistant to TMP/SMZ. One isolate was resistant to ceftriaxone, and three were resistant to ciprofloxacin (Table 4).
Therapy and Outcome
Twenty percent (62/311) of patients received no antibiotic therapy through the whole course of illness. Within the first 3 days, the variable therapeutic regimens were: 37.0% (115/311) of patients did not receive any antibiotics; 40.2% (125/311) received concordant antibiotic therapy; and 22.8% (71/311) received discordant therapy. The most common antibiotic regimen was the combination of ampicillin and gentamicin, which was used in 28.6% of cases. Ampicillin, ceftriaxone, and TMP/SMZ were received by 9.6%, 7.1%, and 4.2%, respectively. Complications were noted in 2% (6/311) of patients, including cecal perforation (one patient), toxic megacolon (one patient), intra-abdominal abscess (one patient), upper gastrointestinal bleeding (one patient), and acute appendicitis (two patients).
The two end points used to assess the impact of resistance were days of fever after admission and duration of hospital stay. Diarrhea was not used as a parameter in assessing outcome because it was too subjective, even though a clear definition of diarrhea was given and some patients did not have any diarrhea. Case fatality rate was not useful as an end point because no patient died. Patients with discordant therapy had more days of fever and longer hospital stay than did patients receiving concordant therapy, whereas patients receiving no antibiotic therapy had the fewest days of fever and shortest hospital stay (P = 0.058) (Fig. 1). Furthermore, when length of hospital stay greater than the median and duration of fever longer than 3 days were combined into a single outcome of prolonged illness, 62% (58/93) of patients without antibiotic treatment, 78% (87/111) of patients with concordant therapy, and 89% (55/62) of patients with discordant therapy had prolonged illness (P = 0.001).
When stratified by severity of illness, among the mild group, patients who did not receive antibiotics had a significantly lower rate of prolonged illness than did patients receiving concordant or discordant therapy (42%v 64%v 75%, P = 0.03). When compared with patients with no antibiotic therapy, patients with concordant therapy had a shorter duration of hospital stay only in the most severely ill group. Patients with discordant therapy had longer hospital stay, regardless of the severity level. As expected, patients with bacteremia had a more prolonged course than did those without. All patients with bacteremia receiving discordant therapy (9/9) had prolonged illness.
The incidence of multiply resistant Salmonella species has increased worldwide (11,12). A significant increase in the isolation of multiresistant strains from 0.6% in 1986 to 18.6% in 1991 has been reported in Spain (11). Multiple resistance in England and Wales has more than doubled, from 5% to 12%, in 1981 and 1988, respectively (12). In Taiwan, the incidence of multiple resistance (resistance to any two of the following antibiotics: ampicillin, chloramphenicol, tetracycline, gentamicin, and TMP/SMZ increased from 10.6% during the period 1989 to 1992 to 19.7% during the period 1993 to 1996 (4). In our study, the resistance rates to ampicillin, chloramphenicol, and TMP-SMX were 60.5%, 66.0%, and 46.1% respectively, which are higher than earlier reports (13). The high prevalence of resistance among nontyphoid Salmonella probably is a reflection of injudicious use of antibiotics in Taiwan, where antibiotics are available over the counter. Furthermore, indiscriminate use of antibiotics in animal feed might have promoted the selection and spread of antibiotic-resistant Salmonella because salmonellosis is a zoonosis (14).
This series of patients did not include children with Salmonella gastroenteritis who were diagnosed and treated as outpatients. By studying a hospitalized cohort in a tertiary referral center we might have underestimated the total incidence of Salmonella enterocolitis. Underlying diseases occur frequently in adult patients with salmonellosis; 91% of patients have been reported to have underlying conditions, such as malignancy, diabetes mellitus, liver cirrhosis, head or spinal cord injury, or uremia (15). Of note, sickle cell disease is significantly associated with the occurrence of salmonellosis. However, in our study, only 8.7% of patients had underlying diseases. None of the patients in this series had sickle cell disease, which is not prevalent in Taiwan (15).
The reported incidence of bacteremia among children with nontyphoid Salmonella gastroenteritis has varied from 2% to 47% (16–20), depending on the presence of underlying diseases, extremes of age, serotypes of the isolates, and geographic differences (18,21–25). Although the prevalence of diseases appears to be higher in males in many intestinal infectious diseases, there were no difference in the severity of presentation or the duration of illness in this cohort of pediatric patients with nontyphoid salmonellosis.
In this series of patients with salmonellosis, the clinical manifestations and laboratory data, including gender, fever days, WBC, percent PMN, and HCO3−, were not significantly different among the three groups of patients: those with positive blood culture and negative stool culture; those with positive blood and stool culture; and those with negative blood culture and positive stool culture. It is biologically plausible that the incidence of diarrhea, bloody stool, and dehydration were higher in the patients with positive stool culture. Whether the patients had occult bacteremia or false-negative stool culture could not be ascertained among patients with positive blood culture but negative stool culture. Blood culture is warranted among patients with nontyphoid Salmonella gastroenteritis because septicemia is impossible to identify on clinical grounds.
The finding in our study that patients receiving no antibiotic therapy had fewer days of fever and shorter hospital stay was in accordance with the belief that human salmonellosis usually is self-limited and antimicrobial therapy is seldom required (26). Moreover, earlier reports have indicated that antimicrobial therapy for uncomplicated gastroenteritis does not reduce the duration or severity of symptoms; in contrast, it may prolong fecal excretion at convalescence and result in the emergence of resistant strains (27). However, antibiotics are recommended in young infants (≤3 months) and other children who are at increased risk of a disseminated disease, such as children with immunocompromise; those with a severe or protracted course; and those with bacteremia or extraintestinal focal infections (28). The increasing prevalence of resistance of nontyphoid Salmonella further complicates the issue of antibiotic therapy. Patients with resistant nontyphoid Salmonella are hospitalized longer and have higher mortality than patients with susceptible organisms (8,29). Inappropriate antibiotic therapy has been reported to be significantly associated with mortality of salmonellosis (22,24). This study showed discordant therapy (treatment with antibiotics inactive in vitro) might prolong the days of fever and hospital stay. We suggest the severity score be used as the initial evaluation for patients with nontyphoid salmonellosis to guide whether antibiotics are mandatory; antibiotics should not be prescribed among immunocompetent patients with a score of 0 or 1. In vitro active antibiotics should be considered among patients whose blood culture is positive or when the severity score is more than 3. Patients with discordant therapy did worse at all levels of illness. Because the prevalence of resistance to ampicillin and TMP/SMX was high, treatment with cefotaxime, ceftriaxone, or a fluoroquinolone generally is preferred when antibiotics are indicated. In vitro susceptibility data should be obtained to guide therapy. Moreover, because this is an observational study, a randomized study is necessary to clarify which antimicrobial agent is most appropriate for patients with resistant nontyphoid salmonellosis.
The problem of antimicrobial resistance in organisms causing diarrheal diseases will continue in developed and developing countries given the data reported (30) and the increased use of antibiotics. In treating antibiotic-resistant nontyphoid salmonellosis, antibiotics active in vitro should be administered when antibiotics are indicated, whereas antibiotics are still not mandatory for the subset of patients who present with primarily gastrointestinal symptoms who have limited signs of systemic inflammation.
The authors thank Dr. Cheng-Hsun Chiu for his expertise and insight in reviewing the manuscript and for his helpful comments.
1. Fasano A. Intestinal infections. In: Walker WA, Durie PR, Hamilton JR, et al., eds. Pediatric Gastrointestinal Disease.
3rd ed. BC Decker, Hamilton, Ontario, Canada: 2000:466–84.
2. Blaser MJ, Feldman RA. From the Centers for Disease Control. Salmonella
bacteremia: reports to the Centers for Disease Control, 1968–1979. J Infect Dis
3. Peng CF. Incidence and antimicrobial resistance
serotypes in southern Taiwan from 1978 through 1987. Kaohsiung J Med Sci
4. Yang YJ, Liu CC, Wang SM, et al. High rates of antimicrobial resistance
among clinical isolates of non-typhoidal Salmonella
in Taiwan. Eur J Clin Microbiol Infect Dis
5. Jewes LA. Antimicrobial therapy of non-typhi Salmonella
infection. J Antimicrob Chemother
6. Neill MA, Opal SM, Heelan J, et al. Failure of ciprofloxacin to eradicate convalescent fecal excretion after acute salmonellosis: experience during an outbreak in health care workers. Ann Intern Med
7. Barry AL, Thornsberry C. Susceptibility tests: diffusion test procedures. In: Balows A, Hausler WJ Jr, Herrmann KL, et al., eds. Manual of Clinical Microbiology.
5th ed. Washington, DC: American Society for Microbiology; 1991:1117–25.
8. Lee LA, Puhr ND, Maloney EK, et al. Increase in antimicrobial-resistant Salmonella
infections in the United States, 1989–1990. J Infect Dis
9. Guerrant RL, Van Gilder T, Steiner TS, et al. Infectious Diseases Society of America. Practice guidelines for the management of infectious diarrhea. Clin Infect Dis
10. Townes JM, Quick R, Gonzales OY, et al. Etiology of bloody diarrhea in Bolivian children: implications for empiric therapy. Bolivian Dysentery Study Group. J Infect Dis
11. Reina J, Gomez J, Serra A, et al. Analysis of the antibiotic resistance
detected in 2043 strains of Salmonella enterica
subsp. Enterica isolated in stool cultures of Spanish patients with acute diarrhea (1986–1991). J Antimicrob Chemother
12. Ward LR, Threlfall EJ, Rowe B. Multiple drug resistance
in England and Wales: a comparison between 1981 and 1988. J Clin Pathol
13. Su LH, Chiu CH, Kuo AJ, et al. Secular trends in incidence and antimicrobial resistance
among clinical isolates of Salmonella
at a university hospital in Taiwan, 1983–1999. Epidemiol Infect
14. Lee CY, Chiu CH, Chuang YY, et al. Multidrug-resistant non-typhoid Salmonella
infections in a medical center. J Microbiol Immunol Infect
15. Chiu CH, Lin TY, Ou JT. Predictors for extraintestinal infection of non-typhoidal Salmonella
in patients without AIDS. Int J Clin Pract
16. Meadow WL, Schneider H, Beem MO. Salmonella enteritidis
bacteremia in childhood. J Infect Dis
17. Torrey S, Fleisher G, Jaffe D. Incidence of Salmonella
bacteremia in infants with Salmonella
gastroenteritis. J Pediatr
18. Yamamoto LG, Ashton MJ. Salmonella
infections in infants in Hawaii. Pediatr Infect Dis J
19. Raucher HS, Eichenfield AH, Hodes HL. Treatment of Salmonella
gastroenteritis in infants: the significance of bacteremia. Clin Pediatr
20. Hyams JS, Durbin WA, Grand RJ, et al. Salmonella
bacteremia in the first year of life. J Pediatr
21. Chiu CH, Lin TY, Ou JT. Prevalence of the virulence plasmids of nontyphoid Salmonella
in the serovars isolated from humans and their association with bacteremia. Microbiol Immunol
22. Lee SC, Yang PH, Shieh WB, et al. Bacteremia due to non-typhi Salmonella
: analysis of 64 cases and review. Clin Infect Dis
23. Galofre J, Moreno A, Mensa J, et al. Analysis of factors influencing the outcome and development of septic metastasis or relapse in Salmonella
bacteremia. Clin Infect Dis
24. Sirinavin S, Jayanetra P, Lolekha S, et al. Predictors for extraintestinal infection in Salmonella
enteritis in Thailand. Pediatr Infect Dis J
25. Sirinavin S, Jayanetra P, Thakkinstian A. Clinical and prognostic categorization of extraintestinal nontyphoidal Salmonella
infections in infants and children. Clin Infect Dis
26. Chiu CH, Lin TY, Ou JT. A clinical trial comparing oral azithromycin, cefixime and no antibiotics in the treatment of acute uncomplicated Salmonella
enteritis in children. J Paediatr Child Health
27. Aserkoff B, Bennett JV. Effect of antibiotic therapy in acute salmonellosis on the fecal excretion of Salmonellae. N Engl J Med
28. Cleary TG. Salmonella
. In: Behrman RE, Kliegman RM, Jenson HB, eds. Nelson Textbook of Pediatrics
. 16th ed. W.B. Saunders Company, Philadelphia: 2000:842–8.
29. Holmberg SD, Wells JG, Cohen ML. Animal to man transmission of antimicrobial resistant Salmonella
: investigations of U.S. outbreak, 1971–1983. Science
30. Murray BE. Can antibiotic resistance
be controlled? (editorial). N Engl J Med
Keywords:© 2004 Lippincott Williams & Wilkins, Inc.
Nontyphoid salmonellosis; Pediatric patients; Resistance