Salmonella/Rotavirus Coinfection in Hospitalized Children

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector Kaohsiung Journal of Medical Sciences (2012) 28, 595e600

Available online at www.sciencedirect.com

journal homepage: http://www.kjms-online.com

ORIGINAL ARTICLE Salmonella/rotavirus coinfection in hospitalized children

Wei-Te Lee a, Pei-Chen Lin b,c, Lung-Chang Lin d,e, Hsiu-Lin Chen d,f, Rei-Cheng Yang d,*

a Department of Pediatrics, Kaohsiung Municipal Hsiao-Kang Hospital, Kaohsiung Medical University, Kaohsiung, Taiwan b Research Education & Training, Kaohsiung Municipal Hsiao-Kang Hospital, Kaohsiung Medical University, Kaohsiung, Taiwan c Graduate Institute of Occupational Safety and Health, Kaohsiung Medical University, Kaohsiung, Taiwan d Department of Pediatrics, Kaohsiung Medical University Hospital, Kaohsiung Medical University, Kaohsiung, Taiwan e Graduate Institute of Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan f Faculty of Respiratory Therapy, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan

Received 27 July 2011; accepted 21 October 2011 Available online 21 September 2012

KEYWORDS Abstract This study’s aim was to analyze the characteristics and severity of acute gastroen- Clinical severity; teritis related to infection with rotavirus (group R), nontyphoid Salmonella (NTS; group S), and Mean temperature infection with both rotavirus and Salmonella (group B) in children in southern Taiwan in order difference; to improve diagnosis and expedite appropriate management. The medical records of children Coinfection; admitted between October 2002 to September 2008 for acute gastroenteritis related to rota- Rotavirus; virus, NTS, or coinfection were collected and analyzed.Among 2040 reviewed medical records, Salmonella 40 patients were infected with both pathogens, while 501 cases were infected with rotavirus alone and 189 were infected with NTS alone. There were no signiﬁcant differences between the three groups in terms of age at admission or sex. The age distribution of the reviewed cases revealed that children between the ages of 12e24 months comprised the largest proportion of cases in each group. Higher concentrations of fecal leukocytes and fecal pus cells and longer hospitalizations were observed in group B in comparison with groups S and R (p < 0.05). Clinical severity was signiﬁcantly higher in groups B and S than group R (p < 0.05). A proportional association was found between the monthly case number of rotavirus infections and the mean monthly temperature difference in southern Taiwan (r Z 0.9248; p < 0.0001). In summary, concomitant rotavirus infection with NTS infection did not affect the clinical manifestations of the reviewed patients. Rotavirus infection was less severe in most clinical manifestations, but vomiting was more severe in rotavirus-infected patients. Positive fecal leukocytes and positive fecal pus cells were more frequent during coinfection. There was a strong positive

* Corresponding author. Department of Pediatrics, Kaohsiung Medical University Hospital, Kaohsiung Medical University, #100, Tzu-you First Road, Kaohsiung City, Taiwan. E-mail address: [email protected] (R.-C. Yang).

Introduction group was plotted against the mean monthly temperature difference. Diarrhea is one of the major causes of morbidity and Clinical symptoms (e.g., fever, vomiting, diarrhea), mortality in children[1], demonstrating a median incidence laboratory data (white blood cell [WBC] count, C-reactive of 3.2 episodes per child-year in children < 5 years old [2]. protein [CRP] level, stool culture for NTS, and the presence It has been estimated that 25% of the growth differential of stool occult blood, fecal leukocytes, fecal pus cells, and between children in developing countries and North rotavirus antigen in stool), and length of hospitalization America is related to diarrhea [3]. were obtained from each patient’s medical chart and S Viruses are the most important pathogens that cause reviewed. Fever was defined as 38 C as measured by an diarrhea in children and are responsible for approximately ear thermometer (FirstTemp Genius, Model 3000A; Sher- 70e80% of all cases [4]. Rotavirus is the leading cause of wood, IMS, Ins., Carlsbad, California, USA). A diarrhea severe diarrhea in infants and young children worldwide episode was defined as the passage of unformed stool for at and is responsible for high morbidity in developed countries least twice the usual daily frequency during a 24-hour and high mortality in developing countries, with an esti- period. Severity was assessed using a modified version of mated annual death toll of 440,000e500,000 worldwide the method described by Flores et al., which assessed the [5,6]. Bacteria are the second most common cause of severity of diarrhea, vomiting, and associated fever [11]. diarrhea. Nontyphoid Salmonella (NTS) is the leading cause The duration of fever (in days) was included in the of bacterial enterocolitis in Taiwan [7,8]. Most cases of NTS- numerical score system because this is used to characterize induced acute diarrhea are generally self-limiting, however the severity of gastroenteritis in clinical practice (Table 1). bacteremia and focal complications, such as meningitis and The degree of severity was expressed as the sum of the osteomyelitis, may occur and result in high morbidity and points attributed to each symptom. mortality without appropriate treatment. Stool and blood culturing were performed on selective As for pneumonia, viral pneumonia could be complicated media plates, including a Salmonella Shigellamedium and by bacterial superinfection (e.g., Streptococcus pneumo- a Hektoen enteric medium. Salmonella serotyping was niae) via the damaged epithelial layer. Whether the same determined using a commercial antiserum assay (Difco conditions could occur in the gastrointestinal tract is Laboratories, Detroit, Michigan, USA). The rotavirus antigen unknown. There are only a few reports in the literature that was detected using a rapid immunochromatographic discuss the characteristics of rotavirus/NTS-caused gastro- assay (Rota-check-1; VEDA Laboratories, Alençon, Basse, enteritis in children [9,10]. In addition, we were unable to France). The monthly temperature data in Kaohsiung were find any reports comparing the severity of disease in provided by the Central Weather Bureau, Taiwan. The hospitalized children infected with rotavirus, NTS, or those mean temperature difference was defined as the mean high who were coinfected. This study analyzed the character- temperature minus the mean low temperature for each istics and severity of hospitalized children infected with month. rotavirus, NTS, and both rotavirus and NTS in southern Statistical analysis included the Chi-square test or Fisher Taiwan in order to improve diagnosis and expedite appro- exact test for categorical data, ANOVA for continuous data, priate management. and Spearman correlation for determining the relationship between continuous variables. All data were analyzed using Statistical Analysis Software (SAS; JMP version 6.0; SAS Patients and methods Institute Inc., Cary, NC, USA). A p value < 0.05 was considered statistically significant. The medical records of children who were admitted to Kaohsiung Municipal Hsiao-Kang Hospital, a community- based teaching hospital with 464 beds in southern Taiwan, Results from October 2002 to September 2008 with acute gastroenteritis were recruited. Approval from the institutional Demographic features review board of Kaohsiung Municipal Hsiao-Kang Hospital was obtained before beginning this study. A total of 2040 cases were reviewed, and 730 of these were Based on the results of the stool examinations, the infected with rotavirus, NTS, or both. Among these 730 patients were divided into group B (positive for both rota- cases, 40 (5.5%) patients were infected with both rotavirus virus and NTS), group R (positive for rotavirus and negative and NTS, 501 (68.6%) were infected with rotavirus alone, for NTS), or group S (negative for rotavirus and positive for and 189 (25.9%) were infected with NTS alone. The average NTS). Demographic chacteristics, including sex, age, and age on admission was 2.41 1.70 years for group B, the duration of hospitalization, were acquired from their 2.73 2.20 years for group R, and 2.53 2.81 years for medical records and initially assessed on a monthly basis for group S. A significant difference was not observed between the 6-year study period. The monthly incidence of each the age groups (Table 2). The 12e24-month age group Salmonella/rotavirus coinfection in children 597

Table 1 Comparison of the clinical features of nontyphoid Salmonella, rotavirus, and coinfection. Severity score Salmonella Rotavirus Coinfection, n(%) p infection, n (%) infection, n(%) Duration of diarrhea (d) <2 1 11 (5.82) 122 (24.55)* 1 (2.50) <0.0001 2e4 2 62 (32.80) 257 (51.71) 12 (30.00) >4 3 116 (61.38) 118 (23.74) 27 (67.50) Maximum number of incidents of diarrhea per day (no.) <4 1 8 ( 4.35) 68 (15.42)* 0 (0.00) <0.0001 4e5 2 22 (11.96) 107 (24.26) 2 (5.00) >5 3 154 (83.70) 266 (60.32) 38 (95.00) Duration of vomiting (d) 0 0 112 (59.26) 91 (18.24)* 26 (65.00) <0.0001 1e2 1 69 (36.51) 345 (69.14) 14 (35.00) >2 3 8 (4.23) 63 (12.63) 0 (0.00) Fever (maximum C) None 0 22 (11.64) 147 (29.46) 6 (15.38)** <0.0001 S38 1 81 (42.86) 246 (49.30) 22 (56.41) <39, S39 2 86 (45.50) 106 (21.24) 11 (28.21) Duration of fever (days) <2 0 21 (11.11) 116 (23.25) 8 (20.00)** <0.0001 2e3 1 56 (29.63) 227 (45.49) 13 (32.50) >3 3 112 (59.26) 156 (31.26) 19 (47.50) Clinical severity score, mean SD 9.23 2.07 7.91 2.28* 8.85 1.95 <0.0001 no. 184 439 39

*Signiﬁcantly different from the other two groups (p < 0.05). **Not signiﬁcantly different from the other two groups (p > 0.05).

Table 2 Comparison of the demographic and clinical features and laboratory data of patients infected with nontyphoid Salmonella, rotavirus, or coinfection. Salmonella infection Rotavirus infection Coinfection p Age (y), mean SD 2.53 2.81 2.73 2.20 2.41 1.70 0.4717 No. 189 501 40 Gender (male/female) 114/75 303/198 25/15 0.9661 Clinical features Vomiting (%) 77 (40.74) 408 (81.76)* 14 (35.00) <0.0001 Fever (%) 167 (88.36) 352 (70.54) 33 (84.62)** <0.0001 Diarrhea (%) 186 (98.41) 461 (92.02) 40 (100)** 0.0017 Laboratory data WBCy (cells/mm3), mean SD 11.10 3.71 11.37 5.58 10.93 3.21 0.7463 no. 189 499 40 CRPz (mg/L), mean SD 71.16 61.12 25.71 39.55* 63.63 47.21 <0.0001 no. 189 498 40 CRPz (>50 mg/L) (%) 105/189 (55.7) 84/498* (16.9) 20/40 (50.0) 0.0002 Positive for stool occult blood (%) 122/185 (65.95) 197/497* (41.13) 32/40 (80.00) <0.0001 Positive for fecal leukocytes*** (%) 41/136 (30.15) 35/387 (9.04) 18/38 (47.37) <0.0001 Positive for fecal pus cells (%) 6/166 (3.61) 7/387 (1.81) 7/39* (17.95) <0.0001 Length of hospital stay*** (d), mean SD 6.99 3.14 5.44 3.31 9.05 3.35 <0.0001

*Significantly different from the other two groups (p < 0.05). **Not significantly different from the other two groups (p > 0.05). ***Significant difference among all three groups (p < 0.05). yWBC Z white blood cell; zCRP Z C-reactive protein. 598 W.-T. Lee et al. comprised the majority of the cases in each group (Fig. 1). S. Using a cutoff value of 50 mg/L, the percentages of Coinfection was not observed in children > 7 years old. The patients with an elevated CRP level were significantly gender ratios (a total of 442 males and 288 females) did not greater in groups B (50%, p < 0.0001 ) and S (55.7%, differ significantly between the three groups (p Z 0.9661). p < 0.0001) than in group R (16.9%); however, there was no significant difference between groups B and S. The percentage of patients that were positive for stool Clinical features, severity, and laboratory data occult blood in group R (41.13%) was significantly lower than in groups B (80%, p < 0.0001) and S (65.95%, Incidences of vomiting, fever, and diarrhea p < 0.0001), but there was no significant difference In group R, the incidence of vomiting (81.76%) was higher between groups B and S. There were significant differences < than those of group B (35%; p 0.0001) and group S in terms of the presence of fecal leukocytes among groups < (40.74%; p 0.0001). A significant difference was not (group B: 47.37%; group S: 30.15%; group R: 9.04%; observed between groups B and S (Table 2). The percent- p<0.0001). The percentage of positive fecal pus cells in ages of fever and diarrhea in group R (70.54% and 92.02%, group B (17.95%) was significantly greater than those respectively) were significantly lower in comparison with measured in groups R (1.81%, p < 0.0001) and S (3.61%, < Z group S (88.36% and 98.41%; p 0.0001 and p 0.0019, p Z 0.0039), with no significant difference noted between respectively). Although the incidences of fever and diar- groups S and R. rhea in group B (84.62% and 100%, respectively) were greater than those in group R (70.54% and 92.02%); Clinical severity score and length of hospital stay however, these differences were not significant. In addition, there were no significant difference in terms of fever The clinical severity scores of groups B (8.85 1.95, or diarrhea between groups B (84.62% and 100%, respec- p Z 0.0129) and S (9.23 2.07, p < 0.0001) were signifi- tively) and S (88.36% and 98.41%). cantly higher than that of group R (7.91 2.28). No significant difference was observed between groups B and S Severity of diarrhea, vomiting, and fever (Table 1). Meanwhile, patients in group B required signifi- As summarized in Table 1, severity of diarrhea was more cantly longer hospital stays (9.05 3.35 days) than those in pronounced in groups B and S than group R. More patients in group S (6.99 3.14 days) or R (5.44 3.31 days, p < 0.001; groups B and S experienced diarrhea lasting for > 4 days Table 2). Furthermore, a good correlation was observed and > 5 times per day (p < 0.0001). However, no significant between the clinical severity score and the length of the difference between groups B and S was found. The duration required hospital stay(r Z 0.279; p < 0.0001). of vomiting was significantly longer in group R (p < 0.0001), but there was no significant difference between groups B and S. The severity of fever in group S, including the Relationship between monthly incidence of disease maximum body temperature and duration of fever, was and temperature significantly more severe than that of group R (p < 0.0001). There was no significant difference in terms of the severity The monthly incidence of disease was plotted by of fever between group B and the other two groups. percentage against the mean monthly temperature difference in the Kaohsiung area (Fig. 2). Rotavirus infection was Laboratory data more frequent in the fall-winter season. The mean monthly The WBC count did not differ between the groups (group S: temperature difference was more strongly associated with 11,100 3710 cells/mm3; group R: 11,370 5580 cells/ the incidence of rotavirus infection in the Kaohsiung area mm3; group B: 10,930 3210 cells/mm3; Table 2). The CRP (r Z 0.9248; p < 0.0001) in comparison with the mean values of groups B (63.63 47.21 mg/L, p <0.0001) and S monthly temperature (r Z -0.8408; p Z 0.0006). NTS was (71.16 61.12 mg/L, p < 0.0001) were significantly higher more common in the summer season. Coinfection than that measured in group R (25.71 39.55 mg/L). No frequently occurred during the peak seasons for rotavirus statistical differences were observed between groups B and and NTS.

30 20 12 ) (

(%) 10 25 (%) 15

cases 8

20 cases 10 6

15 erature difference 4 p

5 tem

10 Percentage of Percentage of 2 Mean 5 0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 Time (month) 0 ~6 6 ~ 12 12 ~ 24 24 ~ 36 36 ~ 60 60 ~ 84 >=84 Age (month) Salmonella Rotavirus Coinfection Mean Temperature Difference Rotavirus Salmonella Coinfection Figure 2. Relationship between monthly incidence of non- Figure 1. Age distribution of patients infected with nontyphoid Salmonella, rotavirus, and coinfection and the mean typhoid Salmonella, rotavirus, and those with coinfection. monthly temperature difference in Kaohsiung. Salmonella/rotavirus coinfection in children 599

Discussion were found more frequently in coinfected patients. There was also a strong positive relationship between the inci- Acute gastroenteritis remains as an important cause of dence of rotavirus gastroenteritis and the mean monthly hospitalization for Taiwanese children. During the 6-year temperature difference. period that was studied, acute gastroenteritis was the second leading cause of hospitalization in children in our hospital. Rotavirus and NTS were the two most common Acknowledgments pathogens found in this study, which is consistent with previous reports [7,8,12]. The incidence of coinfection in The authors would like to thank Dr. John Ebinger and Matt this study was 5.5%, which is within the ranges reported by Nicodemus for editing the English version of this study. previously published studies (0.48e7.4%) [13e15]. It is not surprising that the prevalence of coinfection was higher among males in this study because the prevalences of References rotavirus and NTS infection have been reported as higher in male children [12,16]. However, there were no differences [1] Guerrant RL, Hughes JM, Lima NL, Crane J. Diarrhea in between males and females with coinfection or NTS developed and developing countries: magnitude, special settings, and etiologies. Rev Infect Dis 1990;12(Suppl. 1): infection in terms of clinical severity score or length of S41e50. hospitalization. [2] Kosek M, Bern C, Guerrant RL. The global burden of diarrhoeal Regarding clinical manifestations, NTS and coinfection disease, as estimated from studies published between 1992 were significantly related to a higher frequency of stool and 2000. Bull World Health Organ 2003;81:197e204. passage, longer duration of diarrhea, longer length of [3] Gastanaduy AS, Begue RE. Acute gastroenteritis. Clin Pediatr hospitalization, and higher clinical severity score in (Phila) 1999;38:1e12. comparison with rotavirus infection. However, the severity [4] Sung YL, Huang YF, Huang CF, Chen KT, Chen HY, Lee GH, of illness was similar between coinfection and NTS infection et al. Emergence of G9 serotype rotavirus as a major cause of because NTS infection may mask rotavirus infection, at infectious gastroenteritis in southern Taiwan. J Microbiol e least in terms of clinical severity. Hung et al. report that Immunol Infect 2004;37:322 6. [5] Parashar UD, Hummelman EG, Bresee JS, Miller MA, Glass RI. 70% of patients that were coinfected with NTS and rotavirus Global illness and deaths caused by rotavirus disease in chil- presented with vomiting in comparison with 40% of patients dren. Emerg Infect Dis 2003;9:565e72. with NTS infection only [10]. This phenomenon did not [6] Bass DM. Rotaviruses, caliciviruses, and astroviruses. In: present in our study, indicating that vomiting may not be an Kliegman RM, Behrman RE, Jenson HB, Stanton BF, editors. indicator for differentiating coinfection from NTS infection. Nelson textbook of pediatrics. 18th ed. Philadelphia: Saun- A previous study demonstrated that the WBC count ders; 2007. p. 1399e401. provides only a minimal contribution to the diagnosis of [7] Peng CF. Incidence and antimicrobial resistance of Salmonella bacterial enteritis [17]. However, an elevated CRP serotypes in southern Taiwan from 1978 through 1987. Gaox- level > 50 mg/L and positive stool occult blood test were iong Yi Xue Ke Xue Za Zhi 1992;8:247e54. valuable indicators of bacterial enteritis in children in our [8] Chen SM, Ni YH, Chen HL, Chang MH. Microbial etiology of acute gastroenteritis in hospitalized children in Taiwan. study. Furthermore, positivity for fecal leukocytes and J Formos Med Assoc 2006;105:964e70. fecal pus cells were found more frequently in coinfected [9] Chi H, Sun W, Chan WT, Lee HC, Fang SB. Pediatric Salmonella patients in our study. enterocolitis in a teaching hospital in Taitung: a four-year Rotavirus infections commonly present during the winter analysis. Acta Paediatr Taiwan 2001;42:297e300. months in temperate climates [6]. However, in most trop- [10] Hung TY, Liu MC, Hsu CF, Lin YC. Rotavirus infection increases ical areas, rotavirus causes enteritis throughout the year the risk of bacteremia in children with nontyphoid Salmonella without seasonal variation [18]. Taiwan is located in gastroenteritis. Eur J Clin Microbiol Infect Dis 2009;28:425e8. a subtropical zone at a latitude of 21e25 north of the [11] Flores J, Perez-Schael I, Gonzalez M, Garcia D, Perez M, equator. Therefore, rotavirus infection occurs throughout Daoud N, et al. Protection against severe rotavirus diarrhoea the year in Taiwan, with an epidemic peak during the cooler by rhesus rotavirus vaccine in Venezuelan infants. Lancet 1987;1:882e4. months[8,12,19,20]. The present study, which was con- [12] Chiu TF, Lee CN, Lee PI, Kao CL, Lin HC, Lu CY, et al. Rotavirus ducted in the Kaohsiung area (south of the Tropic of gastroenteritis in children: 5-year experience in a medical Cancer), reveals that the peak incidence of rotavirus center. J Microbiol Immunol Infect 2000;33:181e6. infection is between January and March. Interestingly, this [13] Barnes GL, Uren E, Stevens KB, Bishop RF. Etiology of acute is the first study to report that the mean temperature gastroenteritis in hospitalized children in Melbourne, difference is more strongly related to the incidence of Australia, from April 1980 to March 1993. J Clin Microbiol rotavirus gastroenteritis than the mean temperature. 1998;36:133e8. Further prospective longitudinal surveillance may provide [14] Ling JM, Cheng AF. Infectious diarrhoea in Hong Kong. J Trop additional reliable epidemiological data that could provide Med Hyg 1993;96:107e12. better and faster diagnoses. [15] Khuffash FA, Sethi SK, Shaltout AA. Acute gastroenteritis: clinical features according to etiologic agents. Clin Pediatr In conclusion, this retrospective study confirms that the (Phila) 1988;27:365e8. clinical severity of NTS infection is worse than that of [16] Huang IF, Wagener MM, Hsieh KS, Liu YC, Wu TC, Lee WY, et al. rotavirus infection. The clinical manifestations of rotavirus Nontyphoid salmonellosis in Taiwan children: clinical mani- and NTS coinfection are not worse than those of NTS festations, outcome and antibiotic resistance. J Pediatr Gas- infection. Positivity for fecal leukocytes and fecal pus cells troenterol Nutr 2004;38:518e23. 600 W.-T. Lee et al.

[17] Meloni GF, Tomasi PA, Spanu P, Piga S, Porcu A. C-reactive [19] Lo YS, Huang BL, Tsai LY, Lu CC, Chen TS. Clinical observations protein levels for diagnosis of Salmonella gastroenteritis. of infants and children with rotavirus gastroenteritis in Pediatr Infect Dis J 1999;18:471e3. southern Taiwan. Gaoxiong Yi Xue Ke Xue Za Zhi 1988;4: [18] Rodriguez WJ, Kim HW, Arrobio JO, Brandt CD, Chanock RM, 358e63. Kapikian AZ, et al. Clinical features of acute gastroenteritis [20] Tsai CH, Chiu HH, Abe T. Epidemiologic features of rota- associated with human reovirus-like agent in infants and virus infection in Taiwan: a review. Pediatr Int 2000;42: young children. J Pediatr 1977;91:188e93. 411e4.