Interpersonal relationships and group A streptococcus spread in a Mexican day-care center

Relaciones interpersonales y diseminación del estreptococo del grupo A (EGA) en una guardería de México

Alberto Villaseñor-Sierra, D en C^I; Mayra G. Quiñonez-Alvarado, M en C^I; José Ramiro Caballero-Hoyos, D en Sociol^II

^ICentro de Investigación Biomédica de Occidente, Instituto Mexicano del Seguro Social (IMSS), Guadalajara, Jalisco, México
^IIUIESSA. IMSS. Tonalá, Jalisco, México

ABSTRACT

OBJECTIVE: To study the effect of different degrees of centrality on the carrying of identical group A streptococcus (GAS) clones in the nasopharynx of children from a Mexican public day-care center.
MATERIAL AND METHODS: Nasopharyngeal cultures were performed in children from rooms B (RB) (n = 35) and C (RC) (n = 37). The Restriction Fragment Length Polymorphism (RFLP) patterns were compared among GAS isolates. A social networks questionnaire was filled out for each child and 10 classmates. Structure coefficients were compared among children with and without GAS.
RESULTS: Four GAS clones were identified; clone I in five children from RC; clone II in two from RC and one from RB; clone III in one from RB and one from RC; and clone IV in one from RC. Social network structure: Density of RB and RC = 0.40 (± 0.87) and 0.35 (± 0.80), respectively. In RB, the homophily pattern of interaction was different in carriers (0.00), non-carriers (0.47) and both (0.47) p = 0.35. In RC, the homophily pattern was also different in carriers (0.46), non-carriers (0.68) and mixed (0.19), p = .001. In 4/5 with clone I, the values of degree, closeness and betweenness were above the group mean. In 3/3 with clone II, the values of degree and betweenness were also above the mean. In contrast, in those with clone III and IV, the values of degree, closeness and betweenness were below the group mean.
CONCLUSION: The spread of specific GAS clones was associated with groups of children having a high proportion of ties and a high centrality level. This is evidence that spread of GAS strains among children attending day-care centers is not random but dependent on the degree of communication and physical contact between pairs.

Keywords: Streptococcus pyogenes; networks; child day-care centers; Mexico

RESUMEN

OBJETIVO: Evaluar el efecto de grados diferentes de centralidad con la presencia de clonas idénticas de estreptococo del grupo A (EGA) en la nasofaringe de niños de una guardería pública de México. ]]> MATERIAL Y MÉTODOS: Se realizaron cultivos nasofaríngeos en niños de los salones B (SB) (n = 35) y C (SC) (n = 37). El patrón de polimorfismos de longitud de fragmento por restricción (Restriction Fragment Length Polymorphism, RFLP) fue comparado entre aislamientos de EGA. Un cuestionario de redes sociales fue llenado para cada niño y 10 compañeros. Los coeficientes de estructura fueron comparados entre niños con y sin EGA.
RESULTADOS: Se identificaron cuatro clonas de EGA. Clona I en cinco niños del SC; clona II en dos del SC y en uno del SB; clona III en uno del SB y uno del SC; y clona IV en uno del SC. Estructura de redes sociales: Densidad SB y SC = 0.40 (± 0.87) y 0.35 (± 0.80), respectivamente. En SB, el patrón de homofilia de la interacción fue distinto en portadores (0.00), no portadores (0.47) y ambos (0.47) p = 0.35. En SC, el patrón de homofilia fue distinto en portadores (0.46), no portadores (0.68) y mixto (0.19), p = .001. En 4/5 con la clona I, los valores de grado, cercanía e intermediación estuvieron por arriba de la media grupal. En 3/3 con la clona II, los valores de grado e intermediación estuvieron por arriba de la media grupal. En contraste, en los niños con clonas III y IV, los valores de grado, cercanía e intermediación estuvieron por debajo de la media grupal.
CONCLUSIONES: La diseminación de clonas específicas de EGA se asoció a grupos de niños con gran proporción de lazos entre ellos y un alto nivel de centralidad. Esto evidencia que la transmisión de EGA entre niños de guardería no ocurre al azar sino que depende del grado de comunicación y contacto físico entre éstos.

Palabras clave: Streptococcus pyogenes; relaciones interpersonales; guarderías; México

Group A Streptococcus (GAS) is an important cause of mild, moderate and severe human infections worldwide.^1,2 In day-care centers, GAS is responsible for nasopharyngeal, perianal and respiratory infections in children, along with impetigo, outbreaks of scarlet fever, and invasive infections associated with varicella.^3-7

Several studies have determined the prevalence of GAS carriers and the spread of the bacteria during outbreaks among children attending day-care centers.^4,8-10 Healthy GAS carriers play an important role as potential sources of spread of GAS strains (with diverse degrees of virulence) between individuals.¹¹ Results from these studies showed that some risk factors related to the spread of virulent GAS strains are: a) sharing the index case’s room, and b) the time in hours (more than 24 hours) that the " contacts" spend with the index case during a seven day period. A deeper approach for studying the relationships and interactions between groups of people, and how an action of one may affect others, is called the Social Network Analysis (SNA).¹² This approach makes it possible to identify the " most important or central" actors in a group.¹³ SNA has been used to study the spread of diseases^14,15 and thus can be applied to help understand the effect of different degrees of communication and physical interaction patterns on the spread of a particular GAS strain among children. Common SNA measurements distinguish a specific actor based in their centrality in the network. Centrality measures include degree, closeness, and betweenness, which quantify the prominence of an individual actor embedded in a network.^13,16

The aim of the study was to determine the relationship between sharing the same GAS clones and the interpersonal communication structure in children from a day-care center in Mexico.

A microbiologic survey was conducted from October 31st to November 7th, 2001, to detect nasopharyngeal carriers of GAS among children 3 to 4 years old, from two rooms (room B; n = 35 and room C; n = 37) at a day-care center of the Instituto Mexicano del Seguro Social (Mexican Institute of Social Security). Children attending the day-care center were from a middle socioeconomic stratum of Guadalajara, Mexico. Signed parental consent and approval of the Local Research and Ethics Committee (1305) of the Instituto Mexicano del Seguro Social were obtained prior to the survey, Samples were taken from the nasopharynx using a sterile cotton swab, transported to the microbiology laboratory in Stuart’s medium, and cultured in 5% sheep blood agar within two hours of obtaining the specimen. Agar plates were incubated at 37 °C and 5% CO₂ for 18 hours. Beta hemolytic colonies were isolated and plated on a 5% sheep blood agar to test the sensitivity to bacitracin (TAXO A, BBL, Becton Dickinson. Sparks, MD, USA). Sensitive strains were subjected to latex agglutination (Strepto-Kit, bioMerieux. Marey, France) to confirm their identity as group A streptococci.

Because a history of antimicrobial consumption could affect the rate of GAS isolation in the nasopharynx cultures, parents of each child were asked about antibiotic use during the last two months. The type of antibiotic and the time elapsed since the last dose were recorded.

The polymorphisms of DNA from GAS isolates were analyzed by comparing the RFLP patterns of the VIR regulon by the method described by Gardiner et al¹⁷ and modified by Hartas et al.¹⁸ Briefly, one colony of a GAS strain was cultured overnight (ON) in Todd Hewitt Broth supplemented with yeast extract and glycine (THBYG), at 37 °C and 5% CO2. After a purity check was performed, a 10% volume of the ON culture was sub-cultured in fresh warm THBYG and incubated under the same conditions for two hours to obtain a logarithmic phase culture. The culture was then centrifuged and the bacterial pellet washed twice with Tris-HCl pH 8.2. Bacterial lysis was performed by adding polyethylene-glycol (PEG) and lysozyme, and incubating the mix at 37 °C in an ON water bath. The sample was then centrifuged, the pellet re-suspended in Tris-HCl, EDTA, Proteinase K and SDS, and incubated in water bath at 37 °C. After ON incubation, NaCl was added and the sample centrifuged. The supernatant was discarded and the resulting DNA pellet was washed twice with ethyl alcohol and air dried overnight. The DNA was incubated at 37 °C in TE buffer until dissolved. A sample of DNA was run in a 1% agarose gel and stained with ethidium bromide to visualize the DNA. The DNA was then quantified and the Vir regulon amplified using primers VUF and SBR with the following PCR conditions: one cycle at 95 °C (1’), 30 cycles at 95 °C (15" ), one cycle at 60 °C (2’) and one cycle at 68 °C (6’). The amplified products were verified by running the samples in a 2% agarose gel. The DNA was then digested adding either HaeIII or HinfI and incubated in a water bath for two hours. The digested products were run in a 2.5% agarose gel and the bands obtained between 200-4000 bp were compared visually among each GAS strain. We considered a strain to be a distinct clone if at least one band was different.

Social networks evaluation was performed by one investigator blinded to the culture result. This investigator separately interviewed the two persons in charge of the children of each room by asking the question " with whom does each child have the most communication?" The question was applied to a list of approximately ten classmates of each child. Frequency of communication with each classmate was ranged as extensive, some, or little. Because information regarding the communication profiles of each child was obtained from two persons in charge of the child, discrepancies did arise. These were resolved by re-questioning the caregivers until complete agreement was obtained.

Degree is a term used to describe the extent to which persons are tied to others in the " network" . The closeness is a measure of how close or integrated is an actor to an interaction pattern of a local network of actors. An actor is central if he or she has a better efficiency in the connection to others, due to a more direct communication path. The betweenness refers to how a particular actor might be able to control interactions between pairs of other actors in the network. Density is the proportion of ties in a social network. Finally, homophily is a pattern of communication between actors who share some similar attributes. Measures of centrality (degree, closeness and betweenness) were compared among children with and without GAS. Central actors have a maximal centrality index.

Statistical analysis. Descriptive analysis was performed calculating frequencies, means and standard deviation. Comparative analysis included comparison of proportions among independent groups.

Results

GAS isolates and antimicrobial usage. GAS was detected in 2/35 (5.7%) and in 9/37 (24.3%) of the children from rooms B and C respectively. The percentage of children having used antimicrobials during the last two months was as follows: Group B, 12/35 (34.3%); Group C, 11/37 (29.7%). There were no differences in the history of antimicrobial usage between children harboring GAS and those without GAS from either group B (Chi square 1.38, d.f. = 1; p = .43) or group C (Chi square 0.32, d.f. = 1; p = .69). The time elapsed since the last antibiotic dose ranged from 1 to 60 days (median = 5) in the children from group B, and from 2 to 30 days (median = 12) in the children from group C. The antimicrobials most frequently prescribed for those from group B were b-lactams such as ampicillin or amoxicillin (50%), cephalosporins (25%), macrolides (16.6%) and sulfonamides (8.4%); whereas those used by group C were ampicillin/amoxicillin (54.5%), sulfonamides (27.3%), cephalosporins (9.1%) and macrolides (9.1%) (tables I and II).

GAS isolation in children with and without symptoms of upper respiratory infection. Information regarding symptoms of upper respiratory tract infections in the last 15 days was available in 29/35 (83%) children of group B and in 34/37 (92%) of group C. Among these children, 14/29 children of group B had symptoms of an upper respiratory tract infection; of those, 2 had a GAS isolated (2/14, 14%) and 12 were negative for GAS (Fisher’s exact test, 2.30 d.f. 1; p = .22). In children of group C, upper respiratory symptoms were present in 23/34; of these, GAS was isolated in 7 (7/23, 30%) (Fisher’s Exact Test 0.57, d.f. 1; p = .68).

Identification of GAS clones. The RFLP patterns of the various GAS isolates were heterogeneous and four different clones were identified and named I to IV according to the prevalence of the clones. Clone I was isolated in five children of group C; clone II in two children of group C and in one of group B; clone III was isolated in one child of group B and in one of group C; clone IV was isolated in a single child of group C (figure 1).

Partnership among children of group C with identical GAS clones. Among the 5 children harboring clone I, the degrees of communication between each other was: " extensive" (n = 2), " some" (n = 1), " little" (n = 1) and none (n = 1). The 2 children harboring clone II shared " some" degree of communication. Although children harboring clone III belonged to different rooms, they shared a single dining room at lunchtime.

The social network analysis. The density coefficient values of groups B and C were 0.40 (± 0.87) and 0.35 (± 0.80). This means that in group B the proportion of ties among the group was 40%, whereas in group C the proportion was 35%. The homophily pattern of interaction in group B was different in carriers (0.00), non carriers (0.47) and both groups together (0.47), (p = .35). The homophily pattern in group C was also different in carriers (0.46), non-carriers (0.68) and both groups together (0.19), (p = .001).

The centrality data showed that in 4/5 children with clone I, the values of degree, closeness and betweenness were higher than the group mean. In 3/3 children with clone II, values of degree and betweenness were also higher than the group mean. On the other hand, in the two children with clone III and the one with clone IV (the less frequently isolated clones), the values of degree and betweenness were lower than the group mean. The most widely distributed clones (I and II) were associated with those children having the highest centrality levels (figures 2 and 3).

Discussion

Group A Streptococcus was found in the nasopharynx of children from a day-care center in Mexico, and the clonal diversity between the isolated strains as well as the degree of physical interaction between pairs was evaluated. We found that individuals having a high level of centrality in the interaction shared a common GAS isolate.

The prevalence of GAS isolated from the nasopharynx of the children studied was similar to reports from other day-care centers.^4,9 In contrast with the frequent isolation of a single clone of GAS in outbreaks of pharyngitis and other day-care associated infections, multiple clones GAS have been described in healthy GAS carriers,¹⁹ even among those sharing the same space. Indeed, in our study there were four different GAS clones isolated simultaneously from healthy children sharing the same room.

Regarding the measurement of interaction among the children, the density values of the groups showed an expected proportion of ties. On the other hand, the homophily pattern clearly defined two groups of children, i.e., those with and without a GAS isolate. Belonging to any of these groups integrated by some particular characteristics (i.e. age, gender) could facilitate the natural spread of strains of GAS once it has been introduced into one particular group.

Genetic analysis of the isolated bacteria was crucial for demonstrating epidemiological patterns. For example, of the five children carrying the most widespread strain (clone I), four had values of ties with others (degree), mediation between pairs (betweenness), and efficient communication actors (closeness) that were superior to the mean group values. The next most widespread strain (clone II) was isolated in three children whose values of degree and betweenness were also above the mean of the group. Finally, in the two children with clone III and in the one child with clone IV, the values of degree and betweenness were below the group mean.

In conclusion, the GAS strains isolated simultaneously from children sharing the same room were genetically diverse. The sharing of specific GAS clones between carriers was not random but was associated with natural formation of groups based upon some particular characteristic, high proportion of ties among them and a high level of communication centrality.

Acknowledgments

We thank the critical review and important contributions to this manuscript of Dennis L. Stevens M.D. Ph.D. and Amy Bryant Ph.D. from the Department of Infectious Diseases at the V.A. Medical Center in Boise, Idaho.

Financial support. This work was supported in part by the Research Council (FOFOI) of the Instituto Mexicano del Seguro Social (Grants 2000-249/006 and 2004/126).

This work was presented in part at the 103^rd General Meeting of the American Society for Microbiology. Washington DC. May 20, 2003. (Poster number C-125).

References

1. Stevens DL. Invasive group A streptococcus infections. Clin Infect Dis 1992;14:2-13.        [ Links ]

2. Stevens DL. Invasive group A streptococcal infections: the past, present and future. Pediatr Infect Dis J 1994;13:561-566.        [ Links ]

3. Muotiala A, Saxen H, Vuopio-Varkila J. Group A streptococcal outbreak of perianal infection in a day-care centre. Adv Exp Med Biol 1997;418:211-215.        [ Links ]

4. Yagupsky P, Landau D, Beck A, Dagan R. Carriage of Streptococcus pyogenes among infants and toddlers attending day-care facilities in closed communities in southern Israel. Eur J Clin Microbiol Infect Dis 1995;14:54-58.        [ Links ]

5. Falck G, Kjellander J. Outbreak of group A streptococcal infection in a day-care center. Pediatr Infect Dis J 1992;11(11):914-919.        [ Links ]

6. Hsueh PR, Teng LJ, Lee PI, Yang PC, Huang LM, Chang SC, et al. Outbreak of scarlet fever at a hospital day-care centre: analysis of strain relatedness with phenotypic and genotypic characteristics. J Hosp Infect 1997;36:191-200.        [ Links ]

7. Smith TD, Wilkinson V, Kaplan EL. Group A streptococcus-associated upper respiratory tract infections in a day-care center. Pediatrics 1989;83:380-384.        [ Links ]

8. Engelgau MM, Woernle CH, Schwartz B, Vance NJ, Horan JM. Invasive group A streptococcus carriage in a child care centre after a fatal case. Arch Dis Child 1994;71:318-322.        [ Links ]

9. al Qutob R, Na’was T, Mawajdeh S. The role of day-care givers in the identification and prevention of infections. Soc Sci Med 1991;33:859-862.        [ Links ]

10. Weiss K, Laverdiere M, Lovgren M, Delorme J, Poirier L, Beliveau C. Group A Streptococcus carriage among close contacts of patients with invasive infections. Am J Epidemiol 1999;149:863-868.        [ Links ]

11. Nguyen L, Levy D, Ferroni A, Gehanno P, Berche P. Molecular epidemiology of Streptococcus pyogenes in an area where acute pharyngotonsillitis is endemic. J Clin Microbiol 1997;35:2111-2114.        [ Links ]

12. Social Network Analysis in the Social and Behavioral Sciences. En: Wasserman S, Faust K, eds. Social Network Analysis. Cambridge: Cambridge University Press, 1994:3-27.        [ Links ]

13. Freeman, LC. Centrality in Social Networks: I. Conceptual Clarification. Social Networks 1979;1:215-239.        [ Links ]

14. Fitzpatrick LK, Hardacker JA, Heirendt W, Agerton T, Streicher A, Melnyk H, et al. A preventable outbreak of tuberculosis investigated through an intricate social network. Clin Infect Dis 2001;33:1801-1806.        [ Links ]

15. Edmunds WJ, O’Callaghan CJ, Nokes DJ. Who mixes with whom? A method to determine the contact patterns of adults that may lead to the spread of airborne infections. Proc Biol Sci 1997;264:949-957.        [ Links ]

16. Centrality and Prestige. En: Wasserman S, Faust K, eds. Social Network Analysis. 1st. Cambridge: Cambridge University Press, 1994:169-219.        [ Links ]

17. Gardiner D, Hartas J, Currie B, Mathews JD, Kemp DJ, Sriprakash KS. Vir typing: a long-PCR typing method for group A streptococci. PCR Methods Appl 1995;4:288-293.        [ Links ]

18. Hartas J, Hibble M, Sriprakash KS. Simplification of a locus-specific DNA typing method (Vir typing) for Streptococcus pyogenes. J Clin Microbiol 1998;36:1428-1429.        [ Links ]

19. Kaplan EL. The group A streptococcal upper respiratory tract carrier state: an enigma. J Pediatr 1980;97:337-345.        [ Links ]

Address reprint requests to: Dr. Alberto Villaseñor-Sierra. Centro de Investigación Biomédica de Occidente, Instituto Mexicano del Seguro Social. Paseo de la Reforma 476, Col. Juárez, 6698 México, DF. E-mail: avillase@prodigy.net.mx