Publicação
Modelling the dynamics of Streptococcus pneumoniae transmission in children
| Resumo: | Streptococcus pneumoniae, also known as pneumococcus, is a commensal bacterium very common in the nasopharynx of young children but that can also be found in older children and adults. Carriage may lead to infection. Although this is a rare event, it has a significant impact on human health. Indeed, diseases caused by pneumococcus include infections as common as otitis media and as dangerous as pneumonia or meningitis. It is therefore important to gain more knowledge of its transmission dynamics as modulated by environmental factors and of how it is affected by host population specificities. To characterize pneumococcus transmission dynamics in a Portuguese day-care center, data was used from a one-year longitudinal study on the state of colonization by S. pneumoniae in children attending a day-care center in Lisboa, Portugal [Sá-Leão et al.2008]. The data refer to 1998, before the introduction of the seven-valent pneumococcal conjugate vaccine. A conceptual model for pneumococcus transmission was built, which considered genotype colonizations and clearances as dependent on the number of carriers, the number of non-carriers and the values of four parameters: the clearance rate m, the within-group transmission parameter b, the community rate of acquisition k and the between-genotypes competition parameter f. Bayesian inference was used to estimate these parameters. Colonizations and clearances were modeled as Poisson processes and the joint posterior probability distributions of the model’s parameters were estimated by Markov Chain Monte Carlo sampling. The number of transitions that occurred in each sampling interval was counted directly from the sampled states, assuming that children did not undergo more than one transition per sampling interval. The posterior mean for the transmission parameters were 0.5974for b, 0.0107for k, 0.6280 for f and 0.3059 for m. Data was simulated using the posterior estimates for these parameters from a study of Finnish DCCs [Hoti et al.2009]. Sampling this data monthly, the method was found to give biased estimations, since the assumption that children did not undergo more than one transition per sampling interval did not hold. The precision could be improved by sampling for a longer period, 30 months were used. To significantly improve the accuracy, the sampling interval needed to be extremely short, daily samples were taken from the simulation. The estimation model used was found to be impractical. Another estimation method should be used that infers the possible carriage histories consistent with the observed states. |
|---|---|
| Autores principais: | Pessoa, Delphine |
| Assunto: | Streptococcus pneumoniae Pneumococcus Teses de mestrado - 2010 |
| Ano: | 2010 |
| País: | Portugal |
| Tipo de documento: | dissertação de mestrado |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade de Lisboa |
| Idioma: | inglês |
| Origem: | Repositório da Universidade de Lisboa |
| Resumo: | Streptococcus pneumoniae, also known as pneumococcus, is a commensal bacterium very common in the nasopharynx of young children but that can also be found in older children and adults. Carriage may lead to infection. Although this is a rare event, it has a significant impact on human health. Indeed, diseases caused by pneumococcus include infections as common as otitis media and as dangerous as pneumonia or meningitis. It is therefore important to gain more knowledge of its transmission dynamics as modulated by environmental factors and of how it is affected by host population specificities. To characterize pneumococcus transmission dynamics in a Portuguese day-care center, data was used from a one-year longitudinal study on the state of colonization by S. pneumoniae in children attending a day-care center in Lisboa, Portugal [Sá-Leão et al.2008]. The data refer to 1998, before the introduction of the seven-valent pneumococcal conjugate vaccine. A conceptual model for pneumococcus transmission was built, which considered genotype colonizations and clearances as dependent on the number of carriers, the number of non-carriers and the values of four parameters: the clearance rate m, the within-group transmission parameter b, the community rate of acquisition k and the between-genotypes competition parameter f. Bayesian inference was used to estimate these parameters. Colonizations and clearances were modeled as Poisson processes and the joint posterior probability distributions of the model’s parameters were estimated by Markov Chain Monte Carlo sampling. The number of transitions that occurred in each sampling interval was counted directly from the sampled states, assuming that children did not undergo more than one transition per sampling interval. The posterior mean for the transmission parameters were 0.5974for b, 0.0107for k, 0.6280 for f and 0.3059 for m. Data was simulated using the posterior estimates for these parameters from a study of Finnish DCCs [Hoti et al.2009]. Sampling this data monthly, the method was found to give biased estimations, since the assumption that children did not undergo more than one transition per sampling interval did not hold. The precision could be improved by sampling for a longer period, 30 months were used. To significantly improve the accuracy, the sampling interval needed to be extremely short, daily samples were taken from the simulation. The estimation model used was found to be impractical. Another estimation method should be used that infers the possible carriage histories consistent with the observed states. |
|---|