• Probabilistic record linkage and an automated procedure to minimize the undecided-matched pair problem Articles

    Machado, Carla Jorge; Hill, Kenneth

    Abstract in Portuguese:

    O relacionamento probabilístico permite que fontes de informações do mesmo registro e em bancos de dados distintos sejam unificadas. Apresenta-se um procedimento utilizado quando se espera que um registro de um banco de dados corresponda a apenas um outro num segundo banco. As fontes de dados foram os nascimentos e óbitos infantis da coorte de nascimentos de 1998, na cidade de São Paulo, Brasil. Os dados relacionados com o mais alto escore e relação unívoca foram utilizados como padrão-ouro e concorreram para a decisão sobre pares obtidos sem relação unívoca. Um comportamento esperado dos dados univocamente relacionados em termos da diferença nas datas de registro de óbito e de nascimento, e também dos locais de registro de nascimento e de óbito para óbitos neonatais foi observado e, aplicou-se esta relação aos demais dados. O número de pares com relação unívoca aumentou substancialmente, de 2.249 para 2.827, e diminuiu o número de nascimentos ligados a um óbito. Este procedimento deve ser associado à revisão manual (procedimento padrão na presença de incerteza) a fim de conseguir um pareamento eficiente.

    Abstract in English:

    Probabilistic record linkage allows the assembling of information from different data sources. We present a procedure when a one-to-one relationship between records in different files is expected but not found. Data were births and infant deaths, 1998-birth cohort, city of São Paulo, Brazil. Pairs for which a one-to-one relationship was obtained and a best-link was found with the highest weight were taken as unequivocally matched pairs and provided information to decide on the remaining pairs. For these, an expected relationship between differences in dates of death and birth registration was found; and places of birth and death registration for neonatal deaths were likely to be the same. Such evidence was used to solve for the remaining pairs. We reduced the number of non-uniquely matched records and of uncertain matches, and increased the number of uniquely matched pairs from 2,249 to 2,827. Future research using record linkage should use strategies from first record linkage runs before a full clerical review (the standard procedure under uncertainty) to efficiently retrieve matches.
  • Leisure-time physical activity among women in a neighbourhood in Bogotá, Colombia: prevalence and socio-demographic correlates Articles

    Gómez, Luis Fernando; Mateus, Julio Cesar; Cabrera, Gustavo

    Abstract in Portuguese:

    O estudo mediu a prevalência de exercício físico em mulheres de 18 a 69 anos de idade e identificou correlatos sociais no bairro de Santa Fé, em Bogotá, Colômbia. Os níveis de atividade física foram calculados por meio de um estudo populacional com uma amostra de 1.045 mulheres. Um modelo de regressão logística identificou fatores associados com sedentarismo. De acordo com os resultados, 79,1% das mulheres foram classificadas como inativas; 15,7% relataram atividade física irregular e apenas 5,2% praticavam exercícios físicos regularmente. Após o ajuste das covariáveis, a inatividade física estava associada com a falta de participação em atividades dominicais de lazer e com a falta de planos para perder peso. Os resultados do estudo demonstram a alta proporção de mulheres fisicamente inativas numa comunidade de baixa renda de Bogotá. Estratégias de intervenção devem ser desenvolvidas para reforçar os exercícios físicos promovidos nas áreas de lazer reservadas aos domingos pelo governo municipal de Bogotá.

    Abstract in English:

    This study measured the prevalence of leisure-time physical activity in women from 18 to 69 years of age and identified correlated social factors in the neighborhood of Santafe, Bogotá, Colombia. Levels of physical activity were calculated through a population survey (n = 1,045). Logistic regression modeling identified factors associated with inactivity. Some 79.1% of respondents reported being inactive; 15.7% practiced physical exercise irregularly; and 5.2% regularly practiced physical exercise. After adjustment of covariates, physical inactivity was associated with not participating in recreational weekend activities on Sundays and not planning to lose weight. The results of this study show the high proportion of physically inactive women in a community in Bogotá Intervention strategies should be developed to reinforce recreational weekend activities on promoted by the Municipality.
  • The impact of health-management training programs in Latin America on job performance Articles

    Díaz-Monsalve, Sonia Janeth

    Abstract in Spanish:

    Se desarrolló un estudio en México, Colombia y El Salvador para determinar el impacto de un programa de capacitación en el desempeño de trabajo de los gerentes de salud. El diseño del estudio fue quasi-experimental (no-randomizado) comparando dos grupos: el grupo de intervención con 85 gerentes en los tres países que recibieron la capacitación y, el grupo control con 71 gerentes que no recibieron la capacitación. Después de ejecutar el programa de enseñanza de 18 meses de duración (el cual incluyó talleres de cinco días de duración y compromisos), el mejoramiento en el desempeño fue medido utilizando 12 indicadores. Los resultados mostraron que, en México, el grupo control mostró 8.3 veces más debilidades en el desempeño que el grupo de intervención; en Colombia, el valor fue 3.6 y en El Salvador 2.4 veces mayor. Factores asociados con el éxito de los resultados del programa de capacitación fueron: (a) las técnicas de enseñanza utilizadas, (b) el fortalecimiento de factores asociados con el buen desempeño y, (c) mecanismos de refuerzo.

    Abstract in English:

    A study was undertaken in Mexico, Colombia, and El Salvador to determine the impact of a management training program on health managers' job performance. A quasi-experimental design was used where in the baseline study an intervention group of 85 district health managers in the three countries was compared with a control group of 71 managers who did not receive the training program. After the implementation of an 18-month training program (which included 5-day training workshops and a series of tasks to be carried out between the workshops), the outcome in terms of improved job performance (i.e. use of predefined management techniques) was measured through twelve management performance indicators. The data collection tools were two questionnaires, participant observation in managers' workplaces, focus group discussions, staff interviews, and document analysis. In Mexico, the control group showed 8.3 times weaker management performance compared to the intervention group; in Colombia the value was 3.6 and in El Salvador 2.4. Factors associated with a successful training outcome were: (a) training techniques, (b) strengthening of enabling factors, and (c) reinforcement mechanisms.
Escola Nacional de Saúde Pública Sergio Arouca, Fundação Oswaldo Cruz Rio de Janeiro - RJ - Brazil
E-mail: cadernos@ensp.fiocruz.br