APH 2001, 59, 167-180:
Management of a prostate cancer screening research program in the city of Antwerp.
B. Standaert, A. Alwan, B. Dourcy-Belle-Rose, V. Nelen and L. Denis
Keywords: infant mortality, Brazil, family health program
|Objective: To investigate the value of the
management of a prostate cancer-screening research program in the city
of Antwerp through the evaluation on non-participation.
Methods: The cancer screening study targeted men aged 55 to 74 years living in two districts (Deurne and Borgerhout) of Antwerp city. The districts are subdivided in 37 sectors with a total of 11,382 subjects to be contacted. Retrospective analysis of non-participation per sector was studied regarding sector specific variables and the reasons of non-participation. It was hypothesized that sector specific characteristics may influence the rate of non-participation. Moreover it was searched if some of these variables could be correlated with individual reasons of non-participation. Sector specific variables were: the average distance to the research center; the invitation period; the age-composition; the social class; the population density; the general practitioner (GP) density; and the district area (Borgerhout or Deurne). The individual reasons asked among the home visited non-participants were: "No Interest" in the program; "Consult aGP"; "Being Absent" during the invitation period; and "Other Reasons".
Correlation statistics, non-parametric tests, and regression analysis are used to determine variables that may explain the variation in the rate of non-participation per sector.
Results: Average non-participation rate per sector was
70% (range 61% to 79%). Non-participation was highly correlated
with "Invitation Period" and "Age-Specific"
groups. Individual responses were highly correlated with these
variables and inter-correlated between "Being Absent",
"Consult a GP" and having "No Interest" and
"Other Reasons". Multiple regression analysis suggests
that the rate of non-partcipation per sector increases
significantly with "Being Absent", and a dominant presence in
a sector of 55-59 years old subjects. The fitted model explained
around 60% of the variation.