|
The use of
self-reportable screening criteria as part of a population-based
genital C. trachomatis screening program that targets a
specific age group is not recommended.
Value of
self-reportable screening criteria to identify asymptomatic
individuals in the general population for urogenital Chlamydia
trachomatis infection screening.
Anderson B, van Valkengoed I, Olesen
F, Møller JK, Østergaard.
Clinical
Infectious Diseases 2003;36:837-844.
Summary:
Question
Can the use of self-reportable selective screening criteria based on
sexual behavior, lifestyle, and sociodemographic characteristics limit the
number of individuals needed to be screened for C. trachomatis
infection in a population-based screening program for Danish men and women
aged 21 to 23?
Design
Using information obtained from a questionnaire and laboratory diagnosis
of C. trachomatis, a set of self-reportable screening criteria was
developed for a randomized part of a population of young, asymptomatic
Danish men and women. The usefulness of the criteria and of previously
published screening criteria for selective C. trachomatis screening
were validated on the remaining part of the population.
Participants
One thousand one hundred seventy-five women and 1033 men randomly selected
from among all 21 to 23 year olds living in Aarhus County, Denmark, who
were asymptomatic (i.e., they did not seek medical care at a doctor's
office for a C. trachomatis test prior to the screening offer) were
tested.
Description of Tests and Diagnostic
Standard
Specimens for laboratory detection of C. trachomatis were obtained
using an in-home sampling technique (vaginal flush sample for women and
first void urine sample for men). Specimens were mailed to the laboratory
and analyzed using a transcription-mediated amplification assay (TMA,
Gen-Probe).
Questionnaires were constructed to
include information contained on previously published studies of
self-reportable screening criteria as well as sociodemographic and
lifestyle characteristics. The questionnaires contained 37 questions for
women, 30 for men, could be completed by checking boxes, and were sent to
individuals with their laboratory results. Univariate and multivariate
logistic regression analysis were performed on a randomly allocated subset
of 60% of the participants (development group) with the self-reportable
characteristics as independent variables and the chlamydia diagnosis as
the dependent variable to identify selective screening criteria. The
validity of the selective screening criteria was assessed on the remaining
40% of the participants (validation group).
Main Outcome Measures
The predictive value of each variable defined by the development group was
calculated for each case in the validation group. The predicted values
were used to calculate the area under the receiver operator characteristic
curve (AUC) of the validation group. The percentage of infections detected
and the percentage of the population to be screened by use of the
screening criteria identified in this and previously published studies
were calculated.
Main Results
Among those who were tested and submitted a questionnaire, 78 (7.3%) of
1074 women and 49 (6.0%) of 846 men were C. trachomatis positive by
TMA. Gender was not a significant infection risk so a common model for men
and women was constructed. The results of the multivariate analysis on the
development group (n = 1152) are shown in Table 1. The AUC for the
development and validation (n = 768) groups were the same (0.68),
indicating consistency between them. The proportion of persons who would
need to be tested in each group to find a certain proportion of infections
is shown in Table 2. The previously published screening criteria, when
applied to this population, performed less well than in the original
studies and less well than the criteria established in this study.
Authors' Conclusions
Selective screening criteria are highly
related to the population on which they are based. The predictive value of
this model was not strong enough to justify selective screening over
population-based screening in this population.
Source of funding: Aarhus County,
Danish Medical Association Research Fund, Danish Research Foundation for
General Practice, Nycomed's Research Foundation for members of the Danish
College of General Practitioners, Danish Center for Evaluation and Health
Technology Assessment (DACEHTA).
For correspondence: Berit Andersen,
Research Unit and Dept. of General Practice, Vennelyst Blvd. 6, 8000
Aarhus C, Denmark. E-mail address: ba@alm.au.dk.
|
.gif){kind=small}