Læknablaðið - 15.07.1995, Blaðsíða 31

LÆKNABLAÐIÐ 1995; 81
543
Table I. Performance of thefour rapid tests. Specimensfor the QuickVue test, the SureCell test and culture, were collected in an
alternating sequence but the specimen for PCR was always collected last.
QuickVue SureCell Culture PCR
No positive positive positive positive
Infected 45 43 43 43 41
Not infected 195 2 0 0 0
Total 240 45 43 43 41
Table II. Sensitivity, specificity, predictive value of a positive (PVP) and the predictive value of a negative (PVN) of the
QuickVue test, the SureCell test and culture.
Sensitivity % Specificity % PVP % PVN %
QuickVue 96 99 96 99
SureCell 96 100 100 99
Culture 96 100 100 99
years of age with the median age of 19. Ninety
eight patients had symptoms and 142 were
asymptomatic. Of those 240 patients, 45 were
considered to be infected, giving a prevalence
of 18.8% in this population. The performance
of the two different tests is shown in table I.
The sensitivity, specificity, predictive value of
a positive (PVP) and the predictive value of a
negative (PVN) for the different assays is
shown in table II. The calculation was not per-
formed for PCR because of the bias caused by
the specimen collection sequence. In the two
instances where culture was false negative the
three other tests and the comfirmatory tests
were all in agreement. The majority of the
culture positive patients had high inclusion
counts. Only four had five or fewer inclusions
per cover slip. The two false negative Quick-
Vue tests were among those four as was one of
the false negative SureCell tests. The other
false negative SureCell test had a high inclu-
sion count but the specinten was collected last
in the collection sequence.
Discussion
Many new diagnostic tests have come avail-
able for the diagnosis of Chlamydia trachoma-
tis infections in recent years. None of them are
perfect nor are any of them ideal for use in all
situations. The traditional gold standard, cell
culture, is time consuming and expensive and
although it is highly specific it lacks in sensitiv-
ity (11). Also, the specimens used for culture
need to be transported rapidly to the laborato-
ry and under controlled conditions (12). The
antigen detection tests, which are in most
widespread use at present, give more rapid
results and are the least expensive methods for
diagnosing these infections. Many of the anti-
gen detection tests are intended for use in clin-
ical laboratories but some, such as the two
evaluated here, are intended for use in physi-
cians offices. The antigen detection methods
have traditionally been less sensitive than cell
culture for the detection of these infections
(8). The sensitivity is though, heavily depend-
ent on the types of patients studied, the lack of
sensitivity being most pronounced in asympto-
matic males (4,8). All diagnostic tests perform
best, in terms of sensitivity and predictive val-
ue of a positive, in populations as the one
studied in this evaluation, young high risk fe-
males. The two tests, QuickVue and the Sure-
Cell performed extremely well compared to
culture in this study and the latter better than
previously reported (10). The reason is without
doubt the quality of the specimen and the rapid
delivery to the laboratory. The amount of anti-
gen in an infected cervical canal is related to
age, decreasing with increased age of the pa-
tient. Our patients were mostly young high risk
females with high inclusion counts on culture.
The new DNA amplification methods like
PCR and LCR have been shown to be ntore
sensitive than culture (11-16) and the inferior
performance of PCR in this study is undoubt-
edly caused by bias in the specimen collection
sequence, the specimen for this test always
being collected last.
In conclusion it can be said that both of the