|
The syndromic
approach to diagnosis of cervicitis among female sex workers in
Benin is more sensitive than a laboratory-based approach when the
rate of return visits is taken into account.
Syndromic
versus laboratory-based diagnosis of cervical infections among
female sex workers in Benin.
Mukenge-Tshibaka L, Alary M, Lowndes CM, Van Dyck E,
Guedou A, Geraldo N, Anagonou S, Lafia E, Joly JR.
Sexually Transmitted Disease 2002 29:324-330.
Summary:
Question
What is the performance of syndromic diagnosis of cervical N.
gonorrhoeae and C. trachomatis infections and what is its
practical utility in comparison with a strategy based on locally available
laboratory results for female sex workers in Benin?
Design
A comparison of the diagnosis of cervicitis based on the presence of
clinical signs to diagnosis made by laboratory detection of N.
gonorrhoeae and C. trachomatis using locally available tests
and using nucleic acid amplification tests performed in a reference
laboratory.
Participants
481 female sex workers who presented for the
screening visit of a multicenter randomized clinical trail on the
effectiveness of the microbicide COL-1492 (Advantage 24) in Cotonou,
Benin.
Description of Tests and Diagnostic
Standard
The syndromic diagnosis for cervicitis was
based on the presence of one of the clinical signs of cervicitis (mucopus,
a positive swab test, or the observation of blood on the endocervical
swab), a diagnosis of pelvic inflammatory disease (cervical motion
tenderness and/or lower abdominal pain during the bimanual palpation of
the pelvis), or a count of more than 10 polymorphonuclear leukocytes per
field in a vaginal smear. Other signs were also noted, including the
presence of vaginal discharge. An endocervical swab sample was collected
and immediately inoculated onto modified Thayer-Martin medium for the
detection of N. gonorrhoeae by conventional culture methods. A
second endocervical swab was used to detect C. trachomatis by an
enzyme immunoassay (MicroTrak EIA, Syva, Palo Alto, CA). Positive samples
were confirmed by a blocking assay. A third endocervical swab was frozen,
shipped to the reference laboratory, and used for detection of C.
trachomatis and N. gonorrhoeae by PCR amplification (Amplicor
CT/NG, Roche Diagnostic Systems, Branchburg, NJ). Samples positive for N.
gonorrhoeae were confirmed using a 16S rRNA PCR (Roche Diagnostic
Systems). A true positive test was defined as isolation of N.
gonorrhoeae in culture, a positive Amplicor PCR for N. gonorrhoeae
(confirmed by an N. gonorrhoeae 16S rRNA assay), or an
Amplicor positive for C. trachomatis. The women were given
treatment if indicated by the syndromic diagnosis, and asked to return
within 10 days for their laboratory results and possible treatment
adjustments.
Main Outcome Measures
The sensitivity, specificity, positive
predictive value (PPV), and negative predictive value (NPV) of the
syndromic approach and of laboratory-based detection for the diagnosis of N.
gonorrhoeae or C. trachomatis cervical infection as determined
by the "gold standard" described above.
Main Results
The performance characteristics of the
different approaches are presented in the table. Higher sensitivity was
achieved when vaginal discharge noted on the examination was considered in
the syndromic diagnosis. Within the 10-day period after screening, 215
(44.7%) of the participants returned to the clinic. When the return rate
of the infected women within the fixed 10-day period was taken into
account, the sensitivity of the local laboratory tests was reduced from
74.6% to 28.8%.
| Performance
of different approaches for the diagnosis of C. trachomatis
and N. gonorrhoeae cervical infections among 481 female sex
workers in Benin |
| Diagnostic
Approach |
Sensitivity
%
(n=118) |
Specificity
%
(n=363) |
PPV
(%) |
NPV (%) |
| Syndromic |
48.3 |
74.7 |
38.3 |
81.6 |
| Syndromic
+ vaginal discharge |
66.1 |
58.7 |
34.2 |
not
available |
| Laboratory
tests, if all return |
74.6 |
99.5 |
97.8 |
92.3 |
Laboratory
tests
if 44.7% |
28.8 |
99.7 |
97.1 |
not
available |
| Laboratory
tests (44.7% return) + syndromic |
61.9 |
74.4 |
44.0 |
not
available |
Authors' Conclusions
Despite its poor performance, when the low
rate of return visits by female sex workers to receive their results is
taken into account, the sensitivity of the syndromic approach is higher
than that of the local laboratory tests.
Source of funding: United Nations
Programme on HIV/AIDS
For correspondence: Michel Alary,
Groupe de Recherche en Epidemiologie de l'Universite Laval, Centre de
Recherche, Hopital du Saint-Sacrement du CHA, 1050 Chemin Ste-Foy, Quebec,
QC G1S 4L8, Canada. E-mail address: michel.alary@gre.ulaval.ca.
|
.gif){kind=small}