The results of a proficiency panel of lyophilized urine samples showed high specificity and lower sensitivity when tested for C. trachomatis by laboratories using nucleic acid amplification tests.

The results of a proficiency panel of lyophilized urine samples showed high specificity and lower sensitivity when tested for C. trachomatis by laboratories using nucleic acid amplification tests.
2003
VerKooyen RP, Noordhoek GT, Klapper PE, Reid J, Schirm J, Cleator GM, Ieven M, Hoddevik G.
Journal of Clinical Microbiology
Summary:
Question
How well do 96 clinical laboratories detect C. trachomatis in a panel of urine samples using commercial and in-house nucleic acid amplification tests?
Design
A panel of lyophilized urine samples that were negative, weakly positive, and strongly positive for C. trachomatis were sent to 96 laboratories in 22 countries for analysis using the nucleic acid amplification test (NAT) normally performed by the laboratory.
Participants
The test panel consisted of ten 1.7 ml portions of lyophilized urine samples: three C. trachomatis negative, two strongly C. trachomatis positive, and five weakly C. trachomatis positive, formed by mixing pooled samples obtained from 3 healthy and 40 C. trachomatis positive men and women visiting STD clinics in Rotterdam, The Netherlands, and Oslo, Norway. Samples positive at 100-fold dilutions were considered strongly positive. The weakly positive samples contained 10 times the concentration of the lowest concentration that gave a 50% chance of a positive test result on testing six times by each of four different commercial NATs.
Description of Tests and Diagnostic Standard
The proficiency panel was distributed from the production laboratory by surface mail at ambient temperature to 105 laboratories in 22 countries. Quality assurance for the test panel was performed by four reference laboratories using the Abbott LCx, BD ProbeTec ET, COBAS Amplicor PCR, or GenProbe TMA assay. Laboratories performing the BD ProbeTec received three vials of each specimen, which were pooled after reconstitution to obtain the 4 ml required by the assay.
Main Outcome Measures
The results of all negative, weakly positive, and strongly positive urine samples were determined as a whole and by NAT method used.
Main Results
One hundred two data sets from 96 laboratories showed that 199 (97.5%) of 204 strongly positive, 466 (92.1%) of 506 weakly positive, and 302 (99.3%) of 304 negative samples were reported correctly. Eighty-nine per cent of data sets scored 90% or higher. Four laboratories that reported performing NATs in a single room had a significantly higher number of incorrect results (6 of 39) than other laboratories (41 of 945) (P = 0.008). The results of all samples classified by NAT method used are shown in the table. In-house PCR tests had the lowest score. There was no significant difference in scores between the Roche PCR and the Abbott LCx tests.
Authors' Conclusions
A very low number of false-positive results was reported. Weakly positive samples were missed approximately three times more frequently than the strongly positive samples. Lyophilization of clinical urine samples for NATs is a successful method for obtaining stable, standardized samples for quality assurance testing.