Article Text

Download PDFPDF

Resolution of the recent performance problem of Abbott LCx Chlamydia trachomatis assay. Issues of repeat testing for confirmation of chlamydial infection
  1. H Mallinson,
  2. J Hopwood,
  3. K Mutton
  1. PHLS Liverpool, University Hospital at Aintree, Lower Lane, Liverpool L9 7AL, UK
  1. Correspondence to:
 Dr Mallinson;

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

In February 2001, Abbott Laboratories issued a device correction notice to users of their LCx Chlamydia trachomatis assay suggesting that initially reactive ligase chain reaction (LCR) tests should be repeated on the same sample to validate the test result. A recent alert (December 2001) from the Medical Devices Agency (MDA, DA2001(09)) indicates that the device correction is still in force and points out the resource implications where retesting is required. We offer some data on LCR performance characteristics during this period and before.

The Department of Health pilot study on “Opportunistic screening for genital chlamydial infection in Portsmouth and Wirral” ran for a year up to October 2000. During that study, the standard adopted for reporting chlamydial infection included a repeat LCR test on all first catch urine samples that were initially LCR positive. Samples giving discrepant LCR results were further tested by Roche Cobas (PCR) polymerase chain reaction. Chlamydia LCR urine screening, with repeat LCR/PCR testing of positives, has continued in the Wirral pilot area and is also being used in other research projects locally.

Following the original device correction, we continued to carry out a repeat LCR but additionally included a PCR test on all initially positive LCR urine samples. Analysis of our data (table 1) suggests that compared to the baseline (satisfactory) performance during the Wirral pilot there was indeed a noticeable LCR reproducibility problem when the device correction notice was issued. Since then, however, the LCR performance has improved gradually to be at least as good as in the pilot period.

The MDA alert properly deals with kit performance in generating a valid test result. However, this incident also prompted us to consider the wider issues of repeat testing for confirmation of chlamydial diagnosis.

We have recently also examined the reproducibility of our Roche Cobas chlamydia PCR results and are concerned to have found that of 282 initially PCR positive urine samples only 237 gave repeat PCR positive results.

We sense that there may be a mistaken view adopted by some clinicians that all nucleic acid amplification tests (NAAT) are infallible for sensitivity and specificity. It is important that patients should be made aware (as we did during the screening pilot) that no test is 100% accurate. Problems of reproducibility have been reported for both LCR1 and PCR.2 We recognise the dilemma in repeat testing of samples that give positive reactions in chlamydia NAATs; on the one hand, a low organism load in the specimen makes repeat positivity a matter of statistical chance of retesting a portion with detectable numbers—so cases will be missed. On the other hand, repeat confirmation ensures a more robust diagnosis is made which is so important in the light of the major implications of a chlamydia diagnosis for those who consider themselves well but decide to take a screening test. We would welcome debate on the need for retesting or independent confirmation of positive chlamydia NAATs and support the need for continuous monitoring of all tests to ensure their consistent optimal performance.

Table 1

Repeat LCR testing and PCR testing of initially positive LCR urines during the Wirral Chlamydia Pilot (Sept 1999 to Oct 2000, baseline) and for 3 month periods since the issue of the device correction (February 2001)