Article Text

Download PDFPDF

Screening tests for Chlamydia trachomatis or Neisseria gonorrhoeae using the cobas 4800 PCR system do not require a second test to confirm: an audit of patients issued with equivocal results at a sexual health clinic in the Northwest of England, UK
  1. Mark J Hopkins1,
  2. Godfrey Smith2,
  3. Ian J Hart1,
  4. Fath Alloba3
  1. 1Liverpool Specialist Virology Centre, Royal Liverpool University Hospital, Liverpool, UK
  2. 2Department of Medical Microbiology, Royal Liverpool University Hospital, Liverpool, UK
  3. 3Department of Genitourinary Medicine, Royal Liverpool University Hospital, Liverpool, UK
  1. Correspondence to Dr Mark J Hopkins, Liverpool Specialist Virology Centre, Royal Liverpool University, Hospital, Liverpool L7 8XP, UK; m.hopkins{at}liv.ac.uk

Abstract

Objectives To assess the clinical utility of supplementary PCRs following a positive cobas 4800 CT/NG PCR screening test result.

Methods Laboratory reports, for Chlamydia trachomatis or Neisseria gonorrhoeae, issued to genitourinary medicine patients between April 2010 and April 2011 were reviewed retrospectively. Positive reports were routinely confirmed by supplementary PCRs and N gonorrhoeae culture. Clinical records of patients with unconfirmed positive (equivocal) reports were retrieved to determine if the infection was confirmed by a second sample obtained at patient recall and the impact of this process on antibiotic management.

Results Over 15 000 patients were tested during the study period. The prevalence of chlamydia and gonorrhoea was 972 (5.75%) and 76 (0.50%), respectively. A further 78 chlamydia and 2 gonorrhoea equivocal reports were issued. Only 56 (72%) patients with an equivocal chlamydia report returned to the clinic, and of these, only 41 (73%) gave a second sample to retest. Positive predictive value (PPV) of the PCR screening test was calculated at 98.0% and 97.5% for detection of chlamydia infection from urine and rectal swabs, respectively. Most patients accepted antibiotic treatment before their infection status had been confirmed. Prevalence of gonorrhoea infection was low but the PPV of the screening PCR in urine specimens remained high (98.75%).

Conclusions Equivocal reports introduce delays to patient management, while the risk of unnecessary antibiotic therapy appears acceptable to most patients. The cobas 4800 CT/NG PCR screening assay can achieve UK testing standards (PPV >90%) for chlamydia, and low prevalence gonorrhoea in urine without supplementary tests. A patient-led confirmation algorithm is proposed.

  • Chlamydia infections
  • diagnostics
  • gonorrhea
  • PCR
  • guideline development
  • virology clinical
  • virology HIV
  • virology HPV/HSV
  • virology general laboratory
View Full Text

Statistics from Altmetric.com

Introduction

Laboratories are encouraged to confirm positive Chlamydia trachomatis (CT) or Neisseria gonorrhoeae (NG) results obtained by nucleic acid amplification test (NAAT) platforms by repeat testing. The utility of these confirmation results for chlamydia infection in clinical practice was thoroughly assessed in the USA, and recent UK guidelines suggest repeat testing is not generally required for chlamydia positive results.1–3 However, laboratories should ensure the testing algorithms for both pathogens have positive predictive values (PPVs) >90%, particularly in low prevalence populations or extra-genital specimens, and a supplementary test employed if appropriate.2 ,4

In our laboratory, any NAAT screen-positive sample that was not positive by supplementary testing was reported as unconfirmed positive (equivocal). This prompts a specific patient management pathway in the clinic, whereby patients are recalled, a second sample requested in order to repeat the test, and treatment offered. Patients can choose to delay therapy until the infection status is established, but partner management is only instigated once the result is confirmed.

The aim of this audit was to review performance of the local CT/NG PCR testing algorithm and investigate what proportion of patients issued with an equivocal result were truly infected. This would allow the PPV of the screening PCR assay to be calculated and inform if supplementary testing was still required.

Methods

The laboratory database was used to compile lists of retrospective chlamydia and gonorrhoea reports issued to the Royal Liverpool University Hospital genitourinary medicine clinic over the 12-month period 12 April 2010 to 11 April 2011. Our laboratory employs the Roche cobas 4800 CT/NG PCR assay as the NAAT screen. All PCR screen positive, or invalid results, are confirmed by alternative targets using an in-house CT/NG PCR, and swabs are taken from high-risk patients for N gonorrhoeae culture, in addition to gonorrhoea PCR on urine.5 Electronic patient records were interrogated to establish the clinical impact on the patients issued with an equivocal chlamydia or gonorrhoea report. An anonymised audit proforma was designed to ascertain if a repeat specimen was requested, if the patient returned and how soon, if a second sample was obtained, if infection was confirmed by the second sample, and when the patient was treated.

Results

The results of patient recall, in response to an equivocal result, are shown in table 1. This dataset excludes patients without a valid PCR screening result in order to assess the specificity of this test. Over 28% (22/78) of patients issued with an equivocal chlamydia PCR result did not return to clinic, and of these, a further 26% (15/56) declined to give a second specimen to confirm the infection status. Patients were less likely to submit a secondary swab specimen for chlamydia analysis than a urine sample (55% vs 78%). Most patients accepted antibiotic treatment on their first or second visit to the clinic. Less than 10% (6/65) of the patients preferred to wait for the second specimen chlamydia PCR result before accepting therapy on third visit (range, 1–30 days after 2nd visit). This was higher (50%) in the two patients with equivocal gonorrhoea PCR results; the patient accepting therapy on recall was overtreated, while therapy was delayed a further 7 days after the second visit for one patient who preferred to wait for the gonorrhoea culture result.

Table 1

Summary of audit data showing the number of Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (NG) infections identified by cobas 4800 CT/NG PCR screen and confirmed by second specimens, and the number of patients who waited for the second specimen results before accepting antibiotic therapy

Approximately 50% of patients were likely to have their infection status confirmed as positive when a second urine sample was submitted. This yielded high specificity rates for the screening PCR, with a PPV of >97.5% in all sample types analysed.

A small number of samples (n=161) gave a repeated clot detection error on the PCR screen, and the gonorrhoea infection was confirmed by alternative methods in a relatively high proportion of these results (28/161).

Discussion

Equivocal chlamydia and gonorrhoea reports introduce delays in management decisions for a small number of patients. Data from this audit suggested 91% of the ‘equivocal’ patients preferred to accept a risk of antibiotic overtreatment rather than delay therapy until a third visit to the clinic. However, patient acceptability to instigate contact tracing before their infection status was confirmed, were not assessed at this time.

The low rate of false-positive screens (<2.5%) observed in this audit supports increased confidence in NAAT specificity. Guidance in the US no longer recommends routine repeat testing of NAAT positive specimens for chlamydia or gonorrhoea, and data from this audit also support adequate single PCR assay performance for detection of chlamydia in extra-genital swabs and low prevalence gonorrhoea from urine.6 Performance of the gonorrhoea PCR on other specimen types could not be assessed, as swabs are only tested by N gonorrhoeae culture in our genitourinary medicine clinic.

A limitation of this dataset is that 1.25% (199/16 781) of the samples failed to give a valid result on the cobas 4800 system, and had to be excluded in order to calculate performance as a screening assay. Discussion with the manufacturer identified viscosity as a potential issue in these urine tests and repeat vortexing is recommended. Therefore, a secondary assay must be maintained to analyse these problematic specimens.

Overall, a single step testing algorithm without equivocal reports would yield a faster patient journey to resolution of infection and earlier contact management, but a low number of unidentified false-positive screens would receive unnecessary antibiotic therapy and contact tracing. This simplified policy may enhance the net clinical outcome for genuinely infected patients who fail to return because they originally tested ‘equivocal’. A local decision-making algorithm, which directly involves the patient, is being developed as recommended by National Institute for Health and Clinical Excellence.7 We are considering immediate therapy and contact tracing for all patients with a positive PCR screen result, but also to discuss the small chance of an incorrect result and offer the option to submit a second sample for repeat testing, if any major concerns about the validity of the positive result exist. Swabs for gonorrhoea culture on patient recall should be maintained to inform antibiotic surveillance. Our proposal is based on having ≤2.5% false-positives. If our audit recorded a 10% false-positivity rate, our approach could become ill-advised. This will be the subject of our second audit cycle.

Also, perhaps the time has come to recommend ≥95% PPV for NAAT screening as a national standard.

Key messages

  • A positive cobas 4800 chlamydia PCR screen result does not require a secondary confirmation test because it achieves positive predictive values >97.5%, even in extra-genital swabs.

  • High positive predictive values were demonstrated for low prevalence N gonorrhoeae in urine, without secondary confirmation. Further validation is required for non-genital sites.

  • Secondary assays remain useful to test samples without a valid screen result.

  • Patients informed of an equivocal chlamydia or gonorrhoea result often accept antibiotic therapy rather than wait for a second sample result to confirm infection.

Acknowledgments

The authors are grateful to staff at the Royal Liverpool University Hospital for their assistance with this study.

References

View Abstract

Footnotes

  • Ethics approval Audit approved by the Royal Liverpool University Hospital Clinical Audit Working Group (Ref: 3993-11/12).

  • Provenance and peer review Not commissioned; externally peer reviewed.

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.