Objectives: The ability of molecular methods to detect low levels of nucleic acid has led to the widespread application of techniques based on nucleic acid amplification tests in microbiological diagnosis. This exquisite sensitivity is recognised in the laboratory to require stringent precautions to avoid contamination, but this is not widely appreciated in clinical settings where samples are initially collected, and may be a particular problem in the non-clinical settings used for sampling as part of the National Chlamydia Screening Programme. There is thus the need to characterise the risk of false-positive results caused by environmental contamination in these areas.
Methods: The extent of environmental contamination of Chlamydia trachomatis (CT) nucleic acid in clinical settings was investigated by swabbing surfaces within the vicinity of specimen collection. Laboratory experiments were designed to monitor the persistence of ribosomal RNA under simulated conditions and to investigate whether contamination of patients’ specimens is a risk if environmental surfaces are contaminated. The Gen-Probe APTIMA Combo 2 system was used for CT rRNA detection.
Results: CT rRNA was detected in swabs taken from examination rooms and toilet areas. Tests showed that this could persist for at least 50 days. The potential for clinical samples to become contaminated as a result of the presence of CT rRNA in the immediate environment was demonstrated in this simulated test.
Conclusion: This study demonstrated that there is a risk of false-positive nucleic acid amplification test results, when samples are taken in an area that is contaminated with target nucleic acid.
Statistics from Altmetric.com
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.
Molecular diagnostic tests for the detection of Chlamydia trachomatis (CT) offer a superior level of sensitivity compared with shell vial culture and enzyme-linked immunosorbent assays.1 This also permits the use of non-invasive sample types such as urine. Nucleic acid amplification tests (NAAT) are being used for CT detection in the National Chlamydia Screening Programme (NCSP), which aims to tackle the escalating incidence of CT in asymptomatic 16–25-year-old individuals.2
The exquisite sensitivity of NAAT demands a scrupulous cleaning regime in the laboratory to ensure that traces of target nucleic acid do not contaminate the samples being tested, giving rise to false-positive results. The NAAT procedure takes place in segregated areas separating pre-amplification and post-amplification, to minimise contamination of sample preparation areas with amplified target. Before and after each run in our laboratory, all surfaces and equipment used for APTIMA assays are cleaned with, or immersed in, hypochlorite (1000 ppm). After this, they are rinsed with distilled water to remove hypochlorite residues. Environmental swabs of equipment and surfaces are processed with each batch of samples to ensure that the area has been adequately cleaned.
Such meticulous cleaning regimes are not carried out at the sites used for sample collection, eg toilet areas, therefore there is a possibility that CT ribosomal RNA on the surrounding surfaces could contaminate the specimen during sampling.
Self-collection kits for diagnostic samples have enabled the NCSP to enrol more of the target population compared with clinically collected specimens and the proportion of self-taken samples is likely to increase as the uptake improves.
Numerous studies have reported that self-collected swabs are as good as clinician-taken specimens.3 4 Self-collection may, however, not be so suitable for patients who speak little English who may not fully understand the instructions. Female patients who have never used tampons may find it difficult to take their own vaginal swabs, and put components of the sampling kit on the floor or other potentially contaminated surfaces while they figure out what to do.
Inanimate surfaces in public toilets may be contaminated with CT rRNA target, spread around by urine splashes or improper hand washing. CT nucleic acid has previously been shown to be present in swabs of public toilet bowls using the AMPLICOR Chlamydia trachomatis test (Roche Diagnostics Ltd., Sussex, UK) in a study by Potasman et al.5 in 1999.
A false-positive CT result can have important consequences for the patient concerned. The misguided implication of infidelity may lead to the break up of relationships. Also, anxiety about reproductive health and the social stigma associated with sexually transmitted infections could have a profound psychosocial impact.6 There is also the direct and indirect corollary of inappropriate antibiotic treatment, such as hypersensitivity and the selection of resistant bacterial strains
False-positive test results could also lead to litigation against the agency performing the testing.7
This project aimed to investigate whether environmental contamination with CT rRNA poses a significant risk of obtaining false-positive CT results in patients’ samples that would otherwise be negative.
The objectives were to investigate the extent of environmental contamination in areas used for the collection of patient samples, to establish how long detectable CT rRNA persists on inanimate surfaces and to investigate the likelihood of environmental CT rRNA contaminating patients’ samples during specimen collection.
Molecular detection of CT rRNA
All swabs were tested in the laboratory using the Gen-Probe APTIMA Combo 2 assay (Gen-Probe Inc., San Diego, California, USA). Swabs were taken using APTIMA collection kits and were stored and processed according to the manufacturer’s instructions. An alternative monospecific primer set targeting a different region of CT nucleic acid was used to confirm positive results.
Investigating CT rRNA environmental contamination
A total of 104 environmental swabs were collected on three separate occasions at the genitourinary medicine (GUM) clinic at the Norfolk and Norwich University Hospital. Sites included toilet seats, floors, examination room curtains and light switches.
The GUM facilities were cleaned after every consultation according to infection control recommendations, using isopropan-2-ol. This is sufficient to inactivate CT, but will not degrade the nucleic acid.
Thirteen environmental swabs were also taken during two visits to the local NHS walk-in centre, in which patients regularly self-sample as part of the NCSP.
A higher proportion of swabs were taken at environmental sites for which the recovery of CT rRNA was considered more likely, or would be more likely to result in patient sample contamination, for example examination couches and toilet ledges.
Student’s t-test was used to compare the relative light unit (RLU) values of the CT-positive environmental swabs with an equal number of randomly selected CT-positive patient specimens.
Investigating the environmental persistence of CT rRNA
Before and after each experiment, the area of the bench to be used was thoroughly cleaned with 1000 ppm hypochlorite solution and then washed with distilled water. A negative control swab was taken to ensure that no residual CT rRNA remained on the surface, and a negative control swab spiked with CT rRNA was used to make sure that no residual PCR inhibitors were present on the bench surface.
Two hundred and 500 μl aliquots of pooled previous-positive samples (>50 CT elementary bodies/ml) were dispensed onto specific 7 cm2 areas of the bench, labelled 1–8. A control swab was immersed into the pooled positive sample to check that the material contained a detectable level of CT rRNA.
Swabs of areas 1–8 were taken at 24 hours, and 2, 7, 10, 12, 20, 36 and 50 days, respectively. Each swab was repeatedly swept firmly over the entire allocated area for 15 seconds.
Investigating the spread of detectable CT rRNA from environmental spillages
One millilitre of material from an endocervical swab, containing more than 100 CT elementary bodies per field at 400 times magnification by fluorescent light microscopy, was used to inoculate an area of 25 × 20 cm2 on a bench, which was spread evenly using a pastette. A control swab was taken to ensure that the material contained a detectable level of CT rRNA (swab A). The sample used in this part of the project was from a patient who had recently been tested, and was negative for antibodies to HIV, hepatitis B virus and hepatitis C virus.
After thorough hand washing, a volunteer placed a hand directly onto the contaminated surface. A test swab was taken to check whether the CT rRNA had been transferred onto the hand (swab B).
The hand was placed back on a contaminated area and then placed firmly on to a clean area of the bench for five seconds. Swab C was taken from the bench to check whether any CT RNA from the hand had been transferred to the clean surface.
The hand was then placed back onto a different area of the contaminated surface before being used to handle a swab and replace the cap of a sample tube, re-creating how a patient would handle the specimen (swab D).
CT rRNA in the environment
A total of 117 swabs were taken. At the GUM clinic the floor swabs gave the highest incidence of contamination, with seven out of eight giving a positive result (88%). Five out of the seven swabs of the examination room curtains were also found to be positive (71.4%), as were six out of nine (67%) of the light switches. The incidence of positive results for the examination lamps, toilet seats and urinals were seven out of 14 (50%), three out of nine (33%), and five out of 10 (50%), respectively, and three out of 12 (25%) of the sinks/taps gave positive results. Finally, two out of nine (22%) of the door handles, one out of seven (14%) of the laboratory coats and two out of 15 (13%) of the examination couches gave positive results. Fig 1 illustrates these results.
At the NHS walk-in centre, positive CT results were obtained from two out of three swabs of ledges next to the lavatory in the washroom facilities (a place where components of the specimen collection kit are likely to be placed). One out of two floor swabs was also positive. The results are shown in table 1.
The mean RLU values obtained from the 117 CT-positive environmental swabs and the 117 randomly selected CT-positive patient swabs were 630.12 (range 52–2538) and 1029.09 (range 131–1902), respectively (full dataset not shown). There was strong statistical evidence of a difference between these two groups (p = 0.011).
Persistence of detectable levels of CT rRNA
When the entire area was swabbed with small volumes (<200 μl) of positive material and spread onto 7 cm2 of clean bench, no detection was observed, even immediately after the aliquots were dispensed.
When aliquots of 500 μl were used, detectable levels of CT rRNA were observed up to 50 days later.
Transmission of CT rRNA from sample spillages
The transmission of CT rRNA from a wet contaminated surface on to hands was demonstrated, as was the subsequent transmission from the contaminated hand to a clean surface. A positive CT result was also obtained when the wet contaminated hand was used to simulate specimen collection.
When this experiment was tried using a dry contaminated area (and the same material as used above), transmission was demonstrated from the dirty surface to the clean surface, although a swab of the contaminated hand was negative. No contamination of the specimen was detected when the dry “dirty” hand was used to handle a swab. These results are shown in table 2.
CT rRNA has been found on many surfaces in the vicinity of where clinical specimens are collected, including toilet ledges, examination room curtains and lamps. It is likely that clinicians’ and patients’ hands and/or specimen containers could come into contact with such surfaces before or during sampling.
CT rRNA has been found to contaminate the environment of sampling sites
CT rRNA in the environment may contaminate samples to cause false-positive results
Persistence for at least 50 days has been demonstrated. It is important to note that the material used for the persistence testing was derived from pooled positive samples, contained in a transport medium designed to protect the CT rRNA, as the laboratory no longer receives plain urine samples for CT testing. The stability of nucleic acid in urine is highly variable,8 but the detection of CT in environmental swabs of clinical settings suggests that the nucleic acid can persist for at least a couple of days, and the gradual degradation may be reflected in varying RLU counts. The findings of this part of the study may be more applicable to assays with DNA targets, as this molecule is less prone to hydrolysis and is therefore more stable in the environment compared with RNA.9
The transmission of detectable levels of CT rRNA from surface to surface, and into a negative specimen that had been handled with a wet contaminated hand, has been demonstrated. These findings should be confirmed by repeat experiments, but suggest that false-positive results may be obtained as a result of contamination of the sampling environment when wet spillages have occurred. Realistically, however, patients who come into contact with a wet spill are likely to wash and/or wipe their hands dry before sampling. Contact with dry contamination did not result in false-positive results, but again, these simulated tests should be carried out several times to achieve a more accurate conclusion.
Sodium hypochlorite is effective at degrading nucleic acid and is used routinely in the laboratory to clean surfaces and equipment.10 Bleach-based cleaners are likely to be used in toilet facilities, but the frequency and thoroughness of the cleaning may be a problem, and the logistics of sampling should be taken into account in the light of this study. The investigation of other suitable cleaning materials may be the focus of further studies.
It is impossible to estimate the proportion of patients that may have been affected by false-positive results caused by environmental contamination. Presumably, most screens are from patients who accept that they may have been exposed to CT, or those who have symptoms concordant with the diagnosis. For these reasons it can be expected that most patients who are found to be positive will not challenge the result. The RLU values of CT-positive environmental swabs were well above the threshold value of a positive result, but were generally much lower than a typical CT-positive patient specimen, and there was strong statistical evidence of a difference between these two groups (p = 0.011). The Gen-Probe APTIMA pack insert warns that a high proportion of results with low RLU may be false positives, but it also addresses the fact that the RLU value of a true positive could be lower than that of a false positive, possibly as a result of poor sampling. Indeed, the highest RLU value achieved in this study was not from a clinical specimen but from a floor swab (RLU 2538). By monitoring the RLU values from specific sampling sites, it would be possible to highlight where environmental contamination might be a problem, prompting further investigation at the site. It may also be useful to take routine environmental swabs at sampling sites after screening sessions.
Although the Gen-Probe APTIMA system was used in this project, the findings are just as relevant to all highly sensitive molecular-based detection systems.
Further studies are needed to confirm and characterise the risks identified in this project. It is important to state that no investigations into the viability of the detectable CT material have been included and this study does not suggest that CT infection is acquired from washroom facilities or that clinics are not taking appropriate hygiene measures.
In the light of the findings in this study, it may be necessary to consider a more frequent cleaning regime in examination rooms and toilet areas when swabs intended for molecular tests are being taken. Alternatively, different cleaning products might be more effective. An extra reminder for patients to make sure that their hands are clean and dry throughout the self-collection process may also help to limit the risk of false-positive results from environmental contamination.
NAAT have been selected for the NCSP because of their superior sensitivity over culture, immunosorbent assays and immunofluorescence, and it is of little doubt that a significant number of infections have been correctly identified as positive, which would have otherwise have been missed by conventional methods. The improved sensitivity of these tests, however, brings different problems, and clinicians and patients should be aware of this new perspective on false-positive results. When there are no apparent high-risk factors for sexually transmitted infections confirmatory specimens should be recommended.
These findings also substantiate the need for the cautionary interpretation of NAAT results for medicolegal purposes.
The authors would like to acknowledge the staff from the Norfolk and Norwich University Hospital virology laboratory, the GUM clinic, and the NHS walk-in centre for their help with this project. The authors are also grateful to Gen-Probe for their contribution towards the cost of consumables.
Competing interests: None declared.