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.
More research is needed to determine the cost effectiveness of testing for urethritis
Although more evidence has accumulated since questioning the role of testing for urethritis in asymptomatic men in 2002,1 there is as yet no definitive answer. Men with asymptomatic urethritis have 2–3 times the risk of having Chlamydia trachomatis and/or Mycoplasma genitalium detected compared with those with no urethritis (table 1). I am concerned that abandoning testing for urethritis could do more harm than good in high risk asymptomatic men.
Testing for urethritis in men attending departments of genitourinary medicine has the following purposes.
To allow immediate treatment of men with C trachomatis and/or M genitalium with an associated reduction in on-going transmission in the community.2,3 Currently there is no commercial test for M genitalium.
For men at high risk of HIV, it is a potential marker for increased HIV susceptibility and infectivity.2
ANALYSIS OF CURRENT LITERATURE ON URETHRITIS WITH A UNIFYING HYPOTHESIS ON THE AETIOLOGY
The literature on urethritis is full of contradictory findings, which make interpretation difficult. I believe that we need to be able to explain these conflicting observations, in order to understand the true value of testing for urethritis in clinical practice. For example (1) Angarius et al detected C trachomatis and/or M genitalium in only 26% of men with acute urethritis, whereas Falk et al, Totten et al and Horner et al observed >45%.5,8,9,10 (2) Why do some studies show that urethritis identifies >80% of people with C trachomatis (and M genitalium)5,7,8,11 but others do not?12,13
Possible explanations for conflicting observations
In some studies the urethral smear was more representative of the urethral inflammatory response than others. This will be related to both the technique of obtaining and preparing the urethral smear and probably how long patients have held their urine (given the long-standing clinical practice of undertaking an early morning smear in symptomatic patients who initially test negative for urethritis). There is no internationally recognised standardised technique for testing for urethritis. There are at least five different methodologies in the literature for diagnosing urethritis,1,5,7,9,13 and in one recent study patients only had to be symptomatic to be defined as having urethritis!12 In addition, inter-observer and intra-observer error, especially in samples with low-grade inflammation (5–20 polymorphonuclear leucocytes/ high power field), may also play a role.13–15
The populations studied varied in degree of risk (behaviour and age) for having a sexually transmitted infection (C trachomatis detection is associated with a younger age16).
There is confusion about the term “asymptomatic” which is often assumed to mean that the person does not have a urethral discharge. About 10% of men will have a discharge on examination which is not reported as a symptom.7M genitalium is associated with urethral discharge.5,7
Other causes of urethritis
Partner studies, although limited, suggest that up to 25% of patients with microorganism-negative acute urethritis may have a partner infected with either C trachomatis or M genitalium.2,4–6 Although ureaplasmas can cause urethritis, their exact role remains unclear and probably only account for ∼5–15% of acute urethritis.17,18 The importance of Trichomonas vaginalis probably depends on the prevalence in the local population.17 Herpes, adenovirus and urinary tract infections probably account for <5% each.17,19 What causes the remainder is not known. It remains to be shown whether another major pathogen will be identified.
My group’s work suggests that the risk of an STI increases as the degree of inflammation increases and that the symptoms, discharge, dysuria and/or an observable discharge, are surrogate markers for the degree of inflammation.20 Or looked at the other way round, it implies that urethritis can have non-pathogenic causes—for example, bacterial vaginosis2,21—and this is more likely in men with low-grade urethritis. This challenges the idea of having a simple cut-off and labelling all those with <5 polymorphonuclear cells per high power field as at low risk of having a C trachomatis and/or M genitalium infection, and all those with ⩾5 polymorphonuclear cells per high power field as at high risk. Thus men with asymptomatic urethritis are more likely to have a low-grade urethritis with a lower risk of being caused by an STI than if they were symptomatic, but at increased risk compared with asymptomatic men without urethritis. This reduced risk also probably applies to their partner(s) testing positive for an STI even if they test microorganism negative, although the evidence is conflicting.4,6
To fully assess a patient’s risk (and that of their partner(s)) of having either infection, one needs to consider, age, sexual behaviour, clinical presentation, and the results of testing for urethritis. Table 1 details the estimated risks according to clinical findings based on published evidence currently available.
IS THE GRAM-STAINED URETHRAL SMEAR THE BEST METHOD FOR DETECTING URETHRITIS?
As hypothesised in (1) above, it is likely that a Gram-stained urethral smear is more reliable in some centres than others in detecting urethritis. The technique described by Wiggins et al,20 although too complex for routine clinical practice, offers the opportunity of investigating how best to obtain, and evaluate, a specimen that is representative of the urethral inflammatory response. This would provide an objective evidence base for not only helping to interpret studies but also to develop an international standard for future research which can then be translated into clinical practice.
Potential role of leucocyte esterase testing in asymptomatic men
Given the variability of a Gram-stained urethral smear in detecting urethritis (see above), especially at low grades,15 are there other ways of testing for urethritis? Although the leucocyte esterase test has insufficient sensitivity to detect urethritis,17 Marrazzo et al16 observed in a study of over 1500 asymptomatic men using a nucleic acid amplification technique (NAAT) that the leucocyte esterase test had a positive predictive value (PPV) of 13% and an NPV of 97.7% for the detection of C trachomatis compared with 20% and 97.8% for the Gram-stained urethral smear. Horner and Taylor-Robinson22 have recently argued that the leucocyte esterase test, which is both inexpensive and non-invasive, offers an interim, evidence-based, solution to the issue of whether asymptomatic men attending departments of genitourinary medicine should be screened for the presence of urethral inflammation.
CLINICAL ROLE OF TESTING FOR URETHRITIS IN ASYMPTOMATIC MEN
The questions are therefore (1) is this of benefit to the patient and the public health and (2) could testing do more harm than good (to be addressed by Dr Shahmanesh in accompanying editorial)?
If we consider rationalising/minimising testing for asymptomatic men, there are a number of options available.
Use NAATs for C trachomatis and Neisseria gonorrhoeae on a first catch urine specimen only22 with all gonococcal-positives confirmed by culture. It is well recognised that some men with N gonorrhoeae are asymptomatic2 and would be missed if the microorganism was not tested for, and Horner and Taylor-Robinson22 advocate testing for both, but acknowledge the increased risk of false-positives not only because it is a low-prevalence population23 but also because some NAATs can detect commensal Neisseria species.24–26
As for (1) but examine and only undertake a Gram-stained urethral smear for those with a discharge.
As for (2) but include a leucocyte esterase test on those without a discharge.
The disadvantages of option (1) are:
May result in other pathology being missed in some men
Misses the opportunity of the intimacy of a genital examination to help enable the patients to disclose concerns of a deeply personal nature
A group of men and their partners(s) with 10–20% (urethritis positive) risk of C trachomatis and/or M genitalium as per table 1 will be missed
Option (2) would address the first three of these points, and option (3) all of them. Although the leucocyte esterase test has a lower PPV than a Gram-stained urethral smear (see above), it still has a high NPV (>97.5%)—that is, those with a negative leucocyte esterase test are at a substantially lower risk of having an STI.16
Option (3) was introduced in 2006 in Bristol, with the examination being optional for the patient. This strategy is likely to be most cost effective in: (1) younger men (<25 years old) with high risk behaviour in whom (a) the PPV for an STI will be highest (23% for C trachomatis)16 and (b) the risk of transmitting an STI to a new sexual partner before microbiological results are available is greatest 3,27; (2) men at increased risk of HIV, as inflammation increases both susceptibility and infectivity.2 It is also likely to be preferred by patients who have had a casual relationship within a regular relationship, because of the improved NPV associated with a failure to detect urethritis.
As genitourinary physicians, we need to decide whether it is an effective use of our resources to make a complete assessment of a man’s risk of having or being recently exposed to an STI. In order to do this, I believe that we need to consider, age, sexual behaviour, clinical presentation, and the results of testing for urethritis. A complete risk assessment potentially makes the consultation more complex, but need not be significantly more time consuming during the initial assessment, if we use non-invasive testing for both N gonorrhoeae and C trachomatis and urethritis using a NAAT and leucocyte esterase test respectively.22 Given the increasing pressure to achieve the government’s 48-hour access target for departments of genitourinary medicine31 and the fact that better utilisation of resources must be part of the solution, this would seem a reasonable evidence-based compromise in the debate about testing for urethritis in asymptomatic men.22 Clearly more research, with standardised methodology, to allow rapid translation of findings into clinical practice, is urgently required on the aetiology, diagnosis, acceptability and cost effectiveness of testing for urethritis in departments of genitourinary medicine.
More research is needed to determine the cost effectiveness of testing for urethritis
Dr P J Horner would like to thank the Jefferiss Trust for its support and Dr Judith Berry for her helpful comments on the manuscript.