Factors associated with asymptomatic non-chlamydial non-gonococcal urethritis in heterosexual men: findings from a case-control study

Abstract

Summary

The significance of asymptomatic non-chlamydial non-gonococcal urethritis (NCNGU) is unclear. Organisms associated with NCNGU, e.g. Mycoplasma genitalium, for which there is no widely available test, are linked to reproductive sequelae in women but UK guidance no longer recommends urethral smear microscopy to screen for asymptomatic NCNGU. This case-control study of heterosexual male genitourinary (GU) medicine clinic attenders aimed to identify clinical, demographic and sexual behaviour factors associated with asymptomatic NCNGU so that we could determine whether the presence or absence of symptoms provides a rational basis for deciding to whom we should offer microscopy and whom we should treat. Men with asymptomatic NCNGU were very similar to men with symptomatic NCNGU, except for more consistent condom use. Asymptomatic and symptomatic NCNGU could be different ends of the same clinical syndrome. Until the microbiological basis of NCNGU is understood, we recommend treatment of men with NCNGU irrespective of symptoms.

Keywords

Sexually transmitted infection non-chlamydial non-gonococcal urethritis NCNGU NGU NSU Mycoplasma genitalium diagnosis treatment microscopy urethral smear Case Control Study

INTRODUCTION

Non-chlamydial non-gonococcal urethritis (NCNGU) is a common condition.¹ It is characterized by microscopic urethritis in the absence (asymptomatic) or presence (symptomatic) of symptoms or signs and with negative tests for Chlamydia trachomatis and Neisseria gonorrhoeae . Asymptomatic NCNGU is of debated significance.^2–5 Although numerous sexually and non-sexually transmitted organisms have been associated with NCNGU,⁶ Mycoplasma genitalium, reported in 15–25% of men with mostly symptomatic NCNGU, is of particular importance as it is associated with reproductive health morbidity in women.^7,8 However, testing for M. genitalium is not routine within most settings^9,10 although microscopic urethritis can be used as a suggestive finding. Current UK practice recommends treatment of men with symptomatic NCNGU but no longer recommends microscopy in asymptomatic men, removing the opportunity for asymptomatic NCNGU to be detected or treated.⁹

Given M. genitalium's association with significant pathology in women but in the absence of a readily available diagnostic test, it is unclear whether we should over or under-treat asymptomatic NCNGU. If we could establish the extent to which risk factors are shared between symptomatic men with NCNGU, asymptomatic men with NCNGU and men in the same setting without urethritis, we could determine whether the presence or absence of symptoms provides a rational basis for deciding to whom we should offer microscopy and whom we should treat.

We hypothesized that men with asymptomatic NCNGU are more similar to men with symptomatic NCNGU than they are to men without urethritis. To test this we undertook a case-control study to identify the demographic, behavioural and clinical factors associated with a diagnosis of asymptomatic NCNGU among heterosexual men attending our genitourinary (GU) medicine services in central London. We excluded men who have sex with men because the public health implications of asymptomatic NCNGU for this group are especially poorly understood.¹¹

METHODS

Current clinical practice assumes a hierarchy of urethritis within the clinical population: (1) symptomatic C. trachomatis or symptomatic N. gonorrhoeae or both; (2) asymptomatic C. trachomatis or asymptomatic N. gonorrhoeae or both; (3) symptomatic NCNGU; (4) asymptomatic NCNGU; (5) asymptomatic and no urethritis. We used this hierarchy as the basis for our comparisons, focusing on NCNGU.

We used a retrospective case-control design to compare demographic, behavioural and clinical factors in heterosexual men associated with a diagnosis of asymptomatic NCNGU (cases) with two comparator groups: (1) symptomatic men with microscopic urethritis and negative tests for C. trachomatis and N. gonorrhoeae (symptomatic NCNGU: positive controls); (2) asymptomatic men without microscopic urethritis and negative tests for C. trachomatis and N. gonorrhoeae (negative controls).

Cases and comparators were sequential male attenders of two inner city London sexual health clinics at Barts Health NHS Trust between August 2009 and February 2010 who met the inclusion criteria. Men were identified retrospectively from the clinics’ GUMCAD coding database using national codes for the appropriate diagnoses and local codes for asymptomatic presentations. Men reported the presence or absence of symptoms on a triage form. For men to be considered symptomatic they had to report at least one symptom consistent with urethritis (urethral discharge, penile irritation or itching, dysuria). Men were excluded from the study if they reported ever having a male sex partner and, if they had balanitis, balanoposthitis, penile dermatosis, genital warts, Molluscum contagiosum or active Herpes simplex virus infection.

Urethritis was diagnosed by a Gram-stained urethral smear containing ≥5 polymorphonuclear leukocytes (PMNL) per high powered (×1000) field (HPF) averaged over five fields with the greatest concentration of PMNLs, in the absence of Gram-negative intracellular diplococci.¹² Urethral specimens were collected using a 5-microlitre plastic loop. First void urine was tested for C. trachomatis and N. gonorrhoeae DNA by the BD Viper system (Becton, Dickinson and Company, Franklin Lakes, NJ, USA). All samples were analysed in the same laboratory service by regularly audited, trained staff.

Demographic, behavioural and clinical explanatory variables, available within the clinic database or through the patient self-triage form, were selected based on published literature and consultation with sexual health experts.

Statistical methods

A sample of 115 heterosexual men with asymptomatic NCNGU, 129 men with symptomatic NCNGU (positive controls) and 309 asymptomatic men without urethritis, C. trachomatis or N. gonorrhoeae (negative controls) provided a power of 80% to detect as significant (at the 5% level) differences between groups of 17% versus 33%. Data were manually extracted from the clinic notes, entered into a Microsoft Access database (2003) and analysed using STATA data analysis and statistical software. The data were cross-tabulated and compared using the Chi-squared statistic. Variables that were statistically significant (P < 0.05) in bivariate analyses were then entered into a multivariate logistic regression model and backward stepwise selection used to identify a parsimonious model of the key demographic, clinical and behavioural factors associated with having (1) asymptomatic NCNGU versus symptomatic NCNGU; and (2) asymptomatic NCNGU versus asymptomatic men without urethritis, C. trachomatis or N. gonorrhoeae.

RESULTS

Data for 115 cases (mean age 27; 50.9% white British or Irish), 129 positive controls (mean age 28; 29.7% white British or Irish) and 309 negative controls (mean age 29; 53.3% white British or Irish) were collected. The characteristics of each group are summarized in Table 1.

Table 1

Comparing the demographic, behavioural and clinical profiles of symptomatic men with NCNGU, asymptomatic men with NCNGU and asymptomatic men without urethritis, chlamydia or gonorrhoea

Factors	Symptomatic NCNGU (n = 129)	Asymptomatic NCNGU (n = 115)	Asymptomatic without urethritis, chlamydia or gonorrhoea (n = 309)	P value for comparison 1	P value for comparison 2	P value for comparison 3
Demographic factors:
Age
<25 years	29.5%	26.1%	21.0%	0.676	0.051	0.420
25–34 years	46.5%	52.2%	58.9%
35–44 years	24.0%	21.7%	20.1%
Median (lower, upper quartiles)	28 (24, 34)	27 (24, 34)	29 (25, 33)
Ethnicity				0.011	<0.0001	0.291
White British/White Irish	29.7%	50.9%	53.3%
White, European/Eastern European	20.3%	13.2%	18.8%
Mixed/other ethnicities	8.6%	3.5%	3.9%
Asian	10.2%	10.5%	10.1%
Black	31.3%	21.9%	14.0%
Behavioural factors
Sexual practices, last 3 months:
Given oral sex	42.5%	58.4%	62.2%	0.014	<0.0001	0.478
Received oral sex	52.0%	70.8%	76.6%	0.003	<0.0001	0.227
Vaginal sex	98.4%	95.6%	88.3%	0.187	0.001	0.026
Insertive anal sex	7.1%	7.1%	12.1%	0.276	0.009	0.144
Time (days) since last vaginal or anal sex				0.673	0.066	0.392
≤3 days	27.3%	30.1%	28.9%
>3 but ≤7 days	21.9%	24.8%	21.9%
>7 but ≤14 days	21.1%	18.6%	16.3%
>14 but ≤28 days	13.3%	7.1%	8.3%
>28 days but ≤3 months	12.5%	14.2%	12.0%
>3 months	3.9%	5.3%	12.6%
Median (lower, upper quartiles)	8.5 (3, 18.75)	7 (2, 15.5)	7 (3, 26)
Condom use for in the last 3 months				<0.0001	<0.0001	0.007
Every time/no vaginal/anal sex in last 3 months	20.5%	17.9%	27.0%
Most of the time	6.3%	27.7%	32.6%
Sometimes/occasionally	41.7%	20.5%	21.8%
Not at all	31.5%	33.9%	18.6%
Number of sexual partners, last 3 months				0.373	0.454	0.676
0/1	41.9%	43.4%	45.3%
2	25.6%	32.7%	27.0%
3–4	17.8%	15.0%	18.2%
5 +	14.7%	8.9%	9.5%
Median (lower, upper quartiles)	2 (1, 3)	2 (1, 2)	2 (1, 3)
Number of new sexual partners, last 3 months				0.028	0.036	0.818
0	33.3%	28.3%	28.7%
1	23.3%	38.9%	35.8%
2 +	43.4%	32.7%	35.5%
Median (lower, upper quartiles)	1 (0, 2)	1 (0, 2)	1 (0, 2)
Clinical factors
Had a previous STI check	83.7%	83.5%	72.2%	0.959	0.010	0.016
Time since last screen
≤3 months	20.0%	6.4%	9.3%	0.045	0.061	0.160
>3 but ≤6 months	15.2%	18.1%	12.5%
>6 but ≤12 months	18.1%	27.7%	21.8%
>12 but ≤24 months	21.9%	26.6%	23.2%
>24 months	24.8%	21.3%	33.3%
Median months (lower, upper quartiles)	12 (3, 26)	12 (6, 24)	18 (7, 36)
Previous STI diagnosis/es
Chlamydia	25.6%	20.0%	15.2%	0.301	0.010	0.238
Gonorrhoea	11.6%	6.1%	3.6%	0.131	0.001	0.251
Non-specific urethritis	21.7%	24.4%	13.6%	0.624	0.035	0.008
Genital warts	3.9%	10.4%	5.2%	0.045	0.561	0.053
Genital herpes	1.6%	5.2%	2.9%	0.108	0.406	0.253
Syphilis	0.0%	1.7%	0.0%	0.133	N/A	0.020
Any of above STIs	51.2%	44.4%	31.7%	0.287	<0.0001	0.015
Ever had a HIV test	71.1%	75.4%	63.4%	0.447	0.125	0.020
Time since (last) HIV test
≤3 months	15.7%	5.9%	10.0%	0.030	0.183	0.101
>3 but ≤6 months	21.4%	17.7%	13.1%
>6 but ≤12 months	18.0%	30.6%	24.6%
>12 but ≤24 months	18.0%	28.2%	21.5%
>24 months	27.0%	17.7%	30.9%
Median months (lower, upper quartiles)	12 (4, 30)	12 (6, 24)	13 (6, 36)
HIV status				0.746	0.193	0.037
HIV-negative	70.5%	74.8%	64.1%
HIV status unknown (as not tested)	29.5%	25.2%	35.9%
HIV-positive	0.0%	0.0%	0.0%
Any medical conditions	16.3%	16.7%	12.0%	0.935	0.226	0.206
Regular medication taken	12.5%	11.4%	10.1%	0.793	0.455	0.689

Note: Column percentages unless stated

Factors associated with asymptomatic versus symptomatic NCNGU

In bivariate analyses, eight factors were associated with asymptomatic versus symptomatic NCNGU (Table 1): reporting giving oral sex; receiving oral sex; condom use in the last three months; new sexual partner(s); time since last STI check; and previous warts diagnosis/es. In addition, there were significant differences in the ethnic profile of men in the two groups and also in the time since the patient's last HIV test.

In the multivariate model, two factors remained associated with asymptomatic NCNGU among men with urethritis (Table 2): condom use in the last three months, which was specifically reporting condom use ‘most of the time’ (AOR 4.17; 95% CI 1.76–12.6) and reporting receiving oral sex in the last three months.

Table 2

Key factors* associated with having asymptomatic NCNGU by comparator

	Crude odds ratio (95% CI)	Adjusted^† odds ratio (95% CI)
Comparator 1: symptomatic NCNGU (positive controls)
Received oral sex	P = 0.014	P = 0.015
No (ref.)	1.00	1.00
Yes	1.90 (1.14–3.17)	2.00 (1.14–3.48)
Condoms use in the last 3 months	P = 0.0001	P = 0.0001
Every time/no vaginal/anal sex in last 3 months (ref.)	1.00	1.00
Most of the time	5.04 (1.91–13.3)	4.71 (1.76–12.6)
Sometimes/occasionally	0.56 (0.26–1.21)	0.54 (0.25–1.19)
Not at all	1.24 (0.59–2.57)	1.18 (0.56–2.50)
Comparator 3: asymptomatic without urethritis, chlamydia or gonorrhoea (negative controls)
Condom use in the last 3 months	P = 0.009	P = 0.007
Every time/no vaginal/anal sex in last 3 months (ref.)	1.00	1.00
Most of the time	1.29 (0.68–2.42)	1.15 (0.59–2.21)
Sometimes/occasionally	1.42 (0.72–2.81)	1.36 (0.67–2.73)
Not at all	2.77 (1.46–5.24)	2.70 (1.41–5.18)
Had a previous STI check	P = 0.018	P = 0.028
No (ref.)	1.00	1.00
Yes	1.95 (1.12–3.38)	1.91 (1.07–3.40)

*Significance level for addition to the model was P < 0.05, although the results for comparator 2 were also obtained when significance level for addition to the model was P < 0.10

^†Odds ratio adjusted for all other factors

Factors associated with asymptomatic NCNGU versus asymptomatic men without urethritis, C. trachomatis or N. gonorrhoeae.

Eight factors were associated with asymptomatic NCNGU versus asymptomatic men without urethritis, C. trachomatis or N. gonorrhoeae (Table 1): a previous sexually transmitted infection (STI) check; previous non-specific urethritis (NSU) diagnosis/es; a previous syphilis diagnosis/es; as well as any previous STI diagnosis/es; previously testing for HIV, with a differences in the HIV status profile between these two groups. In addition, vaginal intercourse was more commonly reported among men with asymptomatic NCNGU. In multivariate analysis only two factors remained statistically significant (Table 2); condom use and reporting a previous STI check.

DISCUSSION

Men with asymptomatic NCNGU reported markedly similar clinical, demographic and sexual risk variables to men with symptomatic NCNGU. However, asymptomatic men with NCNGU reported greater condom usage, and more insertive oral sex, than symptomatic men. Men with asymptomatic NGNGU also share many similarities to asymptomatic men with no infection or urethritis but they report less consistent condom use and are more likely to have been previously tested for STIs, although the difference observed was small. In the bivariate analysis a self-reported previous diagnosis of NSU was associated with NCNGU among asymptomatic men but this was excluded in the multivariate model.

In contrast to many other studies of urethritis, we used explicit, recognized criteria for the diagnosis of asymptomatic NCNGU. Inter-observer variation in urethral smear microscopy cannot be altogether excluded¹³ but all our microscopists are regularly audited. The time since men last passed urine also has an effect on microscopy finding. In our clinic microscopy is undertaken at least two hours post-micturition but precise time since micturition was not consistently recorded. Sensitive and specific tests were used to exclude common bacterial STIs and men with symptoms of other conditions were excluded. However, the retrospective design meant that we were limited to variables which were consistently recorded in the clinical notes, and we were unable to include examination findings. Up to 10% of men are reported to misclassify themselves as asymptomatic when they actually have symptoms^14–16 but it may also be true that asymptomatic men report having symptoms when none are present.¹⁷ The study population is likely to be generalizable to other specialist services but is unlikely to represent community settings.

The finding that men with NCNGU, whether symptomatic or asymptomatic, report such similar risk variables may suggest that they represent different ends of the same clinical spectrum. This is consistent with other bacterial STIs, in that we recognize risk factors for an infection as a whole rather than separately for asymptomatic or symptomatic cases. However, men with no infection and no urethritis also reported broadly similar risk variables, which may reflect the ‘higher risk’ nature of GU medicine attenders.¹⁸

The apparent association between reported condom use and likelihood of having NCNGU, and having symptoms could imply that condoms provide partial protection from transmissible causes: those men with least condom use have the greatest exposure with a correspondingly higher organism burden, which may lead to more noticeable pathology, as seen in other infections.¹⁹ Whether or not this exposure is to normal or pathological vaginal microbiota is not clear.

The association between receiving oral sex and urethritis might be explained by an effect of oral pathogens, both sexually transmitted and non-STI associated.²⁰ However we cannot explain why receiving oral sex was more frequently reported in asymptomatic rather than symptomatic men with NCNGU.

Current guidelines only place significance on symptomatic and not asymptomatic NCNGU. While antibiotic therapy is associated with resolution of symptoms for some men,²¹ it is illogical to ignore asymptomatic NCNGU when the importance of other STIs is not symptom-dependent. While symptom-based triage may allow for the detection of men with a higher prevalence of infections, it misses men who are asymptomatic but have significant sexual risk. We must not assume that asymptomatic is the same as low risk. This group of asymptomatic men could benefit from behavioural interventions to help prevent future infections, an opportunity for health promotion that would be missed when purely focusing on symptom-based triage.

Questions remain about the significance of asymptomatic NCNGU and whether or not its detection and treatment is a public health benefit and/or cost-effective. This study may help to inform and interpret future microbial studies which will be essential to understanding the causes of NCNGU. Until then, we suggest that consideration should be given to performing urethral smear microscopy in asymptomatic men with high sexual risk, and that appropriate treatment should be offered if microscopic urethritis is detected.

Footnotes

ACKNOWLEDGEMENTS

We are very grateful to Dr Liat Sarner, Dr John White and Professor George Kinghorn for their helpful advice on interpretation of findings; to Janet Lynch for help with the clinic database; to Krissie Lucas for help preparing the manuscript.

Contributor statement: CSE and JAC had the original idea for the paper. CSE was the principal investigator of the study and obtained funding from the National Institute for Health Research. LJS assisted set up of data collection. JS collected data, wrote the first draft of the paper and coordinated the writing of the paper, with all authors contributing to subsequent drafts. Plans for analysis were led by CHM who undertook all statistical analyses. CSE is the paper's guarantor.

This paper presents independent research commissioned by the National Institute for Health Research (NIHR) under its Programme Grants for Applied Research scheme (RP-PG-0707–10208). The views expressed in this abstract are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health. The authors have no competing interest to report.

References

All new STI episodes seen at GUM clinics in the UK: 1999–2008. See www.hpa.org.uk/web/HPAwebFile/HPAweb_C/1215589014474 (last checked 15 February 2013)

Ross

. Screening asymptomatic men for non-specific urethritis. Sex Transm Infect 2007;83:79

O'Mahony

. Asymptomatic Chlamydia trachomatis-negative non-gonococcal urethritis. Int J STD AIDS 2005;16:330–1

Donovan

. Asymptomatic non-chlamydial, non-gonococcal urethritis – an iatrogenic disease? Sex Health 2004;1:65–7

Maw

Robinson

. Asymptomatic urethritis; the case for a considered view! Int J STD AIDS 2004;15:849–50

Shahmanesh

Moi

Lassau

, 2009 European guideline on the management of male non-gonococcal urethritis. Int J STD AIDS 2009;20:458–64

Haggerty

. Evidence for a role of Mycoplasma genitalium in pelvic inflammatory disease. Curr Opin Infect Dis 2008;21:65–9

Svenstrup

Fedder

Kristoffersen

, Mycoplasma genitalium, Chlamydia trachomatis, and tubal factor infertility – a prospective study. Fertil Steril 2008;90:513–20

Ross

Ison

Carder

Sexually Transmitted Infections: UK National Screening and Testing Guidelines. See www.bashh.org/documents/59/59.pdf : British Association for Sexual Health and HIV, 2006

10.

The Sexual Health Society of Victoria. National Management Guidelines for Sexually Transmissible Infections . See http://www.racp.edu.au/index.cfm?objectid=588CCF7A-07BF-0153-DB433D495BD214EE

11.

Saunders

Hart

Estcourt

. Is asymptomatic non-chlamydial non-gonococcal urethritis associated with significant clinical consequences in men and their sexual partners: a systematic review. Int J STD AIDS 2011;22:338–41

12.

British Association for Sexual Health and HIV. UK National Guidelines on the Management of Non-gonococcal Urethritis (2007) – Dec 2008 Update. See http://www.bashh.org/documents/1955

13.

Smith

Copas

Prince

, Poor sensitivity and consistency of microscopy in the diagnosis of low grade non-gonococcal urethritis. Sex Transm Infect 2003;79:487–90

14.

Falk

Fredlund

Jensen

. Symptomatic urethritis is more prevalent in men infected with Mycoplasma genitalium than with Chlamydia trachomatis. Sex Transm Infect 2004;80:289–93

15.

Leung

Eastick

Haddon

, Mycoplasma genitalium is associated with symptomatic urethritis. Int J STD AIDS 2006;17:285–8

16.

Horner

Thomas

Gilroy

Do all men attending departments of genitourinary medicine need to be screened for non-gonococcal urethritis?

Int J STD AIDS 2002;13:667–73

17.

Serisha

Evans

. How accurate is the information on triage forms? HIV Med 2010;11(Suppl. 1):1–119

18.

Johnson

Wadsworth

Wellings

, Who goes to sexually transmitted disease clinics? Results from a national population survey. Genitourin Med 1996;72:197–202

19.

Michel

Sonnex

Carne

, Chlamydia trachomatis load at matched anatomic sites: implications for screening strategies. J Clin Microbiol 2007;45:1395–402

20.

Bradshaw

Tabrizi

Read

, Etiologies of nongonococcal urethritis: bacteria, viruses, and the association with orogenital exposure. J Infect Dis 2006;193:336–45

21.

Carlin

Barton

. Azithromycin as the first-line treatment of non-gonococcal urethritis (NGU): a study of follow-up rates, contact attendance and patients’ treatment preference. Int J STD AIDS 1996;7:185–9