Comment on Editorial by B. Donovan: Asymptomatic non-chlamydial non-gonococcal urethritis — an iatrogenic disease (Sexual Health 1, 65–67) The role of the urethral Gram stain in non-gonococcal urethritis

We agree with the comments of Donovan in the August issue of Sexual Health, that microscopy of the urethral smear has no place in screening men without symptoms or signs of urethritis, but would also suggest it has a limited role in treatment decisions in men with symptoms of non-gonococcal urethritis (NGU). We have recently completed a case control study of men with urethral symptoms that supports this point.

We compared 80 heterosexual men with symptoms of NGU to 79 controls without symptoms of urethritis. ¹ In this study 39% of cases did not have evidence of urethritis [< 5 polymorphonucleocytes (PMN) per high power field (hpf)] on their urethral gram stain, despite having symptoms consistent with NGU. Men with urethral symptoms, but no urethritis on Gram stain, still had a higher prevalence of Chlamydia trachomatis than the asymptomatic control group (16% versus 1%, P < 0.002). First pass urine samples were also collected from the control group without symptoms of urethritis. Among these individuals 18% had evidence of urethritis (≥ 10 PMN per hpf of centrifuged urine), but no urethral pathogens were detected in this group. Chlamydia trachomatis was isolated from only one control, and he had no evidence of urethritis on microscopy.

It is time to re-examine the relevance of the laboratory definition of urethritis in men with NGU. Studies pre- and post-introduction of molecular testing, report low sensitivities of the urethral Gram stain in predicting the presence of Chlamydia trachomatis. It is apparent that a significant proportion of men with chlamydial infection have <5 PMNs per hpf on their urethral gram stain. Haddow et al. recently examined 363 consecutive men attending an STD clinic in the UK, 8% had chlamydia detected by LCR and 26% had 5 or > PMN phpf. ² Of the men with chlamydial infection a large proportion, 37%, had <5 PMN phpf on their urethral smear. While urethral PMN count was associated with the presence of C. trachomatis, the urethral gram was only 63% sensitive and 77% specific for chlamydial infection, and dysuria and young age were considered as predictive for chlamydial infection. From the premolecular era, Janier concluded that in men with symptoms of NGU but an absence of urethral discharge, the situation was even worse, and the urethral gram had a sensitivity for C. trachomatis of only 29%. ³

In Australia it is common practice to perform a urethral smear in symptomatic men, to assist in the diagnosis of gonorrhoea, however the decision for treatment is based on the presence of urethral symptoms and not on the urethral PMN count. With increasing evidence that urethral pathogens are common in symptomatic males with normal urethral microscopy, it seems appropriate clinical practice to base treatment on the presence of symptoms of NGU, and not to place great importance on the findings of microscopy. This approach is more relevant and feasible in the majority of clinical settings worldwide, where most clinicians do not have access to onsite microscopy.