Literature review > Issue 7 > Review on Smith et al. 

Microscopy remains a cornerstone in the diagnosis of urethritis, with clinical guidelines usually defining non-gonococcal urethritis as the presence of greater than four polymorphonuclear leukocytes (PMN) per high power field (HPF) [1]. However, even with widely used screening and diagnostic tools, issues such as the ability of the diagnostic maneuver to diagnosis the disease and the interobserver reliability need to be examined. Well-established screening tools such as Pap smears have shown limitations in both their sensitivity and interobserver reliability [2, 3]. This paper examines the diagnostic accuracy of microscopy for NGU.

Using ten experienced staff who routinely undertake microscopy, Smith et al. compared the findings of these individual microscopists to two reference standards: the findings of a senior microscopist and to a composite score derived from the findings of all ten microscopists (consensus score). Microscopists were asked to examine Gram stained urethral smear slides that were defined as negative (< 5 PMN/HPF), low grade (5-20 PMN/HPF) and high grade (>20 PMN) and to place the slides into these categories. The microscopists findings were compared to both the senior microscopist findings and the consensus findings. Then, the authors compared the estimated variance between microscopists and the standard error to determine if the intermicroscopist variation was statistically significant.

Readings for excluding urethritis and for high-grade urethritis proved to be satisfactory, with 97% and 94% of slides respectively being correctly diagnosed in these groups. However, low-grade urethritis proved challenging, with less than 70% of slides being correctly diagnosed in this group. This difference between slide grades was statistically significant. In addition, the proportion of slides correctly diagnosed varied substantially across microscopists in this same category.

This study was limited by the number of microscopists that participated. As the authors point out, particularly in the low-grade urethritis slides, the variation across microscopists was substantial, but this did not achieve statistical significance due to the small number of microscopists.

It is difficult to discern if these participating microscopists are representative of the skill level of clinician microscopists globally. Although described as experienced, the background and training of these microscopists was not outlined in detail, leaving readers to interpret whether microscopists outside this setting would have similar, better, or poorer microscopy skills than the participants.

The reference standard(s) for this study also provide a challenge. A consensus reference standard based on interpretations from 3 to 4 blinded senior microscopists who were not study participants, as opposed to a consensus from study participants, could have strengthened the reader's confidence with the reference standard.

Despite its small numbers, clinicians who rely on microscopy to diagnosis urethritis need to consider the findings of this study. Low-grade urethritis was found to be difficult to diagnosis consistently in a group of well-trained, experienced microscopists. Given the personal and medical ramifications of a diagnosis of a sexually transmitted infection, this study further highlights the role of routine use of validated diagnostic tests, particularly in cases of low-grade urethritis. Microscopy seems to be more useful in ruling out the diagnosis of urethritis, or in cases when there is high-grade urethritis. Clinicians who use microscopy for diagnosis need to be mindful of the limitations of this screening tool, even the hands of experienced microscopists.

References:

1. MMWR Recommendations and Reports. Sexually Transmitted Diseases Treatment Guidelines, 2002. CDC

2. Stoler MH, Schiffman M. Interobserver reproducibility of Cervical cytology and Histologic interpretations. JAMA 2-1;285(11) 1500-1505

3. Nanda K, McCrory DC, Myers ER et al. Accuracy of the Papanicolaou Test in Screening for and Follow-up of Cervical Cytologic Abnormalities: A systematic review. Ann Intern Med. 2000;132:810-819

about SDI | newsletters | grants | publications | literature reviews

WHO Home - WHO Search - TDR Home - SDI Home - SDI Contact us
(c) WHO/OMS 2001