Evaluation and comparison of tests to diagnose Chlamydia trachomatis genital infections.

Infection with Chlamydia trachomatis results in intracytoplasmic inclusions and the generation of infectious elementary bodies (EBs). These can be detected by various procedures. Staining of epithelial cells with vital dyes was first used to detect inclusions, but is insensitive. Thus, Papanicolaou-stained cervical smears cannot be recommended. The advent of the ability to grow chlamydiae in cultured cells over 30 years ago had a major impact on chlamydial research and on detection. However, this procedure is probably <70% sensitive for cervical infection and less for urethral infection in men and is now practised infrequently following the advent of other, mostly less laborious and often equally, or more sensitive detection systems. Thus, staining a smear with a specific fluorescent monoclonal antibody to detect EBs is simple and the direct fluorescent antibody tests became a commercial proposition in the early to mid-1980s. Nevertheless, although highly sensitive and specific in competent hands, technical expertise is crucial and even the most experienced may be unable to read a large number of stained smears on slides quickly. In view of this, it is understandable that enzyme-linked immunosorbent assays (ELISAs) gained popularity from the mid-1980s onwards, for they are not very labour intensive and their reading is neither subjective nor tedious. Unfortunately, these aspects outweighed the fact that the ELISAs lack sensitivity, some being very insensitive. The situation has been rescued, however, by the advent in the early 1990s of methods that amplify chlamydial DNA, making it easily detectable by relatively simple procedures. The polymerase chain reaction is such a method and has high specificity and sensitivity, although commercial development has so far not met the high standard expected of it in terms of sensitivity. The ligase chain reaction does not invoke such criticism, and high values for both sensitivity and specificity may be expected, even on urine samples. This augers well for diagnosing an infected individual patient and for effective screening programmes. Antibody tests have no place in a screening programme and are of debatable value in diagnosis.