Answer

a, c—The rapid urine screening test can be administered to large numbers of men much quicker and more cheaply than the PCR diagnostic assay. However, unlike the PCR assay, the rapid urine screening test does not provide an accurate and reliable diagnosis of chlamydia infection. Answer a is, therefore, correct.

The purpose of rapid urine screening is to identify men in the population who are at high risk of chlamydia infection, as indicated by a "positive" result. Such individuals will then be recommended to have a PCR assay to confirm diagnosis. Time and money will be saved by not undertaking diagnostic testing with PCR assay in men for whom there is no suspicion of chlamydia infection, indicated by a "negative" ("low risk") result on rapid urine screening.

The very nature of screening tests means that diagnosis is not always correctly predicted. In particular, a "negative" result may be obtained on rapid urine screening yet the PCR assay diagnoses chlamydia infection. Equally, a "positive" result may be obtained on rapid urine screening yet the PCR assay test does not diagnose a chlamydia infection.

Sensitivity and specificity are indices that can be used to describe the performance of rapid urine screening when identifying those men who should undergo diagnostic testing with PCR assay. To evaluate these indices, a sample of the population will undergo both the rapid urine screening test and the diagnostic PCR assay, as took place in this study.¹

Answer c is the correct definition of sensitivity in this instance. Sensitivity describes the accuracy of rapid urine screening at identifying those men who were diagnosed with chlamydia on PCR assay. Of the 109 men in total diagnosed by the PCR assay, 90 (82.6%) had a "positive" result on rapid urine screening. These 90 individuals were, therefore, correctly identified by the rapid urine screening test. The remaining 19 individuals were incorrectly identified by the rapid urine screening test ("negative" result), as the result disagreed with the diagnostic PCR assay.

View this table: [in this window] [in a new window]   Sensitivity and specificity of the rapid urine screening test

 
Answer b describes positive predictive value, not sensitivity. The positive predictive value reports the accuracy of a "positive" result on rapid urine screening, namely the percentage of men diagnosed with chlamydia infections on PCR assay who also have a "positive" screening result. Of the 107 men in total that had a "positive" ("high risk") result on rapid urine screening, 90 (84.1%) had chlamydia infection according to the PCR assay. The remaining 17 (15.9%) men were incorrectly identified by the rapid urine screening test, because the result disagreed with the result of the diagnostic PCR assay.

Answer d is not the correct definition of specificity in this case. Specificity expresses the accuracy of rapid urine screening at identifying those who were not diagnosed with chlamydia on PCR assay. Of the 1102 men in total who were not diagnosed with chlamydia on the PCR assay, 1085 (98.5%) also had a negative ("low risk") result on rapid urine screening. These 1085 individuals were correctly identified by the rapid urine test, as the result agreed with the diagnostic PCR assay.

Cite this as: BMJ 2009;339:b3941