An evaluation of common standard diagnostic tests for dry eye

Abstract

Introduction Dry eye is a common complaint in patients attending eye clinics. It is characterised by a variety of ocular surface abnormalities and has typical symptoms of dryness, foreign body sensation and discomfort. Although a number of clinical tests and techniques are available for the diagnosis of dry eye, such as the Cotton Thread Test (CTT), the Schirmer Test (ST), Rose bengal staining, fluorescein staining, tear meniscus height (TMH) measurement and the Mc-Monnies' dry eye questionnaire (Mc-DEQ), the condition is nevertheless difficult to diagnose. Most research findings related to dry eye have been on Caucasian eyes and recent studies have shown that the tear stability and tear volume of Hong Kong (HK)-Chinese are significantly less than those reported for Caucasians. In view of the differences in tear characteristics between Caucasians and Chinese, the internationally used Mc-DEQ (whether in English or Chinese (the Mc-DEQC)) may not be applicable to Chinese, and the effectiveness of the commonly used tear assessment tests for discriminating between normal and dry eye in Chinese is unknown. Results The prevalence of Chinese dry eye was determined by a retrospective survey of patient files (dry season 741 subjects, humid season 857 subjects) from a private eye centre in HK. The prevalence of dry eye was about 3% and was not affected by mean humidity within the range of 60% to 72 %. Foreign body sensation was the most common ocular symptom, and dryness was also commonly reported. Based on ten ocular symptoms, high specificities (> 83%) but poor sensitivities (< 65%) were obtained from the predictive equations. The poor sensitivity was influenced by the relatively small number of dry eye subjects (n=25). We found no statistically significant relationship between the results of the SPCTT and the TMH, or between the findings of the PRT Test and the TMH. Our results showed that most responses to the questions in the Mc-DEQC could not discriminate between normal and dry eye subjects. Neither the total scores nor the discriminant scores of the Mc-DEQC were able to discriminate between normal and dry eye. The SP-CTT, the PRT Test and the Mc-DEQC were unable to discriminate between normal and dry eye subjects and had poor specificities and sensitivities. When more stringent diagnostic criteria were applied, the specificities and sensitivities of the SP-CTT and the PRT Test did not improve, however the specificity and sensitivity of the Mc-DEQC did somewhat. The sensitivity of the TMH as a diagnostic tool for dry eye was found to be over 80% at a cut-off value 0.21 mm, however the repeatability of measures is such that one measure could indicate normal and a repeated measure on the same eye could indicate dry eye. Neither the Mc-DEQC nor the SD-DEQC was a good screening test for Chinese dry eye. Both DEQ had poor specificity and sensitivity when we used less stringent diagnostic criteria. When more stringent criteria were used, both the specificity and sensitivity of the SD-DEQC were improved somewhat. Conclusions The predictive equations developed show potential for the diagnosis of dry eye. The SP-CTT and the PRT Test were unable to discriminate between normal and dry eyes. Results from these tests are not related to tear volume. TMH measurement using a photo-slit lamp in conjunction with the IMAGEnet System was not sufficiently repeatable, however, it is a potentially useful test for the diagnosis of dry eye as it is the only test that we used which can discriminate between normal and dry eye. Based on the scoring system, the Mc-DEQC and SD-DEQC are poor screening tools for Chinese dry eye, however discriminant analysis may be a better method than the scoring system to analyse their results. Neither the Mc-DEQC nor the SD-DEQC could effectively discriminate between the normal and dry eye subjects. Clinically, in the absence of better tests, we therefore recommend use of the fluorescein inferior corneal staining and the presence of ocular symptoms. The tests described here are in common use around the world and used universally for dry eye assessment. Our work indicates that the value of these tests is extremely limited in a Chinese population.