We appreciate the interest in our article and the remarks made by Drs Pekel. Indeed, adding more objective tests can always improve a study. In our clinical cohort we chose to include a combination of the most commonly used (and, in our opinion, most validated and practical) objective tests in clinical practice. Since all 8 test results pointed in the same direction, we believe it is unlikely that the conclusion of our article would significantly alter when even more tests were included. Nonetheless, we definitely encourage further research with other and new objective dry eye tests, particularly in finding the missing link between signs and symptoms in dry eye.
We are in agreement with Drs Pekel that dry eye tests in a clinical patient cohort can be confounded by many factors. However, we do not believe there is a simple statistical way to correct for these factors. A similar problem arises when taking into account the multifactorial basis of dry eye disease. In addition to primary and secondary Sjögren disease and diabetes that Drs Pekel describe, there are tens of other causes and risk factors that might bias results on dry eye signs and symptoms. We did examine several subgroups and exclusion scenarios before submitting our article, including looking at primary and secondary Sjögren disease separately. Interestingly, our results and conclusions were consistent in all strata that were big enough to allow statistical inference. In the end, we chose not to make small subgroup analyses or to exclude patients with concomitant dry eye risk factors. The reason for this is that it would inevitably lead to sample sizes of insufficient statistical power when taking into account all the etiologies of dry eye disease, and dilute the message that patients in dry eye clinics with chronic pain syndromes experience more symptoms than those without.