Statistics from Altmetric.com
Is that really the question or do eye symptoms lie?
For those of us who see patients on a regular basis, how many times a day do we hear the recurring complaint, “Doctor, my eyes feel so dry…they itch, they burn, they feel constantly irritated”? As doctors we must ask, with a sense of wonder, do all of these patients really have dry eyes? Are we currently in the midst of a global epidemic of ocular dryness that has caught us unawares and unprepared? Our patients are telling us something when they describe their symptoms, but in the process are they also unintentionally misleading us? Are their eyes truly dry, and if they are not, is it possible for us to tell the difference?
Lee et al in the December issue of the BJO have done a superb job of documenting the prevalence of dry eye symptoms in a village population in Sumatra, Indonesia.1 Their statistical approach is impeccable. Using a one stage cluster sampling procedure, they randomly selected 100 households in each of five rural villages and one provincial town in Riau Province over a 3 month period in 2001. Demographic, life style, and medical data were collected from 1058 participants and dry eye symptoms were assessed using a six item questionnaire. The questions were, essentially, do your eyes ever feel dry, do you experience grittiness, burning, or redness, do you have crusting, and are your eyes ever stuck shut? The frequency of these symptoms was graded for each individual as rarely, sometimes, often, or all the time. One or more of these symptoms occurring often or all the time was felt to be significant, and those subjects who responded positively to these questions were placed in the dry eye symptom group and included in the analysis. The presence or absence of pterygium in each individual was also documented. Lee et al then submitted their data to detailed statistical analysis and arrived at the conclusion that 27.5% of all the individuals questioned experienced one or more of the six ocular symptoms often or all the time, and that the only factors found to be associated with these symptoms were a history of current smoking and the presence of a pterygium. Interestingly, the prevalence for dry eye symptoms was 1.4 times greater in men than in women.
Problems arise, however, when dry eye symptoms become by inference dry eye syndromes. As stated in their abstract the aim of the study was “To determine the prevalence and identify associated risk factors for dry eye syndrome in a population in Sumatra, Indonesia.” Dry eye syndromes by definition encompass a constellation of diverse disease processes that produce objective signs of keratoconjunctivitis sicca (KCS) with or without a concomitant decrease in tear production.2 The classic prototype of the dry eye syndrome is Sjögren’s syndrome, but there are many other causes of KCS including cicatrising conjunctival diseases such as trachoma and pemphigoid, non-cicatrising syndromes causing specific dry eye findings, and atypical syndromes such as keratomalacia in which the eye is symptomatically and objectively dry but tear production is paradoxically normal.2 The diagnosis of dry eye syndrome or KCS implies that the patient has a specific disease entity as an underlying cause and that the condition is potentially treatable.
The dry eye in most cases is truly a diagnostic dilemma until objective tests are performed
Lee et al have documented the prevalence of dry eye symptoms in this village population in Indonesia, but they have not provided us with objective proof that any of the subjects in the study actually had signs of KCS. The authors remarked on this shortcoming in their discussion and stated that local sociocultural sensitivities precluded an interventional study that would allow objective dry eye tests. This is unfortunate because we still do not know the true prevalence of KCS in this interesting group of patients. With minimal intervention this question could have been answered by performing Schirmer’s tests without anaesthesia, fluorescein staining, tear break up time, and rose bengal staining on all the subjects. Using these objective tests the true prevalence of dry eyes in this population could have been easily determined.
Why is it necessary to perform objective tests? Don’t the patients’ symptoms give us a good indication of whether or not they really have dry eyes? Lee et al have provided the answer in Table 3 of their article. The relation between symptoms of dryness and objective findings of dryness is like comparing apples and oranges. Ask the same individuals whether or not their eyes feel dry and then perform objective tests, and you find as much as a sevenfold difference in the outcome (14.6% with symptoms, 2.0% with rose bengal staining).3 In some populations that were questioned about their dry eye symptoms, the prevalence rate was as high as 28.7% (Table 3).4 How can there be such a disparity between the symptoms of dryness and the objective findings of dryness? The simple answer is that the eye is not very smart. When it is stressed it has a very limited range of symptomatic responses—pain, foreign body sensation, burning, itching, dryness. Any of these symptoms can be caused by any number of ocular conditions from infections to allergic reactions to something as simple as conjunctival concretions. It is important to always invoke the innocent until proved guilty principle when suspecting a dry eye. Until ocular dryness can be proved by objective tests, mere testimony (symptoms) can only be used to raise our suspicion that it may exist.
The dry eye in most cases is truly a diagnostic dilemma until objective tests are performed. For years we asked patients in our dry eye clinic a standard series of questions. Those patients both with and without an objective diagnosis of ocular dryness gave positive answers to questions regarding foreign body sensation, burning, redness, inability to tear, itching, etc. It wasn’t until we retrospectively looked at 304 patients with objectively proved KCS and compared them with 195 patients with ocular complaints but who did not have proved KCS that we realised it was the specificity of the questions that was important, not the sensitivity. So what was the most important question to ask a patient who may have a dry eye? “Can you cry?” Patients who were unable to produce tears even under stressful conditions were very likely to have a true dry eye syndrome compared to those who were symptomatic but were not actually dry (p value <0.00001).5
Questions about ocular symptoms of dryness are definitely important, but only as corroborative evidence to point the way to objective tests. Is the eye guilty of being dry? It must remain innocent until proved otherwise beyond the shadow of a doubt. That is the dilemma. Is the eye too dry or not too dry? The answer to this question only serves to guide us as doctors to the objective tests with which we can ultimately determine the true diagnosis.
Note in Proof
Is that really the question or do eye symptoms lie?