Classification of Keratoconjunctivitis Sicca
The condition of keratoconjunctivitis sicca (KCS) is synonymous with that of dry eye. According to the International Dry Eye Workshop (DEWS) the dry eye is a multifactorial disease of the tears and ocular surface that results in symptoms of discomfort, visual disturbance and tear film instability with potential damage to the ocular surface. It is accompanied by increased osmolarity of the tear film and inflammation of the ocular surface. It comprises two subgroups: tear or aqueous-deficient dry eye (aqueous tear deficiency) is due to a failure of lacrimal function while evaporative dry eye is due, predominantly but not entirely, to lipid tear deficiency.

Either form may cause damage to the interpalpebral ocular surface and is associated with symptoms of ocular discomfort. A unifying mechanism is tear hyperosmolarity, which can directly cause damage to surface epithelial cells. However, an additional factor is the release of pro inflammatory cytokines in the lacrimal gland and tears, which can both initiate autoimmune lacrimal damage, as in Sjógren's syndrome, or perpetuate chronic conjunctival inflammation.

The commonest form of teardeficient KCS is non Sjógren's dry eye, whose age related form has a prevalence of about 15% in the older population. It is due to a T-cell infiltration of the lacrimal gland which reduces secretory function. Non-Sjógren's dry eye can be caused by other lacrimal diseases such as graft-versus host disease and sarcoidosis, by lacrimal obstruction in cicatricial conjunctival diseases (e.g. ocular pemphigoid, Stevens Johnson syndrome and trachoma) and also by reduced sensation at the ocular surface, leading to a loss of afferent reflex drive to the lacrimal gland. Sjógren's syndrome is a less common disorder, with a prevalence of about 0.2-0.5%.

It is an autoimmune exocrinopathy giving rise to dry eye and dry mouth and affects other mucous membranes and even the central nervous system. Primary Sjógren's syndrome occurs in the absence of a defined connective tissue disorder, whereas secondary Sjógren's syndrome is accompanied by such a condition, such as rheumatoid arthritis, systemic lupus or Wegener's disease. In general, the onset of Sjógren's syndrome dry eye is earlier than that of non Sjógren's dry eye and it evolves with greater severity.

The commonest form of evaporative dry eye is due to meibomian gland obstruction and this in turn has a strong association with skin disorders such as acne rosacea, atopic dermatitis (affecting the face) and seborrhoeic dermatitis. Evaporative dry eye can also result from lidglobe malposition (e.g. proptosis), contac-lens wear and occupational and environmental stresses. Thus it may be associated with low humidity due to air conditioning, with a reduction in blink rate while performing microscopy or with increased width of the palpebral aperture which occurs when working at a video display terminal. Such events may contribute to the office eye syndrome.

Diagnosis of  Keratoconjunctivitis Sicca
The above definition of the dry eye accentuates the following features of the disease: (1) symptoms; (2) interpalpebral surface damage; (3) tear instability, and (4) tear hyperosmolarity. There are numerous tests for the diagnosis of dry eye and they vary with respect to their invasiveness. The selection and order of these tests is of paramount importance since each test may influence the outcome of the test which follows. In general it is recommended to start with the least invasive procedure and to end with the most invasive test. Occasionally it is necessary to perform some tests on a subsequent day. At the end of a battery of tests it should be possible to confirm the diagnosis, classify the form of dry eye, being conscious of its grade, and initiate appropriate therapy.

Symptoms and History
A record of clinical history and ocular symptoms is required. Several questionnaires have been developed for the assessment of dry eye. A special questionnaire for the detection of psychosomatic alterations exists and can be applied additionally.

Important aspects of the patient's history are: (a) symptoms: burning sensation, foreign body sensation, tired eye, photophobia, epiphora, swelling of the lids; (b) onset of the symptoms, duration; (c) circadian rhythm; (d) environmental conditions at home and in the office (smoke, wind, humidity); (e) contact lens associated problems; (f) cosmetics; (g) systemic diseases; (h) allergic diseases; (i) dermatologic diseases, and (j) drug history.

Examination of the Lids
The dynamics of blinking and of lid position should be observed whilst taking the history in order to prevent conscious alterations. Points of interest are: (a) frequency of blinking; (b) variation of blink intervals; (c) size of the palpebral aperture, and (d) adequacy of lid closure.

The position of the lids may influence the tear turnover, therefore care should be taken to identify the following malpositions: (a) entropion; (b) ectropion; (c) eversión of the lacrimal puncta; (d) cicatrical malposition; (e) dermatochalasis, and (f) swelling of the temporal aspect of the upper lid, implying enlargement of the lacrimal gland.

Slit-Lamp Examination
Slit-lamp biomicroscopy should evaluate the following anatomical structures and their alterations: (a) Lid margins: hyperaemia, telangiectasia, thickening, scarring, keratinization, ulceration, tear debris, abnormalities of the meibomian orifices, metaplasia, character of expressed meibomian secretions, (b) Eyelashes: misdirection, malposition, encrustations, collarettes, (c) Conjunctiva: erythema, swelling, keratinization, papillary/follicular reaction, pinguecula, lid parallel con junctival folds, (d) Cornea: infiltrates, scars, punctuate staining or ulcers, vascu-larization, pannus, and pterygium. (e) Additionally, the tear film should be analysed for: filaments, mucus, and cellular debris, meibomian foam.

Non-Invasive Break-Up Time
The non-invasive break-up time test was created to measure the stability of the precorneal tear film without any dye. It involves projection of a target onto the convex mirror surface of the tear film and recording the time taken for the image to break up after a blink. The test was originally performed with a custom-built 'Toposcope' but has also been performed over a limited zone of the exposed precorneal film, using a keratometer. It can also be measured with the TearscopePlus and is a non-invasive procedure.

Interferometry
Tear film interferometry is a non-invasive technique for grading the behaviour of the tear film lipid layer and estimating its thickness on the basis of the observed interference colours. It is useful for selecting dry eye candidates for punctal occlusion. Apparatus which have been used for this purpose include the TearscopePlus and the Kowa DR-1.

A colour scale which has been used is as follows: (a) greyish colours, uniform: normal; (b) greyish colours, non-uniform: normal; (c) yellow colours: dry eye; (d) brown colours: dry eye, and (e) blue colours: dry eye.

Reflective Meniscometry
Reflective meniscometry is a non-invasive method to measure the radius of the tear meniscus curvature. The radius is directly proportional to the tear meniscus volume and to the total tear volume of the tear sac. A radius <0.25 mm indicates a hyposecretory dry eye. Fluorescein Tests Fluorescein sodium is used for several dry eye tests. They all are mildly invasive tests. At a concentration around 0.1%, the dye is highly fluorescent, staining the tear film and epithelial defects. Once the surface layer of epithelial cells is lost, the dye spreads rapidly in the intercellular space.

Fluorescein is available in the form of fluorescein-impregnated paper strips or as a 1-2% solution in a sterile, unit dose sachet. Fluorescence is with the use of a blue exciter filter in combination with a yellow barrier filter. Most slit-lamps are provided with an adequate blue light source and it is well worth purchasing a suitable Kodak Wratten 12 or 15 barrier filter. To instil fluorescein from an impregnated strip, a drop of sterile saline is applied to the impregnated end and the excess discarded with a rapid flick. The moistened tip is then touched lightly onto the lower tarsal plate of the right and then the left eye, in sequence. Because 1-2% fluorescein is non-fluorescent, it is only appropriate to apply a small volume in order to achieve dilution and fluorescence. A suitable volume is 2-5 |xl applied with a micropipette.

The fluorescein tear film break-up time records the rupture of the tear film after a blink. The tear film should be evaluated after a few blinks. The average of three measurements provides a representative measure of the tear film stability. Evaluation: >10 s: normal; 5-10 s: marginal dry eye; <5 s: dry eye. Fluorescein staining of the interpalpebral surface of the eye has a characteristic pattern in KCS, initially affecting the lower part of the exposed eye and later affecting the cornea and conjunctiva more extensively. In meibomian gland dysfunction the staining pattern is often disposed over the lower cornea, closer to the lower lid margin. A number of suitable grading schemes exist. The Oxford grading scheme consists of a series of panels representing the cornea and the two zones of exposed conjunctiva, on which is displayed a pattern of dots representing increasing staining from grade 0 to 5. The number of dots increases sequentially in a log-linear scale: from grade 0 to 1 there is a 1-log step, which means that 10 dye spots are detected per 1 zone in grade 1. Between grade 1 and 5 there is a 0.5-log unit increase of spots, which equals 32 dye spots in grade 2, 100 dye spots in grade 3 and 316 dye spots in grade 4, always counted per 1 zone. Grade 5 is detected, when the number of dye spots exceeds 316 per 1 zone.

The individual scores for each of the 3 panels are added up to give the total score. The maximum staining score for the exposed conjunctiva and cornea is 15. An important point to note is that if the recommended filter combination is used then grading with fluorescein can be carried out on both the cornea and conjunctiva and use rose bengal can be avoided. This prevents patient discomfort, since in the absence of an anaesthetic, rose bengal causes intense stinging on instillation. Staining of the epithelium can occasionally be obscured by the fluorescence of the tear film. Asking the patient to blink several times allows the staining pattern to be viewed more clearly. The stained meniscus can be used to estimate the meniscus volume, either simply by measuring meniscus height using the width of the slit-lamp beam, or in a more sophisticated fashion, by reflective meniscometry or by assessing its profile photographically in slit section. A meniscus radius of curvatures <0.25 mm suggests a dry eye condition. Fluorescein can also be used for measurements of tear turnover and of the tear fluorescein clearance.

Meibometry Meibometry is a mildly invasive quantitative method for measurement of the basal level of meibomian lipid on the lid margin. In this test, lipid is blotted onto a loop of plastic tape, which produces a strip of increased transparency. The change in transparency is quantified photometrically and provides an index of the uptake of lipid. The system can be calibrated to provide approximate estimate of the amount of lipid on the lid margin, without giving information about chemical composition. An appropriate photometer is available at Courage & Khazaka Electronic GmbH (Cologne, Germany), together with the plastic tape for testing. Schirmer Test The Schirmer I test is one of the oldest tests available for dry eye diagnosis and is a measure of reflex tear secretion. It is performed in the unanaesthetised eye. It is highly invasive, and is therefore performed later in the diagnostic sequence. A standard filter paper is placed with its notched tip bent around the lower lid margin at the junction of the middle and outer third. With the eyes closed, the Schirmer paper is allowed to wet for a period of 5 minutes, after which the length of wetting is measured from the notch to the leading, wetted edge. Evaluation: <6 mm: in the dry eye range; 6-10 mm: dry eye suspect; > 10mm: normal.

The Schirmer test can also be performed after instillation of a topical anaesthetic when it has been said to represent a 'basal' measurement, since sensory reflex stimulation from the eye is suppressed. This is called the Jones test. Although the test value is usually lower than that recorded by the Schirmer I test, the test has not been adequately validated. If the Jones test is performed after nasal stimulation, it is named Schirmer II test. In Sjógren's syndrome in contrast to non-Sjógren's dry eye, it has been shown that the ability of nasal stimulation to increase the tear production of the anaesthetised eye is greatly reduced and is of diagnostic value.

Rose Bengal Staining
The rose bengal test is markedly invasive. Rose bengal is not a vital dye. It stains damaged cells which possess an abnormal mucin coat. It is intrinsically toxic and therefore causes marked stinging on instillation. If used in drop form, its use should be preceded by instillation a topical anaesthetic.
Rose bengal is available in drop form (1%) (Minims Rose Bengal, Chauvin) or as a dye-impregnated paper strip. Staining is a dose-dependent staining effect so that when the paper strip is used, and less dye is delivered, a weaker staining pattern is achieved.

Grading of staining using rose bengal uses the same approach as for staining and grading using fluorescein. The Oxford scale has been described above. The classic, van Bijsterveld schema is also based on an estimate of staining on the cornea and the nasal and temporal part of the exposed bulbar conjunctiva. Each zone is graded from 0 to 3, and the maximum total score is 9. A score >3 is regarded as indicative of dry eye according to the van Bijsterveld schema. It should be noted that the visibility of rose bengal staining is greatest over the white of the bulbar conjunctiva. For grading purposes visibility on the cornea is reasonable when the background is a blue iris, but is poor against a dark brown iris.

Lissamine Green
Lissamine green (Lissaver Plus, Contopharma, Interlaken, Switzerland) stains the eye in a similar way as rose bengal, but it is less toxic and is consequently well tolerated. It is therefore recommended as an alternate test to the rose bengal test. However, the contrast of the dye is less sharp and the detection of the stained areas more difficult. Lissamine green is used in a 1.0% concentration. The test is mildly invasive, similar to fluorescein. The Oxford grading system and the van Bijsterveld system can be applied as described above.

Meibography
Meibography involves the transillumination of the meibomian glands after eversión of the upper and lower lid. The glands are visible in silhouette and the absence of glands or 'drop-out' can be quantified. The manipulation of the lids stimulates reflex tearing and the test should be regarded as mildly invasive.

Ocular Fermng Test
If a tear sample taken from the lower tear meniscus is applied to a glass slide, a characteristic ferning pattern develops as the tears evaporate. This pattern can be viewed under the microscope at a 40-100X magnification and used as an index of dry eye. In dry eye states, the delicate fronded pattern becomes broken up and irregular and the appearances can be graded. Fern formation is influenced by the protein and electrolyte composition of the tears. Since only a small sample of tears is required for the test it is only mildly invasive. Grading is based on the regularity of arborization of the ferning pattern. Classes 1 and 2 are regarded as normal, and classes 3 and 4 represent increasing degrees of dry eye.

Impression Cytology
Impression cytology is a histological method of cytological examination without the disadvantages of an invasive excisional biopsy. The samples can be examined by light microscopical, electron-microscopical, immunological and molecular biological methods. Several methods of interpretation of the results exist. Important features of dry eye include squamous metaplasia, loss of goblet cells and accumulation of inflammatory cells. Ultrastructural signs include changes in the nuclear/cytoplasmic ratio and an increased frequency of 'snake-like' chromatin.

Osmolarity
The measurement of tear osmolarity (mosm/1) is regarded as a gold standard in the diagnosis of dry eye, however it is difficult to measure and no commercial instrument is currently available. Studies using the depression-of-freezing-point osmometer have suggested that a value of >312 mosm/1 is diagnostic of dry eye. Newer techniques for routine clinical use are in development.

General Information and Recommendations
Diagnostic tests of dry eye states are in general intended for two groups of patients: those who come for the first visit with a suspected dry eye and those who have already undergone therapy elsewhere and wish further advice. In the first group of patients the test series for dry eye can be started immediately. In the second group of patients it is recommended to stop the patients' preexisting therapy for 1 week and start the testing afterwards. This opens the pathway towards an exact diagnosis without any therapeutic interference.

For the routine diagnostic way of a suspected dry eye state it is advised to select a number of tests. Non-invasive procedures like questionnaires, symptoms and history and slit-lamp examination are among these and often are called low-tech diagnostic. Considering patients' comfort and economic aspects, simple tests should always be used. These tests provide already a reliable information of the dry eye condition. A questionnaire has a surprisingly high sensitivity of 77% with a specificity of 81%. In combination with data from other non-invasive methods, sensitivity and specificity can even be raised. Slit-lamp characteristics like an irregularity of the black line or hyper-aemia of the conjunctiva result in a sensitivity of 92% and a specificity of 81%. Simple low-tech diagnostic is therefore the basis of the dry eye testing.

Besides that, a battery of dry eye tests exist which are mildly or markedly invasive. The order of tests is of critical importance since one test may influence the result of the next. Therefore it is recommended to start with the least invasive test and to end with the most invasive procedure. Some tests are mutually exclusive, which means that in a certain patient only a selection of dry eye tests is performed.

Within this system of tests  with increasing invasiveness, intervals of 5 min are recommended between invasive tests. This is the time necessary for restoration of the original meniscus height. For the routine dry eye patient a sequence of tests giving the essential information for the classification should be selected.

After having selected the appropriate combination of tests, the grading and interpretation of these tests gains importance. It is essential to know the information we can get from a certain test in order to classify our patient's dry eye form as tear-deficient or evaporative. If you suspect a hyperevaporative dry eye the non-invasive BUT and meibography should be measured. Meniscometry and Schirmer I test are specifically indicative for the hypovolemic dry eye. The same tests give us sufficient information about the severity of the ocular surface disease. Staining with Lissamine and fluorescein allow a more precise quantification of the severity and are therefore recommended. The interpretation is based on a grading system of the tests mentioned above. This system allows us to distinguish normal from marginal dry eyes or manifest dry eye patients.

Once the diagnosis is confirmed and the grade of the disease established the patients need follow-up examinations. The course of the KCS under therapy is documented. We get the best information from a repetition of the tests which have been selected initially. This pathway of examination provides reliable information about the course of the disease and forms the basis for longitudinal observations. However, we have to take into account that certain tests like fluorescein and rose bengal staining show limits with respect to their reliability at different times. Nevertheless, these tests are necessary to provide exact information about the localization of ocular surface defects, whereas the non-invasive tests give us broader information about the whole ocular surface. The diagnosis of dry eye is therefore based on the data of different tests with increasing invasiveness arranged in a way to minimize interference between the tests and on the grading of the results which permit a selection of the appropriate therapy and a long-term observation of the patient.