Otto-Paul van Bijsterveld

"The dry eye is the most common symptom in external diseases of the eye,
but usually no the most theatening"

(J. Murube, modified by O.P. van Bijsterveld)

The study of some pathological conditions of the eye is facilitated by the use of a vital stain. This is done by instilling the dye in the optimal concentration in the lower cul-de-sac. The stained cells or tissue can than be studied by the biomicroscope. Quite a number of stains can be used for the study of the cornea and the conjunctiva, such as: fluorescein, methylene blue, Victoria yellow, sudan red, rose bengal and alcian blue to name a few. The dyes that are often used in the study of the outer eye in keratoconjunctivitis sicca are rose bengal and fluorescein.

As early as 1933 Sj÷gren used the rose bengal stain for the diagnosis of keratoconjunctivitis sicca (KCS). Rose Bengal is a fluorescein derivative, i.e.: tetrachloro tetraiodo fluorescein sodium. Despite their chemical similarity, the two dyes differ completely with regard to their staining properties. Fluorescein does not stain cells or tissues but diffuses into the intercellular spaces from a surface defect. Rose bengal on the other hand has the characteristic to stain surface cells of the cornea and conjunctiva that tend to keratinisation.

In KCS, staining with rose bengal results in punctate staining of the conjunctiva and the cornea particularly in the area of eye that is not covered by the eyelids (Fig. 26-1). Most clinicians use a concentration of 1% rose bengal, but Holm 1949 used a concentration of 2-4% which is excessively painful and does not seem to have any clinical advantage. The clinical procedure for performing the rose bengal test is to instil one drop of a 1 % solution in the lower cul-de-sac and after several minutes study the stained structures.

f26-1.jpg (67801 bytes)
Figura 26-1. Rose bengal has the characteristic to stain surface cells of the cornea and conjunctiva that tend to keratinisation.


Irrigation of the excess dye may be necessary as the conjunctival sac hold more than one fifth of a normal drop (Norn). Instilling rose bengal may be painful even in a 1% solution. The discomfort is roughly proportional to the area of stained epithelium and it may be necessary to use a local anaesthetic in suspected serious cases of keratoconjunctivitis which in our opinion does not interfere with the results of the test.

Van Bijsterveld (1969) introduced numerical scoring for the intensity of staining with rose bengal of both medial and lateral bulbar conjunctiva and of the cornea. Each section was given a score of up to three points, so that a maximum score of nine could be obtained. He found the best staining intensity score limit to differentiate between normal persons and patients with keratoconjunctivitis sicca to be 3.5. In his comparative study, van Bijsterveld found that, with a score limit of 3.5, the combination of the results of the Schirmer┤s test and the rose bengal test was not better than the results of the rose bengal test alone. At the stated limit, he found the probability of misclassification to be around 5%.