Materials and Methods

Birds—Twenty-four Hispaniolan parrots (Amazona ventralis) from a research colony that was housed indoors in a climate-controlled facility were used in the study. The colony was housed at 24°C and was exposed to a constant 24-hour lighting cycle that included 12 hours of light (photopic) conditions and 12 hours of dark (scotopic) conditions. All birds were hand raised from captive stock and were approximately 11 years old. The colony had been established for 10 years, and the birds were used periodically for nutritional studies. The birds were acclimated to restraint as a result of intermittent handling that occurred during previous nutritional studies and that was required for husbandry. At the time of the study, the colony was not involved in concurrent investigations and all birds were fed a commercial avian diet. During the study, 1 bird died as a result of nonocular disease, and recorded values for that bird were included in the data analyses. The investigation was approved by the Institutional Animal Care and Use Committee of the School of Veterinary Medicine, Louisiana State University.

Study procedures—All birds received a thorough ophthalmic examination weekly for 4 weeks to evaluate the repeatability of tear test values in individual birds and to improve the statistical power of the study with fewer experimental animals by use of a repeated measures statistical procedure. All examinations were carried out in the birds' housing area, and measurements were performed under handheld restraint by the same investigators (ESS and DAC) throughout the study to minimize variations in restraint or experimental technique. Birds were excluded from the study if signs of conjunctivitis, keratitis, blepharitis, or intraocular disease (including uveitis, glaucoma, and extensive cataracts) were evident at any of the weekly ophthalmic examinations.

Each ophthalmic examination included various procedures that were performed for both eyes of each bird in a defined sequence as follows: assessment of the palpebral, oculocephalic, and direct pupillary light reflexes; performance of corneal aesthesiometry, a PRTT without topical application of proparacaine hydrochloride (PRTT I), and an STT I without topical application of proparacaine (STT I); topical ocular application of proparacaine ophthalmic solution^b (1 drop/eye was applied once); and performance of another PRTT (ie, with proparacaine [PRTT II]), another STT (ie, with proparacaine [STT II]), applanation tonometry,^c slit-lamp biomicroscopy, and indirect ophthalmoscopy (by use of 20- and 90-diopter lenses). An interval of 10 minutes was allowed to elapse after application of the ocular anesthetic agent prior to performing a PRTT II or STT II. At the end of the examination, fluorescein stain^d was applied to the cornea of each eye to detect corneal ulcers. For purposes of the study, data that were analyzed included PRTT I and II values, STT I and II values, and corneal touch threshold measurements.

Corneal touch threshold measurements were recorded by use of a Cochet-Bonnet corneal aesthesiometer^e to elicit a blink response. To determine the touch threshold, the monofilament was extended and touched onto the central portion of the cornea of each eye. The length of the exposed monofilament was successively shortened and the test repeated until the bird blinked in response to the monofilament touching the central portion of the cornea on 2 of 3 or 3 of 5 attempts. A minimum of 3 attempts to elicit a blink response was used at each filament length. The birds were gently restrained by hand during the procedure. The length of the exposed monofilament was measured with Jameson calipers when it was < 5 mm.

Schirmer tear test strips^f from only 1 manufacturer and of 1 lot number were used for all birds to prevent inconsistencies in the absorptive capacity of tear test strips.¹⁹ The strips were folded at the notch while still within the packaging and placed in the ventral fornix of the eyelid to the level of the notch (Figure 1). After 1 minute, the extent of moisture was measured by use of the scale printed on the strip.

View Full Size

Photograph of a Schirmer tear test strip placed (to the level of the notch) in the ventral fornix of the inferior eyelid in an Amazon parrot. The transparent third eyelid is partially covering the corneal surface of the right eye. Notice that the amount of tear film is indicated by the level of the blue dye on the test strip scale. The width of the palpebral fissure in this species permits the use of a full-sized commercial test strip.

Citation: Journal of the American Veterinary Medical Association 235, 10; 10.2460/javma.235.10.1181

• Download Figure
• Download figure as PowerPoint slide

To perform the PRTT, the cotton thread^g was removed from the packaging with cilia forceps and grasped just behind the bend 3 mm from the proximal end of the thread (Figure 2). The lower eyelid was manually retracted from the globe, allowing the thread to be placed with the notch at the lower eyelid margin (Figure 3). No attempt was made to hold the eyelids closed. If the thread fell from the eye before an interval of 15 seconds elapsed, the test was repeated. The entire length of the straightened moist thread (regardless of color) was immediately measured with Jameson calipers or with the ruler that was included on the packaging.

View Full Size

Photograph of a phenol red thread that has been removed from its packaging with cilia forceps. The thread is grasped just behind the preplaced bend in the thread. The 3-mm-long bent portion of the thread is placed at the inferior eyelid margin and contributes to the measured length of moistened thread after 15 seconds.

Citation: Journal of the American Veterinary Medical Association 235, 10; 10.2460/javma.235.10.1181

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Photograph of the PRTT in progress in the right eye of an Amazon parrot. Notice the change in color from yellow to red in the portion of thread that protrudes from the inferior eyelid, which is indicative of wetting of the thread with tear film.

Citation: Journal of the American Veterinary Medical Association 235, 10; 10.2460/javma.235.10.1181

• Download Figure
• Download figure as PowerPoint slide

After topical ocular application of proparacaine, no attempt was made to absorb the residual tears prior to repeating the PRTT or the STT but an interval of 10 minutes was allowed to elapse. The recorded values were considered representative of the residual tear volume as well as the reflex and basal tear film volumes.

Statistical analysis—All response variable data were initially included in both Pearson and Spearman correlation analyses. Initial analyses also included paired t test comparisons of block variables and a Shapiro-Wilk test of normality. Mean values reported were calculated from all observations of individual eyes (n = 48) at each of the 4 time points for which data were available (1 bird died following the third examination time point). A general linear models procedure was used to analyze the data in an ANOVA of a split-plot arrangement in a repeated-measures design. In the model, group and bird X group were included as main plot effects; in the subplot, main effects were day and eye and the 2- and 3-way interactions of day, eye, and group were included. When overall F tests indicated significance, post hoc pairwise comparisons of main effects were conducted with a Tukey honestly significant difference test. Pairwise comparisons of significant interaction effects were made by use of t tests of least square means. All comparisons were considered significant at a value of P ≤ 0.05. All statistical analyses were performed by use of commercially available software.^h

Results

For the 24 Hispaniolan Amazon parrots, the mean ± SD PRTT I value was 12.5 ± 5 mm/15 s (range, 1 to 25 mm/15 s). A paired t test revealed that there was no significant (P = 0.761) difference between the PRTT I and II values (ie, mean values determined before and after topical ocular application of proparacaine were 12.5 mm/15 s and 12.6 mm/15 s, respectively). The mean STT I value was 7.9 ± 2.6 mm/min (range, 0 to 13 mm/min). The mean STT II value was 5.1 ± 3.3 mm/min (range, 0 to 18 mm/min). The STT I and II values differed significantly, as determined by results of a paired t test (P < 0.001) and ANOVA (P < 0.001) and by the calculated Pearson correlation coefficient (0.609) and Spearman correlation coefficient (0.612). There was weak correlation between the PRTT and STT values (Table 1).

Corneal touch threshold was associated consistently with a filament length < 5 mm (> 15.9 g/mm²) in the Amazon parrots in the present study, thereby making reliable measurement of corneal sensitivity difficult. A mean filament length of 1.33 mm was required to induce a repeatable blink response. No pressure values (g/mm²) or formulas to calculate such values for filament lengths < 5 mm are provided by the manufacturer of the Cochet-Bonnet corneal aesthesiometer used in the study. Manufacturer's instructions for the aesthesiometer provide pressure values corresponding to 0.12-mm-diameter nylon filament lengths in 5-mm increments. By use of those values and curve-fitting software,ⁱ the relationship between the filament length (x) and the pressure (in g/mm² [y]) was best described by the formula as follows: y = 25.66048 − 2.2677677·x + 0.077787855·x² − 0.001177658·x³ + 6.5705128e − 006·x⁴. The SE of this formula when used to fit the curve created by the manufacturer's points was 0.265515. By use of this formula, the mean filament length of 1.33 mm corresponded to a pressure of 21.4 to 23.4 g/mm². There was no correlation between the corneal touch threshold and any of the tear film measurements at any time.

Signs of conjunctivitis, keratitis, blepharitis, or intraocular disease (including uveitis, glaucoma, and extensive cataracts) were not detected in any bird during any of the weekly ophthalmic examinations. Corneal ulcerations secondary to aesthesiometry or other procedures of the examination were not evident at any time.

Discussion

Often in human medicine and sometimes in veterinary medicine, attempts are made to differentiate and quantify the residual, basal, and reflexive tear film volumes to better understand the dynamics of tear film on the ocular surface, the role and value of individual tear test techniques, and how to best treat eyes with tear film inadequacies or ocular exposure. The STT I is purported to evaluate those 3 tear film volumes.²³ The STT II theoretically measures basal tear production and a minimal amount of residual tear film when the conjunctival fornix is gently dried prior to the test.²³ However, some investigators question whether there is any basal tear production rate aside from that induced by corneal reflexes because stimulation of an anesthetized eye continues to induce increased lacrimation.^32,33 The PRTT is not reported to be influenced by topical anesthesia of the eyes, is completed rapidly, and generally is considered to be a measurement of the residual tear film in the conjunctival fornix. This volume is a reflection of the balance among lacrimal secretion, tear drainage, and, primarily in animals with disease, evaporation.^23,34 Efforts to confirm that the PRTT evaluates only the residual tear film have been unsuccessful, and no test can completely eliminate ongoing contributions to the residual tear film in the lacrimal lake from basal tear production.^28,34 However, the brief (15-second) duration of the PRTT minimizes the influence of the continued basal tear production.

In clinical veterinary practice, quantitative clinical evaluation of the precorneal tear film is most frequently limited to use of the STT prior to topical application of an ophthalmic anesthetic agent (ie, the STT I). This is because standard values for STT I in domestic species are published, accepted, and clinically useful for the identification of quantitative tear film deficiencies. The STT I is typically performed in animals by placing a 5-mm-wide strip of filter paper or a commercially available STT strip between the inferior eyelid and the globe at the junction of the middle and lateral thirds of the eyelid for 1 minute. The length of the wet portion of the filter paper is measured from the fold at the lid margin^28,35 or is derived from the scale applied to commercial test strips. Standard commercially available STT strips are 5 mm in width and have a notch 5 mm from the end of the strip that is placed in the eye. Narrower nonstandard widths of 2 and 4 mm have been recommended for use in small domestic, wild, and exotic animals.^28,36,37 It is widely accepted that STT I results are indicative of reflexive tear production in response to irritation induced by the strips themselves, but findings are also influenced both by residual tear volume and basal secretion.^23,26,27,34

Other quantitative methods of clinically evaluating tear films in veterinary medicine include the STT performed following topical application of an ophthalmic anesthetic (ie, the STT II) and the PRTT. To perform the STT II, a drop of topical anesthetic is applied to the eye, and after the residual tears are removed, the STT strip is placed between the inferior eyelid and the globe to the level of the notch for 1 minute. The result of this test is generally accepted as a measurement of the basal tear volume, but some investigators have identified that a reflexive tear component is associated with the STT II value.^33,38 The STT II value may also include a contribution from the residual tears despite attempts to dry the inferior fornix of the eyelids.^26,27 The STT II is not widely used clinically because the values in clinically normal animals may be so low as to make differentiation of normal and abnormal values problematic and subjective.^23,24,27,34 The PRTT measures the residual tear volume and basal tearing. The application of topical ocular anesthetics prior to testing does not significantly alter PRTT values, so reflexive tearing is not thought to contribute to the test result.^26,27 Emphasis is often placed on the contribution of residual tear film to the PRTT value.^26,27,34 Reference range PRTT values are available for dogs,^23,28 cats,²⁷ guinea pigs,²⁸ and various avian species.²⁹ Because of the PRTT's short duration, suitability for small eyelid fissures, and available reference range values, this test is considered a more reliable quantitative tear film test than the STT for use in guinea pigs.²⁸ The PRTT is still not in common clinical use, however, because of practitioners' unfamiliarity with the test and the relatively poor availability of PRTTs in practice.

When performing an STT II, residual tear film is often wicked away from the ventral fornix to limit contribution of residual tears and topical anesthetic solution to the STT II value.^23,24 In the present study, an interval of 10 minutes was allowed to elapse after application of the ocular anesthetic agent to permit restoration of the normal balance of residual tears and remaining tear flow.³⁴ Because of the small eyelid fissures and difficulty retracting the inferior eyelid in birds, no attempt was made to dry the ventral fornix to remove the contribution of residual tear film. Therefore, the STT II and PRTT II values obtained from the birds were thought to reflect residual tear volume and basal tear flow.

In the group of Amazon parrots in the present study, the mean ± SD STT value prior to application of the topical ophthalmic anesthetic was 7.9 ± 2.6 mm/min (range, 0 to 13 mm/min). Given that the data obtained in the study were normally distributed, then statistically, approximately 95% of clinically normal Amazon parrots should have a measurable STT I value in the range of 3 to 13 mm/min. The mean STT I value derived in the present study was greater than 0 mm/min by a magnitude of approximately 3 SDs, and this value may be clinically useful as a cutoff with which to identify birds with low STT I values. Amazon parrots have eyelid fissures that are sufficiently long to permit the use of commercial 5-mm-wide STT strips, which greatly simplifies the use and interpretation of the STT value.

Following topical application of ophthalmic anesthetic to the eyes of the parrots in the present study, the mean STT II value was 5.1 ± 3.3 mm/min (range, 0 to 18 mm/min). Thus, approximately 95% of clinically normal Amazon parrots should have a measurable STT II value in the range of 0 to 12 mm/min. These values illustrate the difficulty associated with this test in differentiating a population of birds with subnormal STT II values. If STT II values as low as 0 mm/min are within the reference range, test results would not be useful for differentiation of low tear production in a parrot with or without clinical evidence of keratoconjunctivitis sicca. In the present study, low STT II values (including 0 mm/min) were recorded, despite the fact that residual tear film was not removed prior to performing the STT II. If an STT is not performed prior to topical application of an ophthalmic anesthetic agent, the STT should be delayed until effects of topical anesthesia have completely worn off. The duration of the effect of various topically administered anesthetics on tear production has not been reported, to our knowledge, but the maximal duration of corneal anesthesia following multiple applications of 0.5% proparacaine is 1 to 2 hours.^39,40

In the Amazon parrots of the present study, mean PRTT value prior to topical application of ophthalmic anesthetic was 12.5 ± 5.0 mm/15 s (range, 1 to 25 mm/15 s). On the basis of statistical analysis, approximately 95% of clinically normal Amazon parrots should have a measurable PRTT I value in the range of 3 to 23 mm/15 s. Following application of the anesthetic agent, mean PRTT value was 12.6 ± 5.4 mm/15 s (range, 2 to 24 mm/15 s). Thus, approximately 95% of clinically normal Amazon parrots should have a measurable PRTT II value in the range of 2 to 23 mm/15 s. This range was not significantly different from the range of PRTT I values (determined without application of proparacaine). In Amazon parrots, the PRTT may be a clinically useful quantitative measurement of residual and basal tear film because values that can be considered normal are greater than 0 mm/15 s by a magnitude of more than 2 SDs, thereby allowing differentiation of a group of birds with subnormal values. The PRTT is also rapid, inexpensive, and requires minimal equipment. The PRTT is clinically useful before or after topical application of an ophthalmic anesthetic agent. Furthermore, it required no modification to be practical for birds with small eyelid fissures in the present study and may be a practical method of quantitative tear film evaluation in birds with yet smaller eyelid fissures.

Fluctuations in the STT data from day to day and from week to week have been detected in dogs,^41,42 and a previous study²⁹ in large Psittaciformes revealed poor reproducibility of the PRTT. In addition, variability in mean PRTT values in the birds in that study²⁹ was attributed to geographic location or seasonal and environmental influences. In the present study, the results of the STT did not vary by date and the significant variability in the PRTT data was considered clinically irrelevant (range, 11.69 to 13.85 mm/15 s). The comparatively better reproducibility in our study, compared with that of the previous study, may be a result of uniformity of environment and birds.

The results of the corneal touch threshold testing with the Cochet-Bonnet corneal aesthesiometer suggested that Amazon parrots have low corneal sensitivity, compared with corneal sensitivity in domestic animals and humans. The monofilament length that consistently induced a blink response was ≤ 5 mm for all birds on all assessment days. The mean length was 1.33 ± 1.06 mm. There was no correlation between the corneal touch threshold and any of the quantitative tear film measurements at any time. By use of a formula derived from the manufacturer's reported conversion values, this corresponded to pressure of 21.4 to 23.4 g/mm².

The ocular surface and lacrimal glands function as a unit that is particularly suited to enable and preserve vision. If basal lacrimation does not exist in the absence of basal corneal or adnexal stimulation, this has important implications for normal physiologic lacrimation. The corneas of the Amazon parrots in the present study were much less sensitive (as determined via corneal aesthesiometry), compared with findings in humans, dogs, cats, guinea pigs, and rabbits.^28,43–46 This may reflect stoicism or a high pain threshold in these birds, and as a result, corneal aesthesiometry may be poorly suited to measuring corneal sensitivity in Amazon parrots. However, it is interesting to consider that the purpose of the relatively insensitive corneal surface may be to permit vision unobstructed by excessive lacrimation during flight.

The usefulness of the PRTT in birds with small eyelid fissures and its ability to segregate birds with clinically important decreases or increases in lacrimation can only be determined with further clinical investigations. For the STT I and the PRTT I and II, the mean values of clinically normal Amazon parrots, including the respective SDs, identified in the present study were greater than 0. Few of these clinically normal birds had an STT I, PRTT I, or PRTT II value of 0 mm at any time point, making these observations statistical outliers. The ability to identify birds with STT I and PRTT values less than the mean by a factor of > 2 SDs suggested these tests are potentially useful for screening for or confirming low quantitative tear values in Amazon parrots.