December 2013
Volume 54, Issue 13
Free
Clinical and Epidemiologic Research  |   December 2013
Corneal Staining Characteristics in Limited Zones Compared With Whole Cornea Documentation for the Detection of Dry Eye Subtypes
Author Affiliations & Notes
  • Beau J. Fenner
    Duke-NUS Graduate Medical School, Singapore
  • Louis Tong
    Duke-NUS Graduate Medical School, Singapore
  • Correspondence: Louis Tong, Singapore National Eye Centre, 11 Third Hospital Avenue, Singapore 168751; louis.tong.h.t@snec.com.sg
Investigative Ophthalmology & Visual Science December 2013, Vol.54, 8013-8019. doi:https://doi.org/10.1167/iovs.13-12802
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Beau J. Fenner, Louis Tong; Corneal Staining Characteristics in Limited Zones Compared With Whole Cornea Documentation for the Detection of Dry Eye Subtypes. Invest. Ophthalmol. Vis. Sci. 2013;54(13):8013-8019. https://doi.org/10.1167/iovs.13-12802.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose.: To determine the reliability of single- and double-zone corneal fluorescent staining compared with five-zone analysis for the prediction of dry eye disease.

Methods.: Prospective study of 510 subjects with dry eye disease characterized using corneal fluorescein staining, Schirmer scores, and tear break-up times. Corneal staining was quantified using Baylor scoring with ROC analysis used to assess predictive power of single- and double-zone compared with five-zone analysis for aqueous, evaporative, and mixed dry eye disease.

Results.: Double-zone analysis predicted each subtype of dry eye disease investigated. Aqueous disease was predicted by superior/inferior zones (AUCSup/Inf 0.797 versus AUCTotal 0.816), evaporative disease by inferior/central zones (AUCInf/Cen 0.759 versus AUCTotal 0.778), and mixed disease by superior/inferior, inferior/nasal, and inferior/central zones (AUCSup/Inf 0.765, AUCInf/Nas 0.771, AUCInf/Cen 0.778 versus AUCTotal 0.795). Inferior zone analysis predicted aqueous (AUCInf 0.751 versus AUCTotal 0.750), evaporative (AUCInf 0.756 versus AUCTotal 0.752), and mixed (AUCInf 0.831 versus AUCTotal 0.788) dry eye disease with similar efficacy to complete analysis in diabetic individuals. Inferior zone analysis also predicted aqueous disease in rheumatoid arthritis patients (AUCInf 0.804 versus AUCTotal 0.785), whereas superior zone analysis predicted evaporative disease in thyroid disease patients (AUCSup 0.765 versus AUCTotal 0.752).

Conclusions.: Double-zone corneal staining predicts the presence of dry eye disease with predictive power similar to complete corneal analysis. Additionally, subtypes of dry eye can be predicted by single-zone analysis among patients with diabetes (inferior zone), rheumatoid arthritis (inferior zone), and thyroid disease (superior zone). Clinical characterization of dry eye can thus be hastened by limiting corneal examination to specific zones.

Introduction
Dry eye disease is a common medical condition with highly variable prevalence in different countries, ranging from 7% in the United States to as high as 33% in Taiwan and Japan. 13 Unsurprisingly, the economic burden of dry eye is substantial, with recent studies indicating that the direct and indirect costs of dry eye disease likely exceed US$1000 per patient per year in industrialized nations. 47 Ophthalmologists currently define dry eye as a multifactorial disease of the tear and ocular surface that produces a variety of symptoms, including discomfort, visual disturbances, and ocular surface damage. Criteria established by the Dry Eye Workshop (DEWS) 8,9 enable separation of dry eye into two subtypes: aqueous and evaporative. Systemic disorders affecting the lacrimal gland, including Sjögren syndrome, are aqueous deficiency diseases, whereas conditions such as blinking disorders and posterior blepharitis are considered evaporative, 10 although patients often have elements of both subtypes. 11  
Clinical diagnosis and evaluation of dry eye remain challenging because of the variety of clinical presentations, etiologies, and multitude of techniques used to quantify disease severity. 12 Moreover, it is widely recognized that patient reports of symptoms correlate poorly with quantitative measurements of clinical severity. Recent work by Sullivan and coworkers 11 highlighted the limited correlation between commonly used measures of dry eye severity and the consensus DEWS severity scale. 
Numerous methods have surfaced in recent years to evaluate dry eye patients, with tear film osmolality being the current gold standard for the quantitative diagnosis of dry eye disease. 1113 A recent comparative study by Sullivan et al. 11 indicated that osmolality correlates more closely with disease severity than other commonly used parameters. That said, measurement of tear film osmolality remains relatively expensive and has yet to find widespread adoption among clinicians. Currently, most ophthalmologists and optometrists rely on a small number of established methods to assess their patients. Tear break-up time (TBUT), corneal and conjunctival staining, tear film assessment, and the Schirmer test are still the most commonly used diagnostic tests for initial assessment of dry eye, in addition to symptom assessment. 12,14 Of these methods, only corneal and conjunctival staining provide an objective measure of ocular surface damage caused by dry eye disease. Consistent recording of corneal staining severity has been facilitated by the development of several grading schemes, including the Oxford scale 15 and the quantitative Baylor scale, 16 which enable reproducible quantification of corneal damage. In these systems, the cornea is divided into five zones and the frequency of fluorescent staining is enumerated in each zone to determine an overall score of corneal damage. Despite their relative technical simplicity, these methods are invariably time-consuming for routine clinical use and ophthalmologists are often reluctant to fully document these staining results in the topographical cornea zones because of the perception that such detailed information may not be worthwhile. 17,18  
In the current work, we have undertaken a prospective examination of a cohort of patients with mild to severe dry eye disease to determine whether corneal staining assays could be further simplified in an effort to hasten the time to diagnosis in a clinical setting. Using receiver operator characteristic (ROC) analysis, we compared the sensitivity and specificity of the conventional five-zone corneal staining analysis with single- or dual-zone analysis for detection of aqueous, evaporative, or mixed dry eye disease. We also investigated the association between systemic diseases known to induce dry eye and corneal staining to determine if certain corneal zones could be used to predict the presence of dry eye disease in these patient groups. 
Methods
Study Population
Patients were prospectively recruited from the dry eye clinic at Singapore National Eye Center during the period of August 2006 to October 2010. All patients enrolled in the study were first-time visitors to the clinic. The study was approved the Institutional Review Board of the Singapore Eye Research Institute, and all study procedures complied with the tenets of the Declaration of Helsinki on human research. Patients were informed that all personally identifiable information would be removed from data before publication. Patients with dry eye symptoms and at least one of the following: abnormal Schirmer I (≤10 mm at 5 minutes), TBUT (≤10 seconds), or positive corneal fluorescein staining test (Baylor >0 any corneal zone) were recruited to the study. Study population characteristics are shown in Table 1
Table 1
 
Characteristics of the Study Population
Table 1
 
Characteristics of the Study Population
Characteristic Value*
n 510
Age, y, mean ± SD 53.0 ± 14.1
Male sex, % 25.5
Race, %
 Chinese 87.8
 Other 12.2
Medical history, n (%)
 Diabetes mellitus 30 (5.9)
 Rheumatoid arthritis 25 (4.9)
 Thyroid disease 41 (8.0)
 Currently smoking 33 (6.5)
Dry eye disease subtype,† n (%)
 Aqueous 171 (33.5)
 Evaporative 180 (35.3)
 Mixed 96 (18.8)
Ocular history, n (%)
 LASIK 23 (4.5)
 Cataract surgery 41 (8.0)
 Contact lens wear 90 (17.6)
Baylor corneal staining scores, mean ± SD
 Superior 0.55 ± 1.07
 Inferior 1.66 ± 1.50
 Temporal 0.91 ± 1.38
 Nasal 1.27 ± 1.45
 Central 0.99 ± 1.40
 Total 5.38 ± 5.48
Schirmer distance, mm, mean ± SD 11.3 ± 8.57
TBUT, s, mean ± SD 3.01 ± 1.58
Ophthalmological Assessments
Assessment of dry eye symptoms was performed as described previously. 19 In addition to demographic information, patients were questioned about dry eye risk factors, including smoking within the past year, contact lens wear within the past month, previous eye surgery, history of rheumatoid arthritis, diabetes mellitus, and thyroid disease. LT performed all clinical examinations. Schirmer testing was performed without anesthesia, and TBUT and corneal fluorescein staining was performed using previously described methods. 19 Staining for each of the five corneal zones (i.e., superior, inferior, temporal, nasal, and central) was scored using the Baylor method as described elsewhere. 16 Briefly, fluorescein-stained corneal dots were scored for each zone using a five-point scale: 0 dots = 0; 1 to 5 dots = 1; 6 to 15 dots = 2; 16 to 30 dots = 3; >30 dots = 4; one area of confluent staining = add one point; two or more confluent areas = add two points; corneal filaments = add two points. 
Statistical Analysis
Statistical analysis was performed using the ROC analysis module of MedCalc version 12 (MedCalc software, Ostend, Belgium), with the method of DeLong et al. 20 software option used for AUC calculation. Data used for analyses were restricted to the right eye to ensure independence of data points. Aqueous dry eye was defined as a Schirmer test of less than 8 mm at 5 minutes and Baylor corneal staining score of greater than 1, whereas evaporative dry eye was defined as a TBUT less than 3 seconds and a Baylor score of greater than 1. Mixed dry eye was defined as Schirmer less than 8 mm, TBUT less than 3 seconds, and a Baylor score of greater than 1. These thresholds were more stringent (requiring fulfillment of at least two criteria) than those used for the initial patient recruitment mentioned above as they were used to identify dry eye subtypes. The initial study population consisted of patients referred for dry eye, and any one of the three characteristics, when present, would suffice for inclusion as long as the clinical impression was predominantly dry eye and not a differential like allergic conjunctivitis. ROC curves were plotted using MedCalc by assigning a value of 1 to patients who met the above criteria for each condition or 0 to those who did not. These values were then paired with the corresponding Baylor scores for each zone, pairs of zones (algebraic sum of 2 scores), and the sum of all five zones to plot ROC curves. Reanalysis of the population after elimination of patients with a history of smoking, contact lens wear, or Lasik or cataract surgery was performed in the same way as described above. For patients with autoimmune conditions, a value of 1 was assigned to patients who had aqueous, evaporative, or mixed dry eye in addition to at least one of the systemic diseases associated with dry eye (rheumatoid arthritis, thyroid disease, and diabetes mellitus), and 0 was assigned to patients who did not meet these criteria. 
Results
Characterization of the Study Population
A total of 510 subjects with dry eye disease were prospectively recruited to the study, with approximately one-third suffering from either aqueous deficiency or evaporative disease, and almost one-fifth having a mixed form of disease (Table 1). Of this initial number, 447 patients met the more stringent dry eye criteria used for one or more of the dry eye subtypes defined in the study. Most recruited patients were postmenopausal females, likely reflecting the prevalence of dry eye disease in this demographic. 2 Mean values for Schirmer score and TBUT were 11.3 mm (SD 8.57 mm) and 3.01 seconds (SD 1.58 second), respectively, highlighting the severity of tear instability in this cohort, whereas Baylor scoring of the five corneal zones revealed more damage to the inferior and nasal zones compared with the superior, temporal, and central zones (Table 1; Fig. 1). 
Figure 1
 
Corneal staining distribution in patients with aqueous, evaporative, and mixed dry eye disease. The mean Baylor staining scores of corneal zones are shown for patients meeting criteria for each of the dry eye types. Error bars represent the SD.
Figure 1
 
Corneal staining distribution in patients with aqueous, evaporative, and mixed dry eye disease. The mean Baylor staining scores of corneal zones are shown for patients meeting criteria for each of the dry eye types. Error bars represent the SD.
One- and Two-Zone Corneal Analysis
We initially compared ROC curves for each of the five corneal zones against the sum of all five zones using the presence or absence of aqueous, evaporative, or mixed dry eye as the disease marker. In all cases, the sum of zones was significantly better than individual zones at predicting the presence of dry eye disease, with P values less than 0.05 for all individual zones when compared with the sum of zones (Table 2). We next repeated this analysis using pairs of zones instead of single zones to compare against the sum of all zones. In the case of aqueous disease, the combination of inferior and superior corneal zones was able to detect the disease as efficiently as using information from all five zones (P = 0.130; Table 3), whereas the inferior/central zone combination predicted evaporative disease (P = 0.059; Table 3). Mixed dry eye disease was predicted by several zone pairs, including superior/inferior (P = 0.565), inferior/nasal (P = 0.088), and inferior/central (P = 0.106), with similar efficiency to five zones (Table 3). ROC curves for aqueous, evaporative, and mixed dry eye are shown in Figure 2 for each of the effective zone pairs, the sum of five zones, and an ineffective zone pair. 
Figure 2
 
ROC curve analysis indicates that fluorescein staining of two corneal zones is statistically similar to full corneal analysis to predict dry eye severity. ROC curves were generated as described in the Methods section. Curves are shown for aqueous (A), evaporative (B), and mixed (C) dry eye disease. Total, five-zone corneal scoring; S-I, superior/inferior zones; S-C, superior/central zones; S-T, superior/temporal zones; I-C, inferior/central zones; I-N, inferior/nasal zones. The marker line represents the curve expected if the diagnostic test holds no predictive value (i.e., AUC = 0.5).
Figure 2
 
ROC curve analysis indicates that fluorescein staining of two corneal zones is statistically similar to full corneal analysis to predict dry eye severity. ROC curves were generated as described in the Methods section. Curves are shown for aqueous (A), evaporative (B), and mixed (C) dry eye disease. Total, five-zone corneal scoring; S-I, superior/inferior zones; S-C, superior/central zones; S-T, superior/temporal zones; I-C, inferior/central zones; I-N, inferior/nasal zones. The marker line represents the curve expected if the diagnostic test holds no predictive value (i.e., AUC = 0.5).
Table 2
 
Fluorescein Staining of Individual Corneal Zones Does Not Predict the Severity of Dry Eye Disease As Well As Combined Corneal Zones in an Undifferentiated Population
Table 2
 
Fluorescein Staining of Individual Corneal Zones Does Not Predict the Severity of Dry Eye Disease As Well As Combined Corneal Zones in an Undifferentiated Population
Zone AUC* 95% CI* P Value* Cutoff Sensitivity, % Specificity, %
Aqueous deficiency, n = 171
 Superior 0.652 0.609–0.693 <0.0001 >0 45.6 82.9
 Inferior 0.773 0.734–0.809 0.0054 >1 77.2 67.6
 Temporal 0.714 0.673–0.753 <0.0001 >0 64.9 73.7
 Nasal 0.731 0.690–0.769 <0.0001 >1 60.2 77.9
 Central 0.685 0.643–0.725 <0.0001 >0 63.7 67.3
 Total 0.816 0.779–0.848 >4 72.5 74.0
Evaporative, n = 180
 Superior 0.595 0.551–0.638 <0.0001 >0 38.3 79.7
 Inferior 0.718 0.676–0.756 0.0001 >1 68.9 64.2
 Temporal 0.673 0.630–0.713 <0.0001 >0 62.2 73.3
 Nasal 0.681 0.638–0.721 <0.0001 >0 75.0 55.2
 Central 0.685 0.624–0.725 <0.0001 >0 65.6 69.1
 Total 0.778 0.740–0.814 >2 89.4 56.4
Mixed, n = 96
 Superior 0.624 0.581–0.666 <0.0001 >0 45.8 77.8
 Inferior 0.747 0.707–0.784 0.0108 >2 65.6 73.9
 Temporal 0.698 0.656–0.737 <0.0001 >0 69.8 67.9
 Nasal 0.724 0.683–0.763 0.0007 >1 64.6 72.0
 Central 0.696 0.654–0.736 <0.0001 >0 70.8 63.3
 Total 0.795 0.758–0.829 >4 81.2 67.6
Table 3
 
Fluorescein Staining of Two Corneal Zones Is Sufficient to Predict the Severity of Dry Eye Disease in an Undifferentiated Population of Patients With Mild to Severe Dry Eye
Table 3
 
Fluorescein Staining of Two Corneal Zones Is Sufficient to Predict the Severity of Dry Eye Disease in an Undifferentiated Population of Patients With Mild to Severe Dry Eye
Zone Pairs AUC* 95% CI* P Value* Cutoff Sensitivity, % Specificity, %
Aqueous deficiency, n = 171
 Superior/inferior 0.797 0.759–0.831 0.130‡ >1 85.4 62.5
 Superior/temporal 0.739 0.699–0.777 <0.0001 >0 74.3 66.4
 Superior/nasal 0.762 0.723–0.799 0.0001 >1 72.5 71.1
 Superior/central 0.720 0.679–0.758 <0.0001 >1 57.3 77.3
 Inferior/temporal 0.794 0.756–0.828 0.0468 >1 85.4 60.5
 Inferior/nasal 0.792 0.754–0.826 0.0436 >3 63.2 81.7
 Inferior/central 0.792 0.754–0.826 0.0171 >2 73.1 73.5
 Temporal/nasal 0.764 0.725–0.801 0.0001 >1 71.3 70.2
 Temporal/central 0.740 0.699–0.777 <0.0001 >2 55.0 83.8
 Nasal/central 0.755 0.716–0.792 <0.0001 >1 73.7 65.5
 Total 0.816 0.779–0.848 >4 72.5 74.0
Evaporative, n = 180
 Superior/inferior 0.731 0.690–0.769 0.0007 >2 64.4 73.6
 Superior/temporal 0.696 0.655–0.736 <0.0001 >0 72.2 66.4
 Superior/nasal 0.710 0.668–0.749 <0.0001 >0 83.9 51.2
 Superior/central 0.714 0.673–0.753 0.0001 >0 78.3 60.9
 Inferior/temporal 0.738 0.698–0.776 0.0004 >1 81.7 59.7
 Inferior/nasal 0.739 0.699–0.777 0.001 >1 87.2 53.6
 Inferior/central 0.759 0.720–0.796 0.0593‡ >1 84.4 56.4
 Temporal/nasal 0.718 0.677–0.757 <0.0001 >1 67.8 69.4
 Temporal/central 0.721 0.680–0.760 <0.0001 >0 80.0 60.3
 Nasal/central 0.738 0.698–0.776 0.0009 >1 72.8 66.1
 Total 0.778 0.740–0.814 >2 89.4 56.4
Mixed, n = 96
 Superior/inferior 0.765 0.725–0.801 0.565‡ >2 77.1 68.8
 Superior/temporal 0.719 0.678–0.758 0.0002 >0 79.2 60.1
 Superior/nasal 0.754 0.714–0.791 0.0110 >1 78.1 64.5
 Superior/central 0.712 0.671–0.751 0.0001 >0 80.2 53.4
 Inferior/temporal 0.761 0.722–0.798 0.0133 >2 77.1 67.1
 Inferior/nasal 0.771 0.732–0.807 0.0887‡ >3 69.8 75.1
 Inferior/central 0.778 0.739–0.813 0.106‡ >2 82.3 67.1
 Temporal/nasal 0.746 0.706–0.783 0.0032 >1 78.1 64.3
 Temporal/central 0.737 0.706–0.783 0.0003 >2 62.5 78.5
 Nasal/central 0.757 0.717–0.794 0.0063 >1 80.2 59.9
 Total 0.795 0.758–0.829 >4 81.2 67.6
Review of the corneal staining patterns for each type of dry eye disease (see Fig. 1) indicated that the predictive zone pairs were not simply those with the highest mean staining, although in all cases the inferior zone was required to predict the presence of a dry eye subtype. Review of the data (Table 3) show that at the ideal points of the ROC, the thresholds for the Baylor scores with optimal sensitivity and specificity differ between the scores from zone pairs and those from sum of the five zones. We found that a Baylor score greater than 1 for the superior/inferior zone pair predicted aqueous disease with comparable efficacy to five zones. Similarly, the combination of inferior/central zones with a Baylor score greater than 1 predicted evaporative dry eye, whereas superior/inferior, inferior/nasal, and inferior/central zone pairs predicted dry eye at Baylor score cutoffs of greater than 2, greater than 3, and greater than 2, respectively. The actual sensitivities and specificities attained by staining grades in zone pairs were surprisingly high. In some instances, the sensitivity achieved by the zone pairs could be as high as 87% and the specificity as high as 84%. 
Common Dry Eye Risk Factors as Potential Confounders
We next sought to address the possibility that confounding factors known to confer a high risk of dry eye disease, specifically a patient history of smoking, contact lens use, or ocular surgery, were influencing our statistical analyses. Thus, we removed patients with these risk factors and repeated the ROC analysis. Zone pairs that predicted dry eye severity in the original undifferentiated population continued to be predictive after elimination of the higher-risk patients (Table 4), while no additional zone pairs became statistically similar to total corneal scoring after this analysis. This indicates that our originally identified zone pairs are predictive of dry eye disease regardless of the presence or absence of smoking, contact lens use, or surgical history. 
Table 4
 
Fluorescein Staining of Two Corneal Zones Predicts Dry Eye Severity With Similar Reliability in an Undifferentiated Patient Population and Following Removal of Patients With a History of Smoking, Ocular Surgery, or Contact Lens Use
Table 4
 
Fluorescein Staining of Two Corneal Zones Predicts Dry Eye Severity With Similar Reliability in an Undifferentiated Patient Population and Following Removal of Patients With a History of Smoking, Ocular Surgery, or Contact Lens Use
Zone Pairs AUC* P Value
Undifferentiated Nonsmoker/Nonsurgical/No Contacts Undifferentiated Nonsmoker/Nonsurgical/No Contacts
Aqueous deficiency
 Superior/inferior 0.797 0.799 0.130 0.256
 Total 0.816 0.816
Evaporative
 Inferior/central 0.759 0.777 0.0593 0.259
 Total 0.778 0.790
Mixed
 Superior/inferior 0.765 0.768 0.565 0.243
 Inferior/nasal 0.771 0.765 0.088 0.120
 Inferior/central 0.778 0.772 0.106 0.158
 Total 0.795 0.791
Single-Zone Analysis in Autoimmune Disease Patients
Because the Baylor score for all five zones is obtained by addition of the scores from each zone, this total score may have lost information on the zone-specific predisposition of the corneal staining. To test the hypothesis that systemic diseases affect corneal epithelial damage via specific mechanisms and may have predilection for specific corneal zones, we next sought to determine in patients with a systemic disease whether dry eye subtype could be predicted by analysis of a specific single corneal zone. 
Interestingly we found two zones producing an AUC value exceeding that of the five-zone AUC value, confirming that in some instances zone-specific information is more valuable than summation over the entire cornea. Because of the relatively small sizes of the patient subgroups used in this analysis, we continued subsequent calculations only for those zones that yielded AUC values greater than those obtained when the total corneal surface was scored, effectively increasing the stringency of the comparisons. Scoring of only the inferior corneal zone was sufficient to predict aqueous (AUCInf 0.751 versus AUCTotal 0.750), evaporative (AUCInf 0.756 versus AUCTotal 0.752), and mixed (AUCInf 0.831 versus AUCTotal 0.788) dry eye in patients with diabetes mellitus (Table 5). On the other hand, the inferior corneal zone detected only the aqueous dry eye in patients with rheumatoid arthritis (AUCInf 0.804 versus AUCTotal 0.785). Surprisingly, the superior zone predicted evaporative disease with greater efficacy than five-zone analysis (AUCSup 0.765 versus AUCTotal 0.752) in patients with a history of thyroid disease (Table 5). 
Table 5
 
Dry Eye Disease in Patients With Specific Systemic Disease Is Predicted by Single Corneal Zone Analysis With Equivalent Significance to Five-Zone Corneal Analysis
Table 5
 
Dry Eye Disease in Patients With Specific Systemic Disease Is Predicted by Single Corneal Zone Analysis With Equivalent Significance to Five-Zone Corneal Analysis
Subgroup/Condition* AUC 95% CI P Value Cutoff Sensitivity, % Specificity, %
Diabetes/aqueous, n = 12
 Inferior 0.751 0.711–0.788 0.978 >2 75.0 67.5
 Total 0.750 0.710–0.787 >3 91.7 50.0
Diabetes/evaporative, n = 10
 Inferior 0.756 0.716–0.793 0.940 >2 80.0 67.4
 Total 0.752 0.712–0.789 >3 100 50.0
Diabetes/mixed, n = 9
 Inferior 0.831 0.795–0.862 0.387 >2 88.9 67.5
 Total 0.788 0.750–0.822 >4 100 59.5
Rheumatoid arthritis/aqueous, n = 11
 Inferior 0.804 0.767–0.838 0.674 >2 81.8 67.5
 Total 0.785 0.747–0.820 >4 100 59.5
Thyroid disease/evaporative, n = 12
 Superior 0.765 0.726–0.801 0.859 >0 75.0 74.5
 Total 0.752 0.712–0.789 >3 91.7 50.0
Discussion
This prospective study of dry eye disease found that the hitherto accepted standard of corneal staining and scoring of the entire corneal surface is not necessary to accurately diagnose the presence of subtypes of dry eye disease. Scoring of only two zones was required to achieve statistical equivalence to five-zone scoring when analyzing an undifferentiated population of patients with mild to severe dry eye disease, although single-zone scoring was insufficient to predict disease subtype in this setting. Furthermore, prediction of dry eye disease subtype with only a single zone is possible with patients with certain systemic diseases. 
Corneal staining of the inferior zone was most predictive of the three forms of dry eye disease examined in our study, and other workers have noted the vulnerability of this zone to damage in dry eyes. 21,22 It has been suggested that a combination of factors, including rubbing due to irritation, blepharitis, and contact lens removal contribute to selective abrasion of the inferior corneal zone, 2124 although it remains unknown whether aqueous deficiency or tear dysfunction can independently contribute to inferior zone damage. A further possibility is the accumulation of proinflammatory cytokines in the inferior and superior tear meniscus/reservoirs, which selectively damages the inferior and superior cornea epithelia respectively. 
The study did not show any patients with predominantly horizontal midzone corneal staining, a characteristic of patients with exposure-type of dry eye. This probably reflects the scarcity of patients with overt lagophthalmos in this clinic. 
Although individual corneal zones were insufficient to predict dry eye in the undifferentiated patient cohort, they were predictive for certain subgroups, suggesting that it is possible to expedite the process of corneal stain scoring in selected patients. Scoring of only the inferior zone was sufficient to predict all three types of dry eye in diabetic patients. Although the increased prevalence of dry eye among diabetic patients has recently been appreciated, 25,26 the pathophysiology behind this phenomenon remains poorly understood. Defects in tear production and corneal repair, as well as increased corneal sensitivity, are noted to occur in type 2 diabetic patients. 27 We further noted scoring of the inferior zone also was sufficient to predict aqueous dry eye in patients with rheumatoid arthritis, whereas the superior zone predicted evaporative dry eye disease in patients with thyroid disease. Associations between dry eye and rheumatoid arthritis are well established, 2830 and the associated autoimmune response significantly impacts tear production even in the absence of Sjögren syndrome. 30 The apparent utility of superior zone scoring for evaporative dry eye in patients with thyroid disease is intriguing, and may be related to the impact of iodine radiotherapy on lacrimal gland function 31 and exophthalmos in patients with ocular manifestations of thyroid disease. 32 Interestingly, superior limbic keratoconjunctivitis is a subtype of dry eye associated with thyroid disease that involves superior corneal staining and laxity of the superior bulbar conjunctiva, probably deleterious to tear spreading. 
The current study used a large prospectively recruited cohort of 510 patients with mild to severe dry eye disease, and single-operator corneal assessment of all patients reduced the variability in corneal scoring that commonly occurs during assessment by multiple clinicians. 11 An additional strength of this study was the creation of a substantial dry eye patient database that enabled us to elucidate features of dry eye disease in certain poorly represented patient subgroups, including those with certain autoimmune diseases coupled with certain subtypes of dry eye disease. In terms of study weaknesses, there was no disease-free control group used in the current study and we were thus unable to perform the ROC statistical analysis using asymptomatic patients as the negative state. Moreover, findings in patients from a tertiary clinic may not be applicable to patients with milder dry eye disease in other populations. 
Our study findings hold special importance for the clinical ophthalmologist confronted with a dry eye patient whose proper workup with corneal and tear film characterization is notoriously time-consuming. These data indicate that clinical characterization of dry eye disease by corneal fluorescein staining may be substantially hastened without compromising sensitivity or specificity. This may encourage more ophthalmologists to pursue corneal staining as part of their routine examination, or by saving time in the fluorescein dye examination, facilitate additional workup of dry eye patients using tests that they may have previously avoided. 
Moreover, we found that dry eye disease subtypes could be adequately predicted in patients with certain systemic conditions by scoring only a single corneal zone, further expediting clinical diagnosis of dry eye. The fact that the single zone is differentially located for different systemic diseases may reflect differences in pathological mechanisms of dry eye in these diseases, which warrants further study. In contrast to dry eye patients in general, in those patients with known thyroid disease, rheumatoid arthritis, or diabetes mellitus, grading the staining in one zone may be of greater value than noting the general level of staining over the whole cornea. In the design of clinical treatment trials of such patients, using the staining severity as an outcome measure, it may be more advantageous to monitor the zone-specific change rather than the change in the sum of the five zones. Future studies are required to determine if these results apply to staining with dyes other than fluorescein, or to zones of conjunctival staining. 
In summary, we find documentation of corneal fluorescein staining to be acceptable in a limited part of the cornea compared with the whole cornea in the detection of dry eye subtypes. In dry eye patients with diabetes mellitus, rheumatoid arthritis, or thyroid disease, examination of an even smaller area of the cornea for fluorescein dye staining may be just as efficient for the same purpose. 
Acknowledgments
Supported by the Singapore National Research Foundation under the Clinician Scientist Award NMRC/CSA/045/2012, the Singapore Ministry of Health National Medical Research Council under the Individual Research Grant NMRC/1206/2009, and Research Center Grant NMRC/CG/SERI/2010. 
Disclosure: B.J. Fenner, None; L. Tong, None 
References
Schaumberg DA Dana R Buring JE Sullivan DA. Prevalence of dry eye disease among US men: estimates from the Physicians' Health Studies. Arch Ophthalmol . 2009; 127: 763–768. [CrossRef] [PubMed]
Gayton JL. Etiology, prevalence, and treatment of dry eye disease. Clin Ophthalmol . 2009; 3: 405–412. [CrossRef] [PubMed]
Lin PY Tsai SY Cheng CY Liu JH Chou P Hsu WM. Prevalence of dry eye among an elderly Chinese population in Taiwan: the Shihpai Eye Study. Ophthalmology . 2003; 110: 1096–1101. [CrossRef] [PubMed]
Yamada M Mizuno Y Shigeyasu C. Impact of dry eye on work productivity. Clinicoecon Outcomes Res . 2012; 4: 307–312. [CrossRef] [PubMed]
Waduthantri S Yong SS Tan CH Cost of dry eye treatment in an Asian clinic setting. PLoS One . 2012; 7: e37711. [CrossRef] [PubMed]
Mizuno Y Yamada M Shigeyasu C. Annual direct cost of dry eye in Japan. Clin Ophthalmol . 2012; 6: 755–760. [CrossRef] [PubMed]
Yu J Asche CV Fairchild CJ. The economic burden of dry eye disease in the United States: a decision tree analysis. Cornea . 2011; 30: 379–387. [CrossRef] [PubMed]
The definition and classification of dry eye disease: report of the Definition and Classification Subcommittee of the International Dry Eye WorkShop (2007). Ocul Surf . 2007; 5: 75–92. [CrossRef] [PubMed]
Lemp MA. Report of the National Eye Institute/Industry workshop on Clinical Trials in Dry Eyes. CLAO J . 1995; 21: 221–232. [PubMed]
Khanal S Tomlinson A Diaper CJ. Tear physiology of aqueous deficiency and evaporative dry eye. Optom Vis Sci . 2009; 86: 1235–1240. [CrossRef] [PubMed]
Sullivan BD Whitmer D Nichols KK An objective approach to dry eye disease severity. Invest Ophthalmol Vis Sci . 2010; 51: 6125–6130. [CrossRef] [PubMed]
McGinnigle S Naroo SA Eperjesi F. Evaluation of dry eye. Surv Ophthalmol . 2012; 57: 293–316. [CrossRef] [PubMed]
Tomlinson A Khanal S Ramaesh K Diaper C McFadyen A. Tear film osmolarity: determination of a referent for dry eye diagnosis. Invest Ophthalmol Vis Sci . 2006; 47: 4309–4315. [CrossRef] [PubMed]
Savini G Prabhawasat P Kojima T Grueterich M Espana E Goto E. The challenge of dry eye diagnosis. Clin Ophthalmol . 2008; 2: 31–55. [CrossRef] [PubMed]
Bron AJ Evans VE Smith JA. Grading of corneal and conjunctival staining in the context of other dry eye tests. Cornea . 2003; 22: 640–650. [CrossRef] [PubMed]
De Paiva CS Pflugfelder SC. Corneal epitheliopathy of dry eye induces hyperesthesia to mechanical air jet stimulation. Am J Ophthalmol . 2004; 137: 109–115. [CrossRef] [PubMed]
Foulks GN. Treatment of dry eye disease by the non-ophthalmologist. Rheum Dis Clin North Am . 2008; 34: 987–1000, x. [CrossRef] [PubMed]
Parkinson J. Taking on dry eye. Ophthalmol Manage . 2005; 6. Available at: http://www.ophthalmologymanagement.com/articleviewer.aspx?articleid=86275 . Accessed November 23, 2013.
Lee SY Petznick A Tong L. Associations of systemic diseases, smoking and contact lens wear with severity of dry eye. Ophthalmic Physiol Opt . 2012; 32: 518–526. [CrossRef] [PubMed]
DeLong ER DeLong DM Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics . 1988; 44: 837–845. [CrossRef] [PubMed]
Liew MS Zhang M Kim E Akpek EK. Prevalence and predictors of Sjogren's syndrome in a prospective cohort of patients with aqueous-deficient dry eye. Br J Ophthalmol . 2012; 96: 1498–1503. [CrossRef] [PubMed]
Begley CG Chalmers RL Abetz L The relationship between habitual patient-reported symptoms and clinical signs among patients with dry eye of varying severity. Invest Ophthalmol Vis Sci . 2003; 44: 4753–4761. [CrossRef] [PubMed]
Begley CG Barr JT Edrington TB Long WD McKenney CD Chalmers RL. Characteristics of corneal staining in hydrogel contact lens wearers. Optom Vis Sci . 1996; 73: 193–200. [CrossRef] [PubMed]
Josephson JE Caffery BE. Corneal staining characteristics after sequential instillations of fluorescein. Optom Vis Sci . 1992; 69: 570–573. [CrossRef] [PubMed]
Manaviat MR Rashidi M Afkhami-Ardekani M Shoja MR. Prevalence of dry eye syndrome and diabetic retinopathy in type 2 diabetic patients. BMC Ophthalmol . 2008; 8: 10. [CrossRef] [PubMed]
Kaiserman I Kaiserman N Nakar S Vinker S. Dry eye in diabetic patients. Am J Ophthalmol . 2005; 139: 498–503. [CrossRef] [PubMed]
Zagon IS Sassani JW Immonen JA McLaughlin PJ. Ocular surface abnormalities related to type 2 diabetes are reversed by the opioid antagonist naltrexone [published online ahead of print June 18, 2013]. Clin Experiment Ophthalmol . doi:10.1111/ceo.12144 .
El-Shazly AA Mohamed AA. Relation of dry eye to disease activity in juvenile rheumatoid arthritis. Eur J Ophthalmol . 2012; 22: 330–334. [CrossRef] [PubMed]
Barendregt PJ van der Heijde GL Breedveld FC Markusse HM. Parasympathetic dysfunction in rheumatoid arthritis patients with ocular dryness. Ann Rheum Dis . 1996; 55: 612–615. [CrossRef] [PubMed]
Fujita M Igarashi T Kurai T Sakane M Yoshino S Takahashi H. Correlation between dry eye and rheumatoid arthritis activity. Am J Ophthalmol . 2005; 140: 808–813. [CrossRef] [PubMed]
Koca G Acar U Atilgan HI Changes in conjunctival cytology and tear function tests with radioiodine treatment for hyperthyroidism. Ann Nucl Med . 2013; 27: 694–699. [CrossRef] [PubMed]
Ismailova DS Fedorov AA Grusha YO. Ocular surface changes in thyroid eye disease. Orbit . 2013; 32: 87–90. [CrossRef] [PubMed]
Figure 1
 
Corneal staining distribution in patients with aqueous, evaporative, and mixed dry eye disease. The mean Baylor staining scores of corneal zones are shown for patients meeting criteria for each of the dry eye types. Error bars represent the SD.
Figure 1
 
Corneal staining distribution in patients with aqueous, evaporative, and mixed dry eye disease. The mean Baylor staining scores of corneal zones are shown for patients meeting criteria for each of the dry eye types. Error bars represent the SD.
Figure 2
 
ROC curve analysis indicates that fluorescein staining of two corneal zones is statistically similar to full corneal analysis to predict dry eye severity. ROC curves were generated as described in the Methods section. Curves are shown for aqueous (A), evaporative (B), and mixed (C) dry eye disease. Total, five-zone corneal scoring; S-I, superior/inferior zones; S-C, superior/central zones; S-T, superior/temporal zones; I-C, inferior/central zones; I-N, inferior/nasal zones. The marker line represents the curve expected if the diagnostic test holds no predictive value (i.e., AUC = 0.5).
Figure 2
 
ROC curve analysis indicates that fluorescein staining of two corneal zones is statistically similar to full corneal analysis to predict dry eye severity. ROC curves were generated as described in the Methods section. Curves are shown for aqueous (A), evaporative (B), and mixed (C) dry eye disease. Total, five-zone corneal scoring; S-I, superior/inferior zones; S-C, superior/central zones; S-T, superior/temporal zones; I-C, inferior/central zones; I-N, inferior/nasal zones. The marker line represents the curve expected if the diagnostic test holds no predictive value (i.e., AUC = 0.5).
Table 1
 
Characteristics of the Study Population
Table 1
 
Characteristics of the Study Population
Characteristic Value*
n 510
Age, y, mean ± SD 53.0 ± 14.1
Male sex, % 25.5
Race, %
 Chinese 87.8
 Other 12.2
Medical history, n (%)
 Diabetes mellitus 30 (5.9)
 Rheumatoid arthritis 25 (4.9)
 Thyroid disease 41 (8.0)
 Currently smoking 33 (6.5)
Dry eye disease subtype,† n (%)
 Aqueous 171 (33.5)
 Evaporative 180 (35.3)
 Mixed 96 (18.8)
Ocular history, n (%)
 LASIK 23 (4.5)
 Cataract surgery 41 (8.0)
 Contact lens wear 90 (17.6)
Baylor corneal staining scores, mean ± SD
 Superior 0.55 ± 1.07
 Inferior 1.66 ± 1.50
 Temporal 0.91 ± 1.38
 Nasal 1.27 ± 1.45
 Central 0.99 ± 1.40
 Total 5.38 ± 5.48
Schirmer distance, mm, mean ± SD 11.3 ± 8.57
TBUT, s, mean ± SD 3.01 ± 1.58
Table 2
 
Fluorescein Staining of Individual Corneal Zones Does Not Predict the Severity of Dry Eye Disease As Well As Combined Corneal Zones in an Undifferentiated Population
Table 2
 
Fluorescein Staining of Individual Corneal Zones Does Not Predict the Severity of Dry Eye Disease As Well As Combined Corneal Zones in an Undifferentiated Population
Zone AUC* 95% CI* P Value* Cutoff Sensitivity, % Specificity, %
Aqueous deficiency, n = 171
 Superior 0.652 0.609–0.693 <0.0001 >0 45.6 82.9
 Inferior 0.773 0.734–0.809 0.0054 >1 77.2 67.6
 Temporal 0.714 0.673–0.753 <0.0001 >0 64.9 73.7
 Nasal 0.731 0.690–0.769 <0.0001 >1 60.2 77.9
 Central 0.685 0.643–0.725 <0.0001 >0 63.7 67.3
 Total 0.816 0.779–0.848 >4 72.5 74.0
Evaporative, n = 180
 Superior 0.595 0.551–0.638 <0.0001 >0 38.3 79.7
 Inferior 0.718 0.676–0.756 0.0001 >1 68.9 64.2
 Temporal 0.673 0.630–0.713 <0.0001 >0 62.2 73.3
 Nasal 0.681 0.638–0.721 <0.0001 >0 75.0 55.2
 Central 0.685 0.624–0.725 <0.0001 >0 65.6 69.1
 Total 0.778 0.740–0.814 >2 89.4 56.4
Mixed, n = 96
 Superior 0.624 0.581–0.666 <0.0001 >0 45.8 77.8
 Inferior 0.747 0.707–0.784 0.0108 >2 65.6 73.9
 Temporal 0.698 0.656–0.737 <0.0001 >0 69.8 67.9
 Nasal 0.724 0.683–0.763 0.0007 >1 64.6 72.0
 Central 0.696 0.654–0.736 <0.0001 >0 70.8 63.3
 Total 0.795 0.758–0.829 >4 81.2 67.6
Table 3
 
Fluorescein Staining of Two Corneal Zones Is Sufficient to Predict the Severity of Dry Eye Disease in an Undifferentiated Population of Patients With Mild to Severe Dry Eye
Table 3
 
Fluorescein Staining of Two Corneal Zones Is Sufficient to Predict the Severity of Dry Eye Disease in an Undifferentiated Population of Patients With Mild to Severe Dry Eye
Zone Pairs AUC* 95% CI* P Value* Cutoff Sensitivity, % Specificity, %
Aqueous deficiency, n = 171
 Superior/inferior 0.797 0.759–0.831 0.130‡ >1 85.4 62.5
 Superior/temporal 0.739 0.699–0.777 <0.0001 >0 74.3 66.4
 Superior/nasal 0.762 0.723–0.799 0.0001 >1 72.5 71.1
 Superior/central 0.720 0.679–0.758 <0.0001 >1 57.3 77.3
 Inferior/temporal 0.794 0.756–0.828 0.0468 >1 85.4 60.5
 Inferior/nasal 0.792 0.754–0.826 0.0436 >3 63.2 81.7
 Inferior/central 0.792 0.754–0.826 0.0171 >2 73.1 73.5
 Temporal/nasal 0.764 0.725–0.801 0.0001 >1 71.3 70.2
 Temporal/central 0.740 0.699–0.777 <0.0001 >2 55.0 83.8
 Nasal/central 0.755 0.716–0.792 <0.0001 >1 73.7 65.5
 Total 0.816 0.779–0.848 >4 72.5 74.0
Evaporative, n = 180
 Superior/inferior 0.731 0.690–0.769 0.0007 >2 64.4 73.6
 Superior/temporal 0.696 0.655–0.736 <0.0001 >0 72.2 66.4
 Superior/nasal 0.710 0.668–0.749 <0.0001 >0 83.9 51.2
 Superior/central 0.714 0.673–0.753 0.0001 >0 78.3 60.9
 Inferior/temporal 0.738 0.698–0.776 0.0004 >1 81.7 59.7
 Inferior/nasal 0.739 0.699–0.777 0.001 >1 87.2 53.6
 Inferior/central 0.759 0.720–0.796 0.0593‡ >1 84.4 56.4
 Temporal/nasal 0.718 0.677–0.757 <0.0001 >1 67.8 69.4
 Temporal/central 0.721 0.680–0.760 <0.0001 >0 80.0 60.3
 Nasal/central 0.738 0.698–0.776 0.0009 >1 72.8 66.1
 Total 0.778 0.740–0.814 >2 89.4 56.4
Mixed, n = 96
 Superior/inferior 0.765 0.725–0.801 0.565‡ >2 77.1 68.8
 Superior/temporal 0.719 0.678–0.758 0.0002 >0 79.2 60.1
 Superior/nasal 0.754 0.714–0.791 0.0110 >1 78.1 64.5
 Superior/central 0.712 0.671–0.751 0.0001 >0 80.2 53.4
 Inferior/temporal 0.761 0.722–0.798 0.0133 >2 77.1 67.1
 Inferior/nasal 0.771 0.732–0.807 0.0887‡ >3 69.8 75.1
 Inferior/central 0.778 0.739–0.813 0.106‡ >2 82.3 67.1
 Temporal/nasal 0.746 0.706–0.783 0.0032 >1 78.1 64.3
 Temporal/central 0.737 0.706–0.783 0.0003 >2 62.5 78.5
 Nasal/central 0.757 0.717–0.794 0.0063 >1 80.2 59.9
 Total 0.795 0.758–0.829 >4 81.2 67.6
Table 4
 
Fluorescein Staining of Two Corneal Zones Predicts Dry Eye Severity With Similar Reliability in an Undifferentiated Patient Population and Following Removal of Patients With a History of Smoking, Ocular Surgery, or Contact Lens Use
Table 4
 
Fluorescein Staining of Two Corneal Zones Predicts Dry Eye Severity With Similar Reliability in an Undifferentiated Patient Population and Following Removal of Patients With a History of Smoking, Ocular Surgery, or Contact Lens Use
Zone Pairs AUC* P Value
Undifferentiated Nonsmoker/Nonsurgical/No Contacts Undifferentiated Nonsmoker/Nonsurgical/No Contacts
Aqueous deficiency
 Superior/inferior 0.797 0.799 0.130 0.256
 Total 0.816 0.816
Evaporative
 Inferior/central 0.759 0.777 0.0593 0.259
 Total 0.778 0.790
Mixed
 Superior/inferior 0.765 0.768 0.565 0.243
 Inferior/nasal 0.771 0.765 0.088 0.120
 Inferior/central 0.778 0.772 0.106 0.158
 Total 0.795 0.791
Table 5
 
Dry Eye Disease in Patients With Specific Systemic Disease Is Predicted by Single Corneal Zone Analysis With Equivalent Significance to Five-Zone Corneal Analysis
Table 5
 
Dry Eye Disease in Patients With Specific Systemic Disease Is Predicted by Single Corneal Zone Analysis With Equivalent Significance to Five-Zone Corneal Analysis
Subgroup/Condition* AUC 95% CI P Value Cutoff Sensitivity, % Specificity, %
Diabetes/aqueous, n = 12
 Inferior 0.751 0.711–0.788 0.978 >2 75.0 67.5
 Total 0.750 0.710–0.787 >3 91.7 50.0
Diabetes/evaporative, n = 10
 Inferior 0.756 0.716–0.793 0.940 >2 80.0 67.4
 Total 0.752 0.712–0.789 >3 100 50.0
Diabetes/mixed, n = 9
 Inferior 0.831 0.795–0.862 0.387 >2 88.9 67.5
 Total 0.788 0.750–0.822 >4 100 59.5
Rheumatoid arthritis/aqueous, n = 11
 Inferior 0.804 0.767–0.838 0.674 >2 81.8 67.5
 Total 0.785 0.747–0.820 >4 100 59.5
Thyroid disease/evaporative, n = 12
 Superior 0.765 0.726–0.801 0.859 >0 75.0 74.5
 Total 0.752 0.712–0.789 >3 91.7 50.0
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×