Tear Production Levels and Dry Eye Disease Severity in a Large Norwegian Cohort

ABSTRACT

Purpose

To determine if the Schirmer I test (without anesthesia) cut-off value is a predictor of dry eye severity in a large Norwegian cohort of dry eye disease (DED) patients, which are grouped into six levels of tear production.

Methods

Patients (n = 1090) with DED of different etiologies received an extensive dry eye work-up: osmolarity (Osm), tear meniscus height (TMH), tear film break-up time (TFBUT), ocular protection index (OPI), ocular surface staining (OSS), Schirmer I test (ST), meibum expressibility (ME), and meibum quality (MQ). Classification of dry eye severity level (DESL) and diagnosis of meibomian gland dysfunction (MGD) were also included. The cohort was divided into six groups: below and above cut-off values of 5 (groups 1 and 2), 10 (groups 3 and 4), and 15 mm (groups 5 and 6) of ST. Mann-Whitney test and Chi-Square test were used for group comparison of parameters (p ≤ 0.05).

Results

The groups 1, 3, and 5 had values indicating more severe DED than the groups 2, 4, 6 with significant difference in DESL, Osm, TFBUT, OPI, OSS, and TMH. Regardless of the choice of cut-off values, there was no statistically significant difference in ME, MQ, and MGD between groups below and above selected cut-off value. When gender difference was considered in each group, significant difference was only observed for DESL (groups 2, 4, and 5), TFBUT (groups 2, 4, and 5), OPI (groups 2 and 6), and ME (group1).

Conclusions

Schirmer I is a robust discriminator for DESL, Osm, TFBUT, OPI, OSS, and TMH, but not for ME, MQ, and MGD. Patients with lower tear production levels presented with more severe DED at all three defined cut-off values. Interestingly, the differences in the mean values of DESL were minimal although statistically significant. Thus, the clinical value of different Schirmer levels appears to be limited.

KEYWORDS:

• Dry eye disease
• Schirmer I test
• tear production levels
• clinical tests
• large population of patients

Introduction

Dry eye disease (DED) is a multifactorial disorder, which can be caused by an alteration in the quality or quantity of tear film’s three layers.¹ DED can result from primary factors such as lacrimal gland atrophy and secondary factors such as pathological changes in the eyelids, conjunctiva, or cornea along with other influencing factors, including immunological processes, neurotransmitters, hormones, pharmaceuticals, contact lenses and environmental pollution.^2–4 In addition to these causes, several risk factors have been proposed as contributing to the development of the disease: gender (i.e., female), increasing age, therapies (e.g., postmenopausal estrogen and radiation), dietary deficiencies (e.g., vitamin A and omega-3 fatty acids), and systemic diseases (e.g., hepatitis C).^5–8

Several clinical tests are available for diagnostic evaluations, clinical trials, and follow-up examinations, including Schirmer I test (ST), measurement of tear osmolarity (Osm), tear film break-up time (TFBUT), ocular surface staining (OSS), tear meniscus height (TMH), meibum expressibility (ME), and meibum quality (MQ). Among them, ST is the most widely used method for assessment of aqueous tear production. It was first introduced by Schirmer in 1903 as installed piece of striped and marked (35 by 5 mm) blotting paper on lower eyelid margin with or without anesthesia to collect tears. ST suffers from several drawbacks such as poor reproducibility, sensitivity and specificity, long testing time (i.e., 5 min), potential for evaporative loss, high variability, discomfort, possibility of uneven absorption of tear by paper strip, and no well-defined cut-off value.^9–13 Nevertheless, ST without topical anesthesia is considered a valid ophthalmological test for patients with severe DED.¹⁴ To assess whether the ST can be utilized as a discriminator for other objective tests, we investigated clinical parameters in a large Norwegian cohort of 1090 dry eye disease (DED) patients grouped into six levels of tear production. The six groups were below and above cut-off values of 5 mm (groups 1 and 2), 10 mm (groups 3 and 4), and 15 mm (groups 5 and 6) of the ST.

Materials and methods

Patients

One thousand and ninety subjects (average age: 52.86 ± 16.03 years, range: 8–95, 796 females and 294 males) diagnosed with DED of different etiologies were consecutively included in the study at the Norwegian Dry Eye Clinic between 2012 and 2016. All subjects received an extensive dry eye work-up.

All examinations were carried out at the same clinic by the same ophthalmologist during normal working hours between 9 AM and 3 PM. The use of the data for the study has been reviewed by The Regional Committee for Medical & Health Research Ethics, Section C, South East Norway (REC). REC found the research project “Evaluation of data from the Norwegian Dry Eye Clinic” to be outside the remit of the Act on Medical and Health Research (2008) and therefore could be implemented without its approval. Prior to data collection, informed consent was obtained from all participants. All procedures performed in this study were in compliance with the Declaration of Helsinki.

Clinical evaluation

All subjects underwent a comprehensive ophthalmic examination by the following stepwise tests: osmolarity (Osm), tear meniscus height (TMH), tear film break-up time (TFBUT), ocular protection index (OPI), ocular surface staining (OSS), Schirmer I test (ST), meibum expressibility (ME), meibum quality (MQ), diagnosis of meibomian gland dysfunction (MGD), and classification of dry eye severity level (DESL).

The tear quality evaluation was performed using TFBUT after instillation of 5 µl 2% fluorescein sodium. Ocular surface staining with fluorescein was recorded after Oxford Grading Scheme: 0–15; range of corneal staining: 0–5).¹⁵ Schirmer test was performed without topical anesthesia in 5 min. The calculation of OPI was based on the ratio of the TFBUT divided by the blink interval.¹⁶ TMH was measured using a slit lamp. ME was recorded based on the number of secreted glands of the lower eyelid viewed at the slit lamp when light pressure applied by cotton tips on central five MGs (0 = all five glands expressible; 1 = three to four glands expressible; 2 = one to two glands expressible; and 3 = no gland expressible). For MQ, the central eight glands of the lower eyelid were scored on a scale of 0 to 3 for each gland (0 = clear; 1 = cloudy; 2 = cloudy with debris (granular); and 3 = thick, like toothpaste). The sum of the central eight glands was then calculated (total score range, 0–24). DESL was assessed according to the Behrens et al.,¹⁷ modified by 2007 International Dry Eye Workshop criteria¹⁸ (Table 1), and diagnosis of MGD was based on the suggestions by the international workshop on MGD.¹⁹ Only 520 subjects were examined for tear film osmolarity measurement using TearLab Osmolarity System (TearLab Corp, San Diego, CA).

Table 1. Dry eye severity grading scheme.

Dry eye severity level	1	2	3	4*
Discomfort, severity, and frequency	Mild and/or episodic; occurs under environmental stress	Moderate episodic or chronic, stress or no stress	Severe frequent or constant without stress	Severe and/or disabling and constant
Visual symptoms	None or episodic mild fatigue	Annoying and/or activity-limiting episodic	Annoying, chronic and/or constant, limiting activity	Constant and/or possibly disabling
Conjunctival injection	None to mild	None to mild	+/–	+/++
Conjunctival staining	None to mild	Variable	Moderate to marked	Marked
Corneal staining (severity/location)	None to mild	Variable	Marked central	Severe punctate erosions
Corneal/tear signs	None to mild	Mild debris, ↓ meniscus	Filamentary keratitis, mucus clumping, ↑ tear debris	Filamentary keratitis, mucus clumping, ↑ tear debris, ulceration
Lid/meibomian glands	MGD variably present	MGD variably present	Frequent	Trichiasis, keratinization, symblepharon
TFBUT (sec)	Variable	≤ 10	≤ 5	Immediate
Schirmer score (mm/5 min)	Variable	≤ 10	≤ 5	≤ 2

*Must have signs and symptoms. TBUT: fluorescein tear break-up time. MGD: meibomian gland disease.

Reprinted with permission from Behrens A, Doyle JJ, Stern L, et al. Dysfunctional tear syndrome. A Delphi approach to treatment recommendations. Cornea 2006;25:90–7 and Lemp, MA, Foulks, GN, The definition and classification of dry eye disease: Report of the definition and classification subcommittee of the international dry eye workshop (2007). The Ocular Surface, 2007;5(2);75–92.

Statistical analysis

Data from right eye were used for statistical analysis. The cohort was divided into six groups: below and above cut-off values (5, 10, and 15 mm) of ST for comparisons. Mann-Whitney test and Chi-Square test were used for group comparison of parameters using SPSS version 21 (SPSS Inc., Chicago, IL, USA). Chi-Square test was used to determine if females were different from males in each group. Data are presented as mean and standard error. P values less than or equal to 0.05 were considered significant.

Results

Table 2 presents the subject demographics (average age: 52.86 ± 16.03 years, range: 8–95). The number of females (n = 796) was higher than males (n = 294). In all groups, except age group 0–19, females outnumbered males. The highest number of patients belonged to age group 40–59 (n = 430). In contrast, age groups 80–99 (n = 34) and 0–19 (n = 9) presented the lowest number of patients, respectively.

Table 2. Subject demographics: The age and gender distribution of 1090 subjects with test for gender difference in each group (Chi-square test, p ≤ 0.05).

	Number of subjects
Age group	Total	Female	Male	p value
0–19	9	4	5	0.739
20–39	241	148	93	< 0.001*
40–59	430	326	104	< 0.001*
60–79	376	290	86	< 0.001*
80–99	34	28	6	< 0.001*
Total	1090	796	294

Table 3 shows the distribution of Schirmer test wetting length. The overall average was 14.67 ± 0.308 mm. Among participants, 20%, 45%, and 62% had Schirmer test of < 5 mm, ≤ 10 mm, and ≤ 15 mm, respectively. The percentage of fully soaked strip (≥ 30 mm) was 12.7%.

Table 3. Distribution of Schirmer I test wetting length.

	Cut-off value: 5		Cut-off value: 10		Cut-off value: 15
	Group 1 (ST≤ 5)	Group 2 (ST≥ 6)	Group 3 (ST≤ 10)	Group 4 (ST≥ 11)	Group 5 (ST≤ 15)	Group 6 (ST≥ 16)
Mean ± SE	5.62 ± 2.417	23.05 ± 3.55	12.81 ± 3.624	15.86 ± 3.073	17.83 ± 3.755	10.83 ± 2.816
Range	0 (16)-5 (63)	6 (45)-36 (1)	0 (16)-10 (59)	11 (34)-36 (1)	0 (16)-15 (35)	16 (22)-36 (1)
Percent	20%	80%	45%	55%	62%	38%

– The frequencies are given in parentheses.

The analysis revealed that groups 1, 3, and 5 (i.e., individuals with values below cut-off levels of ≤ 5 mm, ≤ 10 mm, and ≤ 15 mm, respectively) had values indicating more severe DED with significant difference of p ≤ 0.05 for DESL, Osm, TFBUT, OPI, OSS and TMH compared to the corresponding groups (2, 4, and 6) (Table 4). Regardless of the choice of cut-off values, there was no statistically significant difference in ME, MQ and MGD between the groups. The Chi-Square test indicated significant difference between six tear production level groups for TMH (p < 0.001) in all three defined cut-off values but there was no significant difference for MGD (Table 4).

Table 4. Comparison of clinical parameter results between groups based on three defined cut-off values (5, 10, and 15) for pathological DED measured by Schirmer I test (Mann-Whitney test and Chi-Square test, p ≤ 0.05).

	Cut-off value: 5			Cut-off value: 10			Cut-off value: 15
	Group 1 (ST≤ 5)	Group 2 (ST≥ 6)	p	Group 3 (ST≤ 10)	Group 4 (ST≥ 11)	p	Group 5 (ST≤ 15)	Group 6 (ST≥ 16)	p
DESL	2.15 ± 0.073 (0–4)	2.00 ± 0.025 (0–4)	< 0.001*	2.11 ± 0.038 (0–4)	1.96 ± 0.030 (0–3)	< 0.001*	2.07 ± 0.032 (0–4)	1.95 ± 0.035 (0–3)	< 0.001*
Osm^1	320.08 ± 2.247 (287–380)	314.77 ± 0.848 (280–398)	0.003*	318.38 ± 1.357 (283–385)	313.72 ± 0.968 (280–398)	< 0.001*	316.76 ± 1.118 (280–398)	313.90 ± 1.115 (280–386)	0.021*
TFBUT	3.07 ± 0.338 (1–14)	3.88 ± 0.159 (1–20)	< 0.001*	3.30 ± 0.234 (1–20)	4.06 ± 0.183 (1–20)	< 0.001*	3.49 ± 0.191 (1–20)	4.13 ± 0.221 (1–20)	< 0.001*
OPI	1.243 ± 0.131 (0.07–5)	1.656 ± 0.082 (0.1–10)	< 0.001*	1.413 ± 0.109 (0.07–10)	1.715 ± 0.096 (0.1–10)	< 0.001*	1.519 ± 0.093 (0.07–10)	1.702 ± 0.116 (0.1–10)	< 0.001*
OSS	2.10 ± 0.322 (0–13)	1.62 ± 0.099 (0–10)	< 0.001*	2.08 ± 0.184 (0–13)	1.44 ± 0.104 (0–10)	< 0.001*	1.93 ± 0.142 (0–13)	1.37 ± 0.120 (0–10)	< 0.001*
TMH^2	(0–1)	(0–1)	< 0.001*	(0–1)	(0–1)	< 0.001*	(0–1)	(0–1)	< 0.001*
ME	1.43 ± 0.106 (0–3)	1.67 ± 0.043 (0–3)	0.155	1.54 ± 0.064 (0–3)	1.69 ± 0.052 (0–3)	0.054	1.61 ± 0.54 (0–3)	1.67 ± 0.061 (0–3)	0.257
MQ	8.18 ± 0.637 (0–22)	8.61 ± 0.247 (0–24)	0.155	8.67 ± 0.374 (0–22)	8.48 ± 0.292 (0–24)	0.201	8.69 ± 0.310 (0–24)	8.37 ± 0.343 (0–24)	0.799
MGD^2	(0–1)	(0–1)	0.623	(0–1)	(0–1)	0.277	(0–1)	(0–1)	0.116

– Data are presented as the mean ± standard error and the range of values are given in parentheses.

– Asterisk (*) above bar denotes p ≤ 0.05, Chi-Square test for TMH and MGD and Mann-Whitney T-test for the remaining clinical parameters.

– ¹ Osm was only assessed in 520 subjects in comparison with other clinical tests (n = 1090).

– ² TMH and MGD are dichotomous (0 and 1), and there is no mean ± standard error.

– Abbreviations: Schirmer I test (ST), dry eye severity level (DESL), Osmolarity (Osm), tear film break-up time (TFBUT), ocular protection index (OPI), ocular surface staining (OSS, Oxford Grading Scheme), tear meniscus height (TMH), meibum expressibility (ME), meibum quality (MQ), and meibomian gland disease (MGD).

When gender difference was considered in each group (Table 5), Chi-Square test indicated significant difference only for DESL (groups 2, 4, and 5), TFBUT (groups 2, 4, and 5), OPI (groups 2 and 6), and ME (group1).

Table 5. Comparison of gender difference in each group below and above Schirmer I test cut-off values (5, 10, and 15 mm) for clinical parameter results (Chi-Square test, p ≤ 0.05).

		Cut-off value: 5		Cut-off value: 10		Cut-off value: 15
		Group 1 (ST≤ 5)	Group 2 (ST≥ 6)	Group 3 (ST≤ 10)	Group 4 (ST≥ 11)	Group 5 (ST≤ 15)	Group 6 (ST≥ 16)
DESL	F	2.02 ± 0.029 (0–4)	2.09 ± 0.083 (0–3)	1.40 ± 0.245 (1–2)	2.04 ± 0.028 (0–4)	2.12 ± 0.038 (0–4)	1.94 ± 0.040 (0–3)
	M	1.92 ± 0.044 (1–3)	2.31 ± 0.151 (2–4)		2.03 ± 0.041 (1–4)	2.00 ± 0.07 (1–4)	1.95 ± 0.056 (1–3)
	p	0.510	< 0.001*	0.691	< 0.001*	< 0.001*	0.247
Osm^1	F	316.13 ± 1.020 (276–398)	317.11 ± 2.240 (287–355)	312 ± 7.287 (289–330)	316.33 ± 0.939 (276–398)	317.72 ± 1.357 (280–398)	315.27 ± 1.251 (280–386)
	M	310.18 ± 1.421 (283–368)	328.44 ± 5.415 (304–380)	304.5 ± 2.766 (295–312)	313.12 ± 1.603 (283–380)	316.86 ± 2.423 (283–380)	308.56 ± 1.760 (289–368)
	p	0.538	0.354	0.073	0.693	0.347	0.391
TFBUT	F	3.52 ± 0.162 (1–20)	3.18 ± 0.411 (1–14)	15.60 ± 1.860 (11–20)	3.28 ± 0.124 (1–10)	3.11 ± 0.201 (1–20)	3.88 ± 0.227 (1–20)
	M	4.85 ± 0.369 (1–20)	2.75 ± 0.581 (1–8)	14.67 ± 1.358 (12–20)	4.02 ± 0.258 (1–10)	4.21 ± 0.481 (1–17)	4.92 ± 0.464 (1–20)
	p	0.833	< 0.001*	0.210	< 0.001*	< 0.001*	0.158
OPI	F	1.5478 ± 0.085 (0.1–10)	1.340 ± 0.164 (0.2–5)	7.248 ± 1.179 (4.24–10)	1.428 ± 0.064 (0.10–9.09)	1.417 ± 0.097 (0.10–10)	1.635 ± 0.123 (0.10–10)
	M	1.94 ± 0.191 (0.1–10)	0.9713 ± 0.189 (0.07–2.67)	5.671 ± 1.00 (2.40–8.57)	1.601 ± 0.148 (0.07–10)	1.650 ± 0.238 (0.07–8.57)	1.967 ± 0242 (0.20–10)
	p	0.347	< 0.039*	0.111	0.251	0.188	0.038*
OSS	F	1.71 ± 0.116 (0–10)	1.96 ± 0.305 (0–9)	1.60 ± 1.122 (0–6)	1.74 ± 0.109 (0–10)	2.07 ± 0.171 (0–10)	1.42 ± 0.128 (0–10)
	M	1.32 ± 0.181 (0–10)	2.50 ± 0.894 (0–13)	0.17 ± 0.167 (0–1)	1.55 ± 0.209 (0–13)	1.66 ± 0.330 (0–13)	1.31 ± 0.230 (0–8)
	p	0.548	0.287	0.104	0.251	0.332	0.099
TMH^2	F	(0–1)	(0–1)	(0–1)	(0–1)	(0–1)	(0–1)
	M	(0–1)	(0–1)	(0–1)	(0–1)	(0–1)	(0–1)
	p	0.409	0. 451	0.867	0.944	0.931	0.660
ME	F	1.71 ± 0.047 (0–3)	1.38 ± 0.111 (0–3)	1.60 ± 0.510 (0–3)	1.67 ± 0.044 (0–3)	1.62 ± 0.062 (0–3)	1.71 ± 0.064 (0–3)
	M	1.55 ± 0.095 (0–3)	1.56 ± 0.258 (0–3)	1.00 ± 0.258 (0–2)	1.59 ± 0.092 (0–3)	1.57 ± 0.123 (0–3)	1.54 ± 0.129 (0–3)
	p	0.044*	0.081	0.170	0.225	0.204	0.097
MQ	F	8.47 ± 0.280 (0–24)	8.09 ± 0.702 (0–18)	9.20 ± 1.20 (6–12)	8.40 ± 0.264 (0–24)	8.43 ± 0.369 (0–24)	8.35 ± 0.376 (0–24)
	M	9.07 ± 0.494 (0–24)	8.44 ± 1.458 (0–22)	9.17 ± 1.108 (5–12)	8.97 ± 0.491 (0–24)	8.93 ± 0.634 (0–22)	9.03 ± 0.694 (0–24)
	p	0.338	0.228	0.416	0.254	0.409	0.505
MGD^2	F	(0–1)	(0–1)	(0–1)	(0–1)	(0–1)	(0–1)
	M	(0–1)	(0–1)	(0–1)	(0–1)	(0–1)	(0–1)
	p	0.858	0.454	0.743	0.619	0.076	0.967

– Data are presented as the mean ± standard error and the range of values are given in parentheses.

– Asterisk (*) above bar denotes p ≤ 0.05, Chi-Square test.

– ¹ Osm was only assessed in 520 subjects in comparison with other clinical tests (n = 1090).

– ² TMH and MGD are dichotomous (0 and 1) and there is no mean± standard error.

– Abbreviations: Schirmer I test (ST), dry eye severity level (DESL), Osmolarity (Osm), tear film break-up time (TFBUT), ocular protection index (OPI), ocular surface staining (OSS, Oxford Grading Scheme), tear meniscus height (TMH), meibum expressibility (ME), meibum quality (MQ), and meibomian gland disease (MGD).

Discussion

This study revealed that patients with lower tear production levels measured by ST presented with more severe DED than patients with higher tear production. All the three defined cut-off values (≤ 5, ≤ 10, and ≤ 15 mm) gave rise to significant difference for DESL, Osm, TFBUT, OPI, OSS, and TMH, but not for ME, MQ, and MGD, in the six tear production level groups. There was no clear pattern in the gender difference of each group and significant difference was only observed for DESL (groups 2, 4, and 5), TFBUT (groups 2, 4, and 5), OPI (groups 2 and 6), and ME (group1).

An association between ST and other commonly used clinical DED tests has previously been published. For example, three out of four statistical analyses performed by Nichols et al.²⁰ revealed a significant relationship between the ST (cut-off ≤ 5 mm) and TFBUT (cut-off ≤ 10 s) in DED patients (n = 75) compared to other objective tests. These two parameters were significantly correlated (r = 0.40) and lower ST results were found to predict lower TFBUT in the stepwise, multivariate logistic regression models. Also, Wilcoxon rank sum analysis showed that subjects with an abnormal TFBUT had lower mean Schirmer scores.

Sullivan, et al.²¹ determined correlation coefficients between ST (cut-off < 7 mm) and other tests: Osm (0.05), TFBUT (0.08), fluorescein corneal staining (0.14), lissamine green conjunctival staining (0.13) and Bron/Foulks meibomian gland grading (0.05) in 344 subjects with (262) and without (82) DED. In order to compare their results with an independent data set, the authors used the data collected from a previously published study in Germany with 200 subjects (184 DED and 16 health controls).²² Accordingly, correlation coefficients were reported between ST and other clinical parameters, such as Osm (0.00), TFBUT (0.06), fluorescein corneal staining (0.03), lissamine green conjunctival staining (0.03), and Bron/Foulks meibomian gland grading (0.03). These results indicate that this test of aqueous deficiency can be utilized as a discriminator for some other objective tests.

Among all the parameters investigated in the current study, ST was unable to significantly discriminate only those related to meibomian gland function (ME, MQ and MGD). Patients with MGD have an unstable tear lipid layer but are generally considered to have normal tear production. This instability in the lipid layer may cause a decrease in tear volume, which in turn results in higher rates of evaporations and eventually damage to the ocular surface.²³ In contrast, Tung et al.²⁴ observed a different pattern in MGD wherein higher tear volume correlates with worse ocular surface disease. Thus, the authors hypothesized that changes in tear composition such as hyperosmolarity, lactoferrin concentration and inflammatory mediators may play a more important role in causing corneal epithelial damage in MGD than increased evaporation. This may also be the case in our study, as ST could not significantly determine meibomian gland parameters in DED patients. Therefore, measuring tear production may not be an efficient discriminator for ME, MQ, and MGD tests.

Historically, a gradual decrease in cut-off value for ST has been observed. For example, Schirmer,²⁵ De Roetth,²⁶ and Van Bijsterveld²⁷ used 15, 10, and 5.5 mm as normal cut-off values in their studies, respectively. The lack of well-defined cut-off values for the available DED tests such as ST complicates research making it difficult to compare results from different studies. The choice of cut-off value by researchers has been based on the purpose of their studies, for example, increased sensitivity for screening and specificity for therapeutic purposes. Irrespective of whether the cut-off value is defined as 5, 10, or 15 mm, the present study showed that ST is a robust discriminator for some clinical DED tests in a cohort of 1090 DED individuals.

For ST without anesthesia, the application of 10 mm cut-off value was considered sufficiently high not to overlook DED for screening purposes, but a proper value should be used to increase specificity in comparing inter-institutional data and evaluating efficacy of drugs even though sensitivity decreases to some extent.²⁸ However, Lee and Hyun²⁹ suggested that there is no available normal cut-off value for ST due to the high number of false positive or false negative values, leading to the ratio of 48.4% misdiagnosis in their study. They concluded that ST was not a reproducible test to differentiate DED patients (n = 15) from normal individuals (n = 110) and therefore analogized ST to the toss of a coin in the diagnosis of DED. In contrast, Van Bijsterveld’s (1969) cut-off value of 5.5 mm was suggested to be acceptable because of impossibility in ruling out both false positivity and false negativity in ST test due to its inaccuracy.^28,30,31 In the current study, a possible explanation for obtaining significant differences using all three defined cut-off values might be related to the sample size. We recruited 1090 subjects in contrast to other studies with a significantly lower number of individuals. Clearly, larger sample size increases statistical robustness.³² Thus, by increasing the number of subjects, the conclusions may change and researchers should be mindful of this when designing their studies.

In response to the limitations of ST, such as invasive nature, unavoidable variability, poor reproducibility, long testing time (i.e., 5 min), low sensitivity and specificity, alternative strategies have been proposed by researchers. A study evaluating the diagnostic usefulness of ST by using the Japanese diagnostic criteria on DED concluded that the combination of ST and TFBUT increased the diagnostic predictability.²⁸ Other authors prefer to use less invasive tests, for example, phenol red thread test³³ or noninvasive methods, such as thermography, keratography, and Hartmann-Shack wavefront sensor.^34–36 Nevertheless, based on our results, ST should be chosen as a clinical standard for measuring tear production until a more accurate test appears.

Conclusion

Schirmer test is a robust discriminator for DESL, Osm, TFBUT, OPI, OSS, and TMH, but not for ME, MQ, and MGD. Our results demonstrate that patients with lower tear production levels presented with more severe DED, irrespective of different cut-off levels (5, 10, or 15 mm). Interestingly, the differences in the mean values of DESL were minimal although statistically significant. Thus, the clinical value of different Schirmer levels appears to be limited. No clear pattern of sex differences of results were observed in the groups.

Acknowledgments

This work was supported by Oslo University Hospital, Norway, and the Norwegian Dry Eye Clinic, Oslo, Norway.

Disclosure statement

We declare the authors have no competing interests as defined by Current Eye Research, or other interests that might be perceived to influence the interpretation of the article. The authors alone are responsible for the content and writing of the paper.