Dry Eye COncordance and Discordance Evaluator

How does DECODE work?

DECODE (Dry Eye COncordance and Discordance Evaluator) is a proof-of-concept framework designed to quantify the relationship between dry eye symptoms and clinical signs.

DECODE transforms symptom questionnaires and clinical signs to a common 0-to-1 severity scale, where 0 represents lower severity and 1 represents greater severity. Discordance is then calculated:

Discordance = normalized symptom severity − normalized sign severity

Positive values indicate symptom-predominant discordance (symptoms are more severe than signs), while negative values indicate sign-predominant discordance (signs are more severe than symptoms). Values closer to zero indicate greater symptom–sign concordance.

What does DECODE calculate?

DECODE provides results at three complementary levels:

• Overall discordance: summarizes the relationship between the available symptom questionnaires and clinical signs.
• Sign-specific discordance: compares symptom severity with each individual clinical sign.
• Domain-specific discordance: groups related measurements to evaluate specific clinical domains, including tear evaporation (TBUT/NIBUT), aqueous deficiency (Schirmer/TMH), ocular surface staining (corneal/conjunctival staining), LWE (horizontal length/sagittal width), and MGD (meibum quality/expressibility/meibomian gland loss).

How are multiple measurements handled?

By default, DECODE uses the mean of the available normalized symptom questionnaires and clinical signs to calculate overall discordance. Users may alternatively select the worst approach, in which the most severe available normalized symptom score and the most severe available normalized clinical sign are used.

When data from both eyes are available, DECODE uses the mean of both eyes by default. Users may alternatively select the worst eye approach. Eye-specific results can also be displayed separately when desired.

The calculation method selected by the user is recorded in the final report to facilitate transparency and reproducibility.

What information is required?

DECODE does not require every clinical measurement to be available. To calculate overall discordance, at least one symptom questionnaire and either TBUT or NIBUT must be entered. OSDI-6 is the recommended symptom questionnaire, according to DEWS-III. Additional available signs are incorporated into the corresponding overall, domain-specific, and sign-specific calculations.

Missing measurements are not treated as zero and are excluded from the corresponding calculations.

Individual patients and research datasets

DECODE can be used to evaluate an individual patient or to process an anonymous research dataset using the DECODE data template. In batch analysis, the same calculation rules are applied consistently across all eligible participants.

Important

DECODE is currently a pilot, proof-of-concept framework and has not yet been clinically validated as the optimal method for measuring symptom–sign discordance. The best approaches for combining measurements, selecting mean versus worst values, handling bilateral data, and interpreting the magnitude of discordance require validation in independent cohorts.

When using DECODE for research or publications, users should report the measurements included and the calculation settings used.