Screening utility of a rapid non-invasive dry eye assessment algorithm

Area of Science:

• Ophthalmology
• Diagnostic Accuracy
• Ocular Surface Disease

Background:

• Dry eye disease (DED) diagnosis can be complex and time-consuming.
• Current diagnostic methods may not be suitable for all clinical settings.
• There is a need for rapid, non-invasive DED assessment tools.

Purpose of the Study:

• To evaluate a rapid, non-invasive dry eye assessment algorithm for detecting DED.
• To assess the discriminative ability of the algorithm's components against TFOS DEWS II criteria.

Main Methods:

• A prospective diagnostic accuracy study included 235 participants.
• Evaluated symptomology (SANDE, OSDI, DEQ-5), tear film breakup time, tear osmolarity, staining, lipid layer, meibum expressibility, and tear meniscus height.
• Utilized the global evaluation component (SANDE, non-invasive tear film breakup time) and subtype classification component (lipid layer, tear meniscus height).

Main Results:

• All components of the algorithm demonstrated discriminative abilities significantly greater than chance (AUC > 0.80).
• The global evaluation component achieved 86% sensitivity and 94% specificity at optimal thresholds (SANDE ≥30, tear breakup time <10s).
• Positive and negative likelihood ratios were 15.0 and 0.15, respectively.

Conclusions:

• The abridged non-invasive dry eye assessment algorithm shows promise as a rapid screening instrument.
• It can be beneficial in resource or time-constrained clinical settings.
• The algorithm aids in the diagnostic process for dry eye disease based on TFOS DEWS II criteria.