External lacrimal punctum grading: reliability and interobserver variation.

PURPOSE To test the reliability and interobserver variation of external lacrimal punctum (ELP) grading. METHODS In a population-based epidemiologic study, ELP was visually graded based on slit lamp examination (grade 0 to 5), as described previously. Grades 0, 1, and 2 are for closed or stenotic ELP and grade 3 is for normal ELP (subjective grades). Grades 4 and 5 are for larger than normal (slit) ELP measured in millimeters by slit lamp beam (objective grades). Upper and lower ELP were graded by an oculoplastic surgeon and a senior ophthalmology resident. Both observers were masked to the reading of the other. Both observers were familiar to the grading system for past few years, although the oculoplastic surgeon had more experience in this regard. Patients with previous periocular surgery, medial lid mass obscuring the ELP, anatomically distorted punctum, and subjects with grade 4 and 5 ELP (objective grades) were excluded. The intraclass correlation coefficient test was used for the statistical analysis. RESULTS There were 632 eyes from 174 (55.10%) male and 142 (44.90%) female subjects. Subjects' ages ranged from 7 to 87 years (mean +/- SD = 54+/-17.56). The intraclass correlation coeficient test showed a positive and strong correlation between two observers for grading the upper (value=+0.87, 95% CI: 0.84 and 0.88, p=0.000) and lower (value=+0.91, 95% CI: 0.90 and 0.92, p=0.000) ELP. In general, the less experienced observer tended to overestimate both upper and lower ELP grading. CONCLUSIONS Reliability of the new ELP grading makes it a recommendable option for use in the clinic and comparing reports on ELP pathologies.