Ocular and oral grading of mucous membrane pemphigoid.

Background: A variety of methods have been described for grading ocular mucous membrane pemphigoid (MMP), each with their own limitations. In contrast, there are no reported grading systems for involvement of the oral mucosa. We wished to evaluate two ocular (one established and one proposed) and an oral mucosal grading system for MMP.

Methods: Patients with MMP were assessed by three ophthalmologists and two oral medicine physicians. Ocular disease was graded using the system described by Rowsey and a proposed system based on measurement of vertical depth and horizontal width measured from the bulbar conjunctival aspect. Oral assessment used a 'mucosal disease severity score' originally described for lichen planus, in which 17 areas of the mouth are scored for involvement, together with a pain score. Levels of agreement were evaluated using Fleiss' Kappa Statistic (k).

Results: Forty-four patients with MMP encompassing mild to severe disease were included. Good levels of agreement were observed between observers for both vertical (k:0.86) (upper 95% CI: 1.03 mm) and horizontal (k:0.80) (upper 95% CI: 3.01 mm) involvement for the proposed ocular system and the Rowsey system (k: 0.83) (upper 95% confidence interval: 3.19 mm). There was a high coefficient of determination (R(2)) between the ocular grading systems (0.81, p < 0.01). Oral grading showed excellent levels of agreement (k: 0.71) between observers. There was no significant association between the severity of oral and ocular disease using described grading systems.

Conclusions: The proposed grading systems for both oral and ocular involvement in MMP are easy to use, and show good agreement between observers. The proposed ocular system correlates well with a currently used system, and overcomes some of the difficulties encountered with existing systems. For the individual patient, changes greater than 1.5 mm (vertical) and 3 mm (horizontal) are significant. This may increase our ability to detect change or disease progression. Although the risk of ocular involvement in patients with only oral involvement has been demonstrated, the severity of oral and ocular disease are not well-correlated, due in part to an absence of an ocular disease activity score.