Randomised trial of the clinical utility of an eyelid massage device for the management of meibomian gland dysfunction.

PURPOSE

To compare the single application and two week treatment effects of device-applied (Eyepeace) and manually-applied eyelid massage techniques, as an adjunct to warm compress therapy, on ocular surface and tear film parameters.

METHODS

Twenty participants (11 females, 9 males; mean age, 27 ± 11 years) with dry eye symptoms were recruited in a two week, investigator-masked, randomised, contralateral-eye trial. Following 10 min of warm compress therapy application (MGDRx EyeBag) on both eyes, eyelid massage therapy was applied to one eye (randomised) by device, and to the fellow eye by manual eyelid massage, once daily for 14 days. Ocular surface and tear film measurements were conducted at baseline, and 15 min post-application by a clinician, then again after 14 days of self-administered daily treatment at home.

RESULTS

Baseline clinical measurements did not differ between the treatment groups (all p > 0.05). Following two weeks of treatment, tear film lipid layer grade improved significantly with device massage (p = 0.008), and was marginally greater than manual massage by less than 1 grade (p = 0.03). Although immediate post-treatment improvements in tear film stability were observed in both groups (both p < 0.05), no significant long-term cumulative effects or inter-treatment differences in stability measures were detected (all p > 0.05). Visual acuity, tear meniscus height, conjunctival hyperaemia, ocular surface staining, and meibomian gland dropout did not change during the treatment period (all p > 0.05).

CONCLUSIONS

Two weeks of treatment with the eyelid massage device, as an adjunct to warm compress therapy, effected marginally greater improvements in tear film lipid layer thickness than the conventional manual technique, which were statistically but not clinically significant. Future parallel group trials with longer treatment periods and a greater range of disease severity are required.