BACKGROUND:Refractory non-infectious uveitis is a serious condition that leads to ocular complications and vision loss and requires effective systemic treatment to control disease. The effectiveness of long-term infliximab [IFX] in refractory non-infectious childhood uveitis and the impact of treatment adherence on disease control were evaluated. METHODS:Retrospective, single-center study between December 2002 and April 2016 of 27 children with refractory non-infectious uveitis [17 with juvenile idiopathic arthritis, JIA] treated with long-term IFX [9+ months]. Disease activity was assessed prior to and while on IFX using the Standardization of Uveitis Nomenclature [SUN]. Number of visits per year with active uveitis was analyzed by repeated measures logistic regression analysis from 2 years prior to IFX initiation or from onset of uveitis until most recent visit on IFX. Incomplete treatment adherence was assessed for each visit and defined as any deviance in corticosteroid use, prescribed infusion frequency, and/or follow-up examination frequency. RESULTS:Primary outcomes were sustained uveitic and systemic disease control prior to and during IFX treatment and the impact of incomplete adherence on uveitic disease control while on IFX. Secondary outcomes included corticosteroid and glaucoma medication requirement, ocular complications and need for surgical intervention. Mean age at IFX initiation was 10.4 ± 4.5 years; initial mean dose was 6.6 ± 2.2 mg/kg [and given at weeks 0, 2, 4 and q4 weeks thereafter for 93%]. Median duration on IFX was 35 [range 9-128] months. Prior to IFX, 14/27 patients had failed adalimumab ± methotrexate [MTX]; 21/27 failed MTX. IFX led to uveitis control in 89% and arthritis control in 76% (13/17). The odds ratio of having controlled disease after IFX was 4.1 (2.6, 6.4) compared to pre-treatment visits. Topical corticosteroids and glaucoma medications were statistically decreased (p = 0.007 right eye [OD], 0.003 left eye [OS] and p = 0.001 OD, p = 0.028 OS respectively). Incomplete adherence to treatment showed 10.3 times greater odds (7.1, 15.0) of having disease activity than full adherence. CONCLUSIONS:This study adds significantly to the IFX literature by documenting outstanding uveitis control with long-term IFX treatment in non-infectious pediatric uveitis patients. Higher dosage and shorter interval were utilized without adverse effects. Importantly, this is the first study, to our knowledge, to document the significant impact of treatment adherence on uveitis control.