Abstract
Purpose :
Noncompliance during prospective studies can bias results and limit conclusions. The current study retrospectively investigated the relationship between patient characteristics and rates of noncompliance among subjects enrolled in interventional trials involving common causes of blindness.
Methods :
Demographic and longitudinal clinical information, including visual acuity (VA) and central retinal thicknesses (CRT) of 191 patients (202 eyes), were examined using Student’s T test, Pearson’s test and logistic regression. Subjects were enrolled in 1 of 4 investigator-initiated trials studying proliferative diabetic retinopathy, neovascular age-related macular degeneration, diabetic macular edema, and retinal vein occlusion, with a cumulative duration of 7-study years. Records were reviewed for evidence of noncompliance, defined as either missed visits (MV) or loss to follow-up (LTFU).
Results :
Among 191 patients, 77 subjects (40%) were noncompliant: 19 (9.9%) were LTFU, 28 (14%) had 1 MV, and 30 (16%) had >1 MV. 114 patients (60%) were completely compliant. Of 15 variables analyzed, age, baseline VA, baseline CRT, change in CRT, and diabetic status were statistically significantly correlated with noncompliance. Noncompliant patients were younger than compliant patients, with a mean age of 61 years (27-90) vs 67 years (25-93; p=0.007). Noncompliant patients had a higher mean baseline CRT, 394 μm (159-1034; p=0.002), and lower mean baseline VA, 64 ETDRS letters (7-94; p=0.028) compared to compliant patients, with a mean baseline CRT of 312 μm (149-1240) and VA of 69 letters (13-90). Noncompliant patients experienced a greater decrease in CRT of -130 μm (-792-63) vs compliant patients, who had a mean CRT change of -54 μm (-1103-298; p=0.002). Patients were more likely to be compliant if they were not diabetic (p<0.001). Diabetic status was the only significant risk factor in multivariate regression (p=0.006).
Conclusions :
The current analysis found that noncompliance was frequent among prospective trials, with 40% of subjects either LTFU or missing scheduled visits. Noncompliance was found to be associated with age, vision, anatomy and diabetic status. Appreciation of the demographic and clinical characteristics associated with compliance may allow implementation of approaches aimed at improving compliance within prospective clinical trials.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.