Participants and study design
This cross-sectional study comprises patients with clinical diagnosis of AMD, DME or RVO in at least one eye, receiving intravitreal treatment with anti-vascular endothelial growth factor (VEGF) in the period between August and December 2018. Adult Danish-speaking patients of both genders without any known severe mental disorders or dementia were eligible for participation. We explained the nature of the study to all patients and verbal informed consent was obtained prior to their participation.
Patients were selected on a first-come basis, as the interviewer included the first patient arriving at the department from when the interviewer was available. The study complies with the tenets stated in the Declaration of Helsinki. Since the study includes data from a questionnaire, it is by Danish law (Komitéloven § 14, stk. 2) exempted from registration at the Regional Research Ethics Committee.
Patients were interviewed regarding education level, self-perceived health, comorbidities, and postal code. Education levels were based on the highest achieved level of education, defined as described by the Danish Ministry of Education. The levels of education were: primary school, vocational, short-cycle higher education (approximately two years at Academy profession programmes), medium-cycle higher education (approximately 3.5 years at Professional Bachelor’s programme), and long-cycle higher education (approximately 5 years at University programme). 
Postal codes were classified by municipalities: peripheral, rural, intermediate, or urban as suggested by the Danish Ministry of Food, Agriculture and Fisheries. 
Best corrected visual acuity (VA) was measured based on the Early Treatment Diabetic Retinopathy Study (EDTRS)-protocol by trained nurses. The patient is scored by how many letters are correctly identified, after reading from the top of the chart. Patients read down the chart until a minimum of three letters cannot be read. Further, we converted the ETDRS value into the logarithm of minimum angle of resolution (LogMAR) from each patient.
Comorbidities were converted based on the Charlson’s Comorbidity Index (CCI). CCI applies comorbidities into four categories; none (0), low (1-2), medium (3-4) or high (>4). [12, 13] Diabetic maculopathy was not registered as a diabetic complication in this index, to better assess other disorders besides the known retinal disease.
Health literacy assessment
The European Health Literacy Survey Questionnaire (HLS-EU-Q47) was developed to measure self-perceived health literacy across Europe[14, 15] and was translated into Danish and validated by Maindal et al.  The short form called HLS-EU-Q16 consists of 16 items focusing on four health literacy dimensions reflecting perceived ease or difficulty in an individual’s ability to obtain, understand, process, and apply health information.
A Danish-speaking physician or nurse administered the survey and patients got the option of having the survey read aloud to them accommodating patients at all vision and literacy levels.
The responses were dichotomized with “very easy” and “easy” given a score of 1 and “difficult” and “very difficult” given a score of 0.
The scores of the patients were then divided into three qualitative categories of health literacy: ‘inadequate’ (1-8), ‘problematic’ (9-12), ‘sufficient’ (13-16).
Statistical analysis were performed using the SPSS version 24 (IBM Corporation, Armonk, NY, USA). Normally distributed data was presented in mean and standard deviation (SD) and compared using parametrical tests. Not normally distributed data were presented using median and interquartile range (IQR) and compared using non-parametrical tests. We used Spearman rank correlation to assess possible correlations. Categorical data were tested using χ2 test, unless sample size was small (<5 cases) in which case Fisher’s exact test was used.