eTable 1. Questionnaire used to determine vision preference value and treatment preferences.
eTable 2. Visual function scores of individuals with vitreomacular interface abnormality (VIA).
eTable 3. Mean visual function scores by type of vitreomacular interface abnormality (VIA) and clinical site.
eTable 4. Enthusiasm for treatment for each type of vitreomacular interface abnormality (VIA) in each clinical site.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Linz MO, Bressler NM, Chaikitmongkol V, et al. Vision Preference Value Scale and Patient Preferences in Choosing Therapy for Symptomatic Vitreomacular Interface Abnormality. JAMA Ophthalmol. 2018;136(6):658–664. doi:10.1001/jamaophthalmol.2018.1272
What preference value do individuals across 3 continents with vitreomacular interface abnormalities assign to their visual states, and what are their treatment preferences?
In this cross-sectional questionnaire study including 213 participants, with 1 indicating perfect health and 0 indicating death, the mean (SD) preference value was 0.76 (0.15), with no difference across continents identified. Most participants (171 [81.0%]) were enthusiastic about being treated; 150 (71.1%) were enthusiastic about vitrectomy and 120 (56.9%) about intravitreal injections.
These data suggest vitreomacular interface abnormalities affect preference-related quality of life similarly across continents at a level slightly better than reported for common retinal diseases, such as retinal vein occlusions; there may be a slight preference for surgical treatment over injections.
While symptomatic vitreomacular interface abnormalities (VIAs) are common, assessment of vision preference values and treatment preferences of these may guide treatment recommendations by physicians and influence third-party payers.
To determine preference values that individuals with VIA assign to their visual state and preferences of potential treatments.
Design, Setting, and Participants
In this cross-sectional one-time questionnaire study conducted between December 2015 and January 2017, 213 patients from tertiary care referral centers in Thailand, the United Kingdom, and the United States were studied. Patients with symptomatic VIA diagnosed within 1 year of data collection, visual acuity less than 20/20 OU, and symptoms ascribed to VIAs were included. Data were analyzed from January 2017 to November 2017.
Main Outcomes and Measures
The primary end points were overall mean preference value that individuals with VIA assigned to their visual state and patients’ preferences for potential treatments. Preference values were graded on a scale from 0 to 1, with 0 indicating death and 1 indicating perfect health with perfect vision.
Of the 213 included patients, 139 (65.3%) were women, and the mean (SD) age was 65.6 (7.7) years. Diagnoses included epiretinal membrane (n = 100 [46.9%]), macular hole (n = 99 [46.5%]), and vitreomacular traction (n = 14 [6.6%]). The mean (SD) vision preference value was 0.76 (0.15), without differences identified among the 3 VIA types. More participants were enthusiastic about vitrectomy (150 [71.1%]) compared with intravitreal injection (120 [56.9%]) (difference, 14.2%; 95% CI, 5.16-23.3; P = .002). Adjusted analyses showed enthusiasm for vitrectomy was associated with fellow eye visual acuity (odds ratio, 10.99; 95% CI, 2.01-59.97; P = .006) and better-seeing eye visual acuity (odds ratio, 0.03; 95% CI, 0.001-0.66; P = .03). Overall enthusiasm for treatment was associated with fellow eye visual acuity (odds ratio, 7.22; 95% CI, 1.29-40.40; P = .02). Overall, most participants (171 [81.0%]) were enthusiastic about surgery, injection, or both.
Conclusions and Relevance
Study participants reported similar preference values among 3 types of VIAs. The data suggest that most patients with these conditions would be enthusiastic about undergoing vitrectomy or an injection to treat it, likely because of the condition’s effect on visual functioning, although there may be a slight preference for vitrectomy at this time.
Vitreomacular interface abnormalities (VIAs), including epiretinal membrane (ERM), vitreomacular traction (VMT), and macular hole (MH), are common degenerative vitreoretinal disorders affecting aging populations in many countries around the world. Posterior vitreous detachment occurs naturally with aging and is generally asymptomatic, with the exception of floaters or flashes.1 An incomplete posterior vitreous detachment, with persisting attachments between the vitreous and retina, may cause traction leading to symptomatic VIA, depending on the orientation of traction.2 Visual acuity (VA) with VIA may vary, depending on severity of vitreous traction and area of affected macula. Vitreomacular interface abnormalities may also be associated with blurry vision, metamorphopsia, and central scotomas.
Vision preference values are validated methods to determine what value an individual places on visual function (VF) and the effect a specific disease process has on a patient’s perceived quality of life. Measurements of preference assign a value to the extent to which an individual prefers a specified health outcome.3 First validated by the Submacular Surgery Trials Research Group in 2004,4 the vision preference value scale determines preference values patients would assign to their vision. It is useful for determining preference values for ocular diseases that may be associated with low VA, as it does not require visual aids or in-person interviews, unlike other methods.4 Relative to other methods, the vision preference value scale is also less affected by variances in participant demographic characteristics, such as socioeconomic status and education.5 Compliance with follow-up, compliance with therapy, and attitudes toward therapies may be affected by preference values. To our knowledge, preference values of patients with symptomatic VIA are currently unknown.
For the past several decades, there have been only 2 options for treatment available to individuals with symptomatic VIAs: observation or surgery. Observation with “watchful waiting” is a preferable option for some patients, as VIAs can spontaneously resolve without treatment6 or the condition may not cause severe symptoms.7 Patients with more severe symptoms may undergo vitreoretinal surgery to remove ERM or VMT.8 Surgery generally has favorable visual outcomes with some resolution of visual symptoms; however, there are associated risks.9 Most patients experience a development or progression of cataracts; about half may require cataract surgery.7,8,10 Additionally, surgery can increase risk of retinal tear or detachment,9 irritation and discomfort,11 high intraocular pressure,12 vitreous hemorrhage,9 and formation of ERM.9
Recently, intravitreal injections have been used to treat certain types of VIAs. Pneumatic vitreolysis, injection of an air bubble or gas bubble into the vitreous cavity, may be useful in the treatment of VIA.13 Additionally, pharmacologic vitreolysis administered through intravitreal ocriplasmin injection has been approved by the US Food and Drug Administration and other regulatory agencies around the world for use in symptomatic vitreomacular adhesion.14,15 Phase 3 clinical trials have supported ocriplasmin’s efficacy in resolution of vitreomacular adhesion under certain conditions.16 Indications for ocriplasmin include vitreomacular adhesion with diameter less than 1500 μm, absence of ERM, phakic lens status, age younger than 65 years, and full-thickness macular hole less than 250 μm.16 Pharmacologic vitreolysis treatment may allow earlier intervention for VIAs, has a shorter recovery period, and does not carry increased likelihood of cataract formation compared with vitrectomy. Risks tend to be mild, commonly including floaters, photopsia, temporary acute visual loss, dyschromatopsia, subretinal fluid, eye pain, and conjunctival hemorrhage.17 Less commonly, ocriplasmin intravitreal injection can cause more severe risks, including endophthalmitis, retinal tear, and retinal detachment.17 Although much still needs to be understood about long-term safety and efficacy of ocriplasmin, to our knowledge, there is no currently available published information as to whether patients affected by VIA would preferentially choose intravitreal injection over vitreous surgery. Herein, we assess a vision preference value scale and patients’ preferences in choosing therapies for symptomatic VIA among patients from 3 regions of the world.
This study was approved by the Research Ethics Committee of Chiang Mai University, the National Research Ethics Service Committee, and the Johns Hopkins Medicine Institutional Review Board. All participants provided written informed consent.
Adult patients were included if they had a primary diagnosis of VIA, such as idiopathic VMT, ERM, or MH, within the past 1 year and had a VA less than 20/20 OU. The VIA had to be evident on spectral-domain optical coherence tomography and confirmed by a retinal specialist prior to enrollment in the study. Patients with an ERM had a qualifying central subfield thickness on spectral-domain optical coherence tomography beyond 2 SDs of the mean value among normal eyes or symptoms of metamorphopsia accompanied by wrinkling of the inner retina. Patients with a history of treatment for VIA or those with coexisting vision-threatening conditions (eg, dense cataract or diabetic macular edema in the VIA-affected eye) were excluded from this study. Participants were recruited during their scheduled standard clinic visit or by telephone within 6 months of their clinic visit. After giving informed consent, participants were interviewed by study team members using 2 questionnaires: first, a previously validated vision preference value score questionnaire (the Submacular Surgery Trials Vision Preference Value Scale version 24), including treatment enthusiasm questions not previously validated (eTable 1 in the Supplement), and then the previously validated 25-item National Eye Institute Visual Functioning Questionnaire version 2000 (NEI VFQ-25).18Enthusiasm, as used in the questions evaluating treatment enthusiasm, is an accepted term used to allude to a patient’s preference of treatment. The NEI VFQ-25 measures not only VF but also the effect of eye disease on the patient’s emotional health.5,18 Participants were queried about their treatment preference and current symptoms. Following the interviews, demographic data, medical and ocular history, treatment history, ocular examination data, and imaging information were collected from medical records.
The primary end points were overall mean preference values that individuals with VIA assigned to their visual state and patients’ preferences for potential treatments. Secondary end points included factors associated with preference values (such as affected eye VA, better-seeing eye VA, and clinical site in Thailand, the United Kingdom, or the United States), NEI VFQ-25 composite and subscale scores, and factors associated with treatment preferences.
The primary outcome variable was defined as the overall preference value assigned by study participants with VIA to their visual state. Secondary outcomes included overall level of enthusiasm for potential treatments of VIA and how patient characteristics were related to level of enthusiasm. Preference values were graded on a scale from 0 to 1, with 0 indicating death and 1 indicating perfect health with perfect vision. Visual acuity was analyzed as a continuous variable using converted logMAR equivalent values of habitual VA. Counting fingers was given a value of −2.0, and hand motion was given a value of −3.0.19 For patients with VIAs in both eyes, the study eye VA was determined to be the lower VA of the 2, and these patients were excluded in analyses related to fellow eye VA. Patients with a diagnosis of more than 1 type of VIA in 1 eye (eg, MH and ERM) were assigned a single diagnosis category; MH took priority, followed by ERM, and then VMT. Participants who did not identify as white or Asian in this study, including a smaller number of participants who identified as African American, black, other, or declined to identify, were grouped into 1 category for ethnicity to allow for analysis.
Enthusiasm for surgical or injection treatment was classified as either enthusiastic if participants answered “moderately enthusiastic” or “very enthusiastic” for that particular treatment or not enthusiastic if participants answered “not enthusiastic” or “slightly enthusiastic” for that particular treatment. Overall enthusiasm was defined as participants answering “moderately enthusiastic” or “very enthusiastic” for at least 1 invasive treatment (ie, surgery or injection).
Unadjusted regression analysis was used to assess the association of preference value with each patient characteristic, such as age, race/ethnicity, sex, clinical site, symptoms (ie, blurry vision, metamorphopsia, central scotoma, and floaters), study eye phakic status, disease duration, VA (study eye, fellow eye, and better-seeing eye), NEI VFQ-25 composite score, VIA diagnosis (study eye), and bilaterality. Results from the unadjusted analyses were used to build adjusted regression models. The final models were determined using the backward variable selection method based on Akaike Information Criterion. Age, sex, and clinical site were adjusted for in the final model. Coefficients, 95% confidence intervals, and P values are presented.
Unadjusted logistic regression models were used to examine the association of enthusiasm for surgery, enthusiasm for injection, and overall enthusiasm for treatment with the patient’s demographic and clinical characteristics. These variables then were included into adjusted logistic regression models with enthusiasm for each treatment and overall treatment enthusiasm as the dependent variable for each model. To make the 3 final models uniform, all the variables that were chosen from each of the dependent variables by the backward variable selection method based on Akaike Information Criterion were included. Age, sex, and clinical site were adjusted for in the final model. Odds ratios, 95% confidence intervals, and P values are presented.
Wald χ2 tests were used to test the significance of individual coefficients in the model. P values less than .05 were considered statistically significant, and all P values were 2-tailed. Data analysis was performed using STATA version 14.1 (StataCorp).
A total of 213 patients were enrolled between December 2015 and January 2017; diagnoses included ERM (n = 100 [46.9%]), MH (n = 99 [46.5%]), and VMT (n = 14 [6.6%]). Table 1 shows overall demographic and clinical data for the study participant population.
The mean (SD) vision preference value among 209 participants from the 3 sites was 0.76 (0.15), and no differences were identified among the 3 VIA types. No substantial differences in mean (SD) preference value were detected among the sites (Thailand, 0.75 [0.13]; United Kingdom, 0.73 [0.18]; and United States, 0.80 [0.13]). The range of scores was 0.1 to 1. Visual function was graded on a scale from 0 to 100, with 0 indicating the worst possible score and 100 indicating the best possible score. The mean (SD) composite score for VF was 80.69 (14.07), and the median was 84.13. Visual function composite and subscale scores are given in eTable 2 in the Supplement. Visual function subscales varied slightly based on type of VIA and clinical site (eTable 3 in the Supplement). The mean (SD) composite VFQ scores varied slightly for each type of VIA, including VMT (85.37 [12.83]), ERM (81.16 [13.36]), and MH (79.59 [14.88]). The number of participants with VIA at each level of enthusiasm for surgery or injection, subdivided by diagnosis type and clinical site, is presented in eTable 4 in the Supplement. More participants were enthusiastic about vitrectomy (150 [71.1%]) compared with intravitreal injection (120 [56.9%]) (difference, 14.2%; 95% CI, 5.16-23.3; P = .002). Overall, 171 of 211 participants (81.0%) were enthusiastic about surgery, injection, or both compared with 40 participants (19.0%) who were not enthusiastic about either treatment (difference, 62.0%; 95% CI, 54.6-69.4; P < .001).
In unadjusted analyses, preference values were associated with VFQ-25 composite score. Adjusting for potential confounders, preference values were associated with VFQ-25 composite score (Table 2). In unadjusted analyses, enthusiasm for surgical treatment was associated with study eye VA (odds ratio [OR], 0.22; 95% CI, 0.09-0.52), better-seeing eye VA (OR, 0.09; 95% CI, 0.02-0.41), VFQ-25 composite score (OR, 0.97; 95% CI, 0.94-1.00), disease duration (OR, 0.92; 95% CI, 0.86-0.99), Asian race (OR, 4.02; 95% CI, 2.07-7.83; reference, white race), clinical site (Thailand vs United Kingdom: OR, 3.49; 95% CI, 1.60-7.61; Thailand vs United States: OR, 0.17; 95% CI, 0.08-0.37), and blurriness (OR, 3.14; 95% CI, 1.37-7.17). After adjusting for potential confounders, enthusiasm for surgical treatment was associated with better-seeing eye VA and fellow eye VA (Table 3). Based on unadjusted analyses, enthusiasm for treatment overall was associated with study eye VA (OR, 0.26; 95% CI, 0.09-0.71), Asian race (OR, 5.27; 95% CI, 2.39-11.65), and clinical site (United States vs United Kingdom: OR, 0.32; 95% CI, 0.13-0.76; Thailand vs United States: OR, 0.15; 95% CI, 0.06-0.36). Enthusiasm for treatment overall was associated with fellow eye VA in adjusted analysis (Table 4). Enthusiasm for intravitreal injection was associated with Asian race (OR, 1.80; 95% CI, 1.01-3.20) in unadjusted analyses and with black, other, or unspecified race (OR, 11.54; 95% CI, 1.41-94.68) in unadjusted analyses and adjusted analysis (Table 5).
In our study, participants with VIA among the 3 sites assigned a preference value of 0.76 on average to their visual state value, on a scale from 0 (death) to 1 (perfect health with perfect vision). Preference value did not seem to vary by clinical site or by VIA type, suggesting that VIAs have a similar effect on vision preference value independent of diagnosis. The association between preference value and NEI VFQ-25 composite score suggests there are other factors not included in this study affecting preference value. The mean VFQ-25 composite score was 80.7 and the median composite score was 84.1, suggesting that VIA moderately affects VF. The subscale scores indicate that symptomatic VIA can negatively affect driving ability, mental health, near vision, distance vision, and role performance (eTable 2 in the Supplement). When participants were asked about their treatment preferences, although the information given to participants about the injection during the interview indicated more favorable outcomes for VA and less associated discomfort, more participants reported being enthusiastic about vitrectomy (71.1%) than intravitreal injection (56.9%). The data suggest that most patients with VIA are enthusiastic about undergoing vitrectomy or an injection to treat it, likely because of the condition’s effect on VF, and there may be a slight preference for vitrectomy.
The association of fellow eye VA with both enthusiasm for surgical treatment and enthusiasm for treatment overall may be due to the tendency for the eye with better VF to compensate for the eye with poor VF.20 Alternatively, a fellow eye with poor VF may contribute to worse vision, prompting a patient to prefer treatment over observation. In most study participants, the better-seeing eye was the fellow eye; however, this was not always the case. This may explain why fellow eye VA had a greater association with enthusiasm for treatment than better-seeing eye VA.
This study has several limitations. All 3 sites were tertiary care sites, and patients may have primarily been referred for consideration for treatment given their severity of VIA and thus may already have undergone a period of observation with watchful waiting for VMT release. Some participants discussed the possibility of surgery with a retinal specialist prior to the interview and may be more comfortable with choosing surgical intervention if their physician recommended it. Patients also tend to have a greater background knowledge of surgery than intravitreal injections. These factors may bias results by depressing vision preference value and increasing enthusiasm for treatment, specifically surgical treatment. The wording of the questionnaires, the participants’ moods or personalities,21 and the use of multiple interviewers across the study may have also affected patient response. Although it is likely that some participants were good candidates for intravitreal injection, importantly, not all participants in this study were eligible for treatment with intravitreal injection. Our study population included individuals with nonsevere cases of VIA that may not necessitate treatment as well as individuals with advanced VIAs (eg, MH ≥400 μm and MH with ERM present) who are not good candidates for intravitreal injection.16 There was a small sample size of participants with VMT, limiting generalizability of the results to this population. Because the study inclusion criteria only specified that the primary cause of vision loss in the study eye had to be due to VIA, some participants may have other eye diseases with mild effect on vision in the study eye or may have disease in their fellow eye that is more substantial. This may bias mean preference values, NEI VFQ-25 composite scores, and NEI VFQ-25 subscale scores to be lower than if we selected participants whose only cause of vision loss was VIA. It is unknown how MH size or the existence of more than 1 VIA in the study eye, such as coexisting MH and ERM, affected responses from study participants.
Previous studies have determined the preference values of other ocular conditions, including branch vein occlusion (0.65),22 central vein occlusion (0.65),22 preoperative cataract (0.68),23 and subfoveal choroidal neovascularization (0.64).4 In perspective, the average preference value of patients with VIA was higher than the average preference value assigned to each of these ocular conditions. Other than factors described above contributing to this value, a higher preference value among patients with VIA is likely because these other conditions cause relatively more severe vision loss and tend to be long term. Additionally, these other studies included patients who had the disease for more than 1 year. Similar to our findings, the VFQ-25 composite score was also the only variable associated with preference value in patients with choroidal neovascularization.4
Considering that there were no differences identified in treatment preferences among the 3 sites and that the United Kingdom clinical site covered patients’ treatment costs, finances were unlikely a factor in patients’ treatment preference. Additionally, preference for surgery over intravitreal injection may be due to lingering anxiety about an injection to the eye. Anxiety associated with intravitreal injections to treat other eye diseases has been reported owing to concerns over going blind, treatment effectiveness, the thought of an injection to the eye, and fear of the unknown.22,24,25 Our results vary from a previous study26 that showed that treatment-naive patients with diabetic retinopathy preferred intravitreal injections over laser treatment, although this may be due to the difference in surgical treatment for each disease state, which would affect benefits and risks shared with the participant.
In summary, this study provides important information regarding patients’ preference for VIA treatment. These data may be useful in informing future research investigating prevention and treatment of VIAs as well as potentially influencing third-party payers in their understanding of how much a patient with these conditions in different parts of the world would be willing to pay for treatments relative to other disease states. These results may help physicians better understand the effect that VIAs may have on a patient’s life and choose an appropriate treatment plan.
Accepted for Publication: March 9, 2018.
Corresponding Author: Adrienne W. Scott, MD, Retina Division, Wilmer Eye Institute, Johns Hopkins University School of Medicine, 600 N Wolfe St, Maumenee 719, Baltimore, MD 21287 (firstname.lastname@example.org).
Published Online: May 10, 2018. doi:10.1001/jamaophthalmol.2018.1272
Author Contributions: Ms Linz and Dr Scott had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Bressler, Chaikitmongkol, Patikulsila, Choovuthayakorn, Scott.
Acquisition, analysis, or interpretation of data: Linz, Bressler, Chaikitmongkol, Sivaprasad, Watanachai, Kunavisarut, Menon, Tadarati, Delalíbera Pacheco, Sanyal, Scott.
Drafting of the manuscript: Linz, Chaikitmongkol, Watanachai, Kunavisarut, Sanyal, Scott.
Critical revision of the manuscript for important intellectual content: Bressler, Sivaprasad, Patikulsila, Choovuthayakorn, Menon, Tadarati, Delalíbera Pacheco, Sanyal, Scott.
Statistical analysis: Sanyal.
Obtained funding: Bressler, Patikulsila, Choovuthayakorn, Scott.
Administrative, technical, or material support: Bressler, Chaikitmongkol, Sivaprasad, Patikulsila, Choovuthayakorn, Watanachai, Kunavisarut, Tadarati, Delalíbera Pacheco, Scott.
Study supervision: Bressler.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Drs Bressler, Chaikitmongkol, Sivaprasad, Patikulsila, Choovuthayakorn, Watanachai, Kunavisarut, and Scott received grants from ThromboGenics during the conduct of the study. Dr Chaikitmongkol has received grants from Bayer and personal fees from Novartis and Allergan. Dr Sivaprasad has received grants from Bayer, Boehringer Ingelheim, and Roche; personal fees from Novartis, Bayer, Heidelberg Engineering, and Boehringer Ingelheim; and nonfinancial support from Bayer. Dr Patikulsila has received personal fees from Bayer and Novartis as well as nonfinancial support from Alcon, Bayer, and Novartis. Drs Choovuthayakorn and Watanachai have received nonfinancial support from Alcon, Allergan, Bayer, and Novartis. Dr Kunavisarut has received nonfinancial support from Novartis and Allergan. Dr Scott is a consultant for Allergan. No other disclosures were reported.
Funding/Support: This study was supported by ThromboGenics, and the biostatistician who performed data analysis was funded by core grant EY01765 from the Wilmer Biostatistics Consulting Center.
Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. The sponsor was allowed to review the manuscript prior to submission for publication, and sponsor feedback was considered, though was not mandatory.
Disclaimer: Dr Bressler is the Editor of JAMA Ophthalmology. He was not involved in the evaluation or decision to accept this article for publication.
Create a personal account or sign in to: