Visual Acuity Screening Among Asymptomatic Older Adults

The US Preventive Services Task Force (USPSTF) is a critical contributor in translating research into practical recommendations for policy makers, payers, care professionals, and the public. By using techniques grounded in best-practice, evidence-based medicine, the USPSTF’s Evidence Reports and Recommendation Statements for screening seek to provide clear insights into the state of documented knowledge and associated implications.

As reported in this issue of JAMA, based on an updated review of the evidence on screening for visual acuity impairment associated with uncorrected refractive error, cataracts, and age-related macular degeneration (AMD) in adults 65 years or older in the primary care setting, the USPSTF concluded that “current evidence is insufficient to assess the balance of benefits and harms of screening for impaired visual acuity in older adults. (I statement)”¹ The updated Evidence Report on visual acuity screening² and the USPSTF Recommendation Statement¹ illustrate the strengths of the USPSTF approach.

First, the assessment is explicitly focused on a portion of the high-risk population that might significantly benefit from some form of vision screening short of a comprehensive eye examination: (1) individuals 65 years and older (2) who have not sought care for evaluation of vision problems and (3) who do not have any symptoms or known impairments in visual acuity. Those excluded are either already in the care system or have both the means (Medicare insurance for those 65 years and older) and motivation (symptoms or known impairment) to be more likely to seek care. Sloan et al³ found that 85% of Medicare beneficiaries in the Health and Retirement Study population saw a physician (not limited to eye care) within 2 years of eligibility, suggesting that most—but not all—patients could overcome barriers to utilization. Existing data also show that approximately half of those 65 years and older report having had an eye examination in the prior year and that those without vision or eye problems are much less likely to have seen an eye care professional.⁴ Thus, the report clearly delineates the population for whom vision screening in a non–eye care setting might be very relevant.

Second, the Evidence Report clearly assesses the value or benefit of treatment of diseases that screening should detect: cataract, refractive error, and AMD. The report details the extensive body of data that demonstrate the value of vision care along important outcome dimensions for these 3 conditions. It is also known that more regular eye examinations in the adult population are associated with a decreased risk of developing new-onset limitations in instrumental activities of daily living.⁵ Further, the rate of self-reported visual impairment among older adults in the United States has declined by 50% in the last 20 years with the advent of extensive implementation of therapies for visually disabling eye conditions.⁶

Third, the analysis explicitly reviews the limited data that have been published about the harms associated with the screening process or a false-positive result. Given that the result of a positive screening test is a comprehensive eye examination by an eye care professional, it is not surprising that there are few direct data regarding the risks of an eye examination done by an ophthalmologist or optometrist.

Fourth, within a defined population, the report addresses a specific intervention for the detection of a specific set of conditions: that of visual acuity for the detection of uncorrected refractive error, cataract, and AMD. However, visual acuity alone (at distance) is only modestly related to the detection by a comprehensive eye examination for the presence of these 3 diseases.¹ In addition, visual acuity is poorly related to other assessments of visual function such as screening questions or other tests, and the examined approaches were not all well suited to the detection of the 3 eye conditions of interest.¹ As such, visual acuity was included as but one of a group of measures by a working group convened by the Centers for Disease Control and Prevention on the development of a surveillance system for visual impairment for the United States.⁷

Fifth, as with all USPSTF efforts, the Evidence Report suggests clear areas of research to fill in the data gaps. Among these, “Well-designed studies in primary care settings are needed to identify optimal methods for vision screening and to develop effective strategies for linking older adults with impaired visual acuity to appropriate care.”² The report also notes that studies are needed to determine the diagnostic accuracy and utility of alternative screening modalities that can supplement or replace current methods of screening, such as the Snellen visual eye chart.²

The strengths of the USPSTF process are also the same attributes that can create difficulties in the view of some patients, clinicians, and policy makers, including in carrying out the recommended studies. First, the high degree of methodological rigor makes it challenging to complete randomized clinical trials involving older adults who, at entry, do not have a known eye condition or visual acuity impairment, have not sought eye care, and are asymptomatic. Under intent-to-treat analyses, study participants in the control group of no screening could well develop conditions or symptoms that bring them to eye care, reducing the benefit differential (as hypothesized) of the screening group. Furthermore, older adults who are not in the eye care system may be more difficult to convince to obtain a comprehensive eye examination, even with a positive screening result.^1,8 Moreover, while those without symptoms and prior eye care visits may be assumed to be least likely to have conditions to be detected by screening, it is the disadvantaged populations that may be less likely to obtain care and to have greater disease burden, symptomatic or not.^9,10

Second, the specific nature of the questions asked does not allow clinicians to draw conclusions about the value of other questions, such as whether a comprehensive eye examination, either on entry to Medicare or potentially on a specified interval, might be appropriate. As noted in the report, the American Optometric Association recommends an annual eye examination after age 60 years, and the American Academy of Ophthalmology recommends a comprehensive examination every 1 to 2 years for adults 65 years or older who do not have risk factors. Data indicate there may be merit to this approach; incremental cost-effectiveness analysis with Monte Carlo simulation by Rein et al¹¹ suggested that an initial dilated eye examination yielded an estimated cost of $12 000 per quality-adjusted life-year gained. In addition, as the defined population ages, the underlying prevalence of disease increases markedly, increasing the positive predictive values of any adequate screening approach.

Third, the lag time for the conduct of appropriate studies may not keep pace with changes in current health delivery approaches. The combination of the rise of patient-centered care, empowered by the growth of digital medicine through online resources, mobile applications on smartphones, tablets, and computers, and data capture in electronic records (and subsequent analyses), portends a potentially very different future for health care and health maintenance.¹² It is likely that vision and other screening activities in the future, to the extent they occur, will not be performed in just the primary care setting, but widely distributed in homes, community centers, and other settings. Already, many apps and e-health approaches exist to assess vision, visual functioning, and even aspects of the comprehensive eye examination; with time, the performance specifications of these and other approaches will improve to serve as appropriate screening tools.^13,14 An underappreciated factor that is likely to push any assessment of vision out of the traditional primary care setting is the time required to perform visual acuity testing in the already overburdened primary care encounter.¹⁵

To the extent these changes occur, they may well serve to change the nature of the discussion around the importance of ratings for preventive services, which has engendered so much controversy in the past. Today, screenings that are “recommended” by an authoritative source such as the USPSTF are mandated to be covered under health insurance. In a fee-for-service environment, this represents added revenue to clinicians and health care organizations. In a future of population health and a focus on quality, such measures will likely be part of a quality report card, providing impetus for plans to include means of accomplishing the goal. As with diabetic retinopathy screening, telehealth or other approaches that create new pathways for such screening to occur will likely be seen as effective and economical methods to meet growing patient demand.¹⁶ Instead of increased “revenue,” health system organizational concerns about “costs” will help drive changes in screening. With the reduced (marginal) costs per unit of service associated with these innovative approaches, expanding screening to areas of concern of empowered patients and their families may become an attractive advantage for plans and clinicians. In other words, screening services with intuitive appeal to patients and some practitioners may be facilitated by these technological means, reducing the friction between recommendations of insufficient evidence and patient preferences and professional judgments. What will likely be unchanged is the importance of the contributions of the USPSTF to the translation of research into practical recommendations.