Capabilities and Limitations of Student-Led Free Vision Screening Programs in the United States

Abstract

Purpose: The Consortium of Student-Led Eye Clinics (CSLEC), founded in 2021, administered a comprehensive survey to document the types of services, most common diagnoses, and follow-up care protocols offered by student-led free vision screening programs (SLFVSP) in the United States.

Methods: An 81-question institutional review board (IRB)-approved survey was administered to student-led vision screening eye clinics from October 1, 2022 to February 24, 2023.

Results: Sixteen SLFVSPs were included in the final analysis, of which 81% (n = 13) conducted variations of fundoscopic examinations and 75% (n = 12) measured intraocular pressure. Cataracts and diabetic retinopathy were reported as the most frequent diagnoses by the majority of SLFVSPs (n = 9, 56%); non-mobile SLFVSPs more commonly reported cataract as a frequent diagnosis (P < 0.05). Most patients screened at participating programs were uninsured or met federal poverty guidelines. Prescription glasses were offered by 56% of the programs (n = 9). SLFVSPs that directly scheduled follow-up appointments reported higher attendance rates (66.5%) than those that only sent referrals (20%). Transportation was the most cited barrier for follow-up appointment attendance.

Conclusions: SLFVSPs, one community vision screening initiative subtype, vary significantly in scope and capabilities of identifying vision threatening disease. The follow-up infrastructure is not uniformly robust and represents a key target for improving care delivery to at-risk populations.

Translational Relevance: The CSLEC aims to develop a consensus-based standardization for the scope of screening services, offer guidelines for diagnostic criteria, promote real-time data stewardship, and identify means to improve follow-up care mechanisms in member communities.

Introduction

Community vision screening programs in the United States address the access and cost barriers of utilizing routine eye care to identify and treat preventable causes of blindness, such as diabetic retinopathy and glaucoma.¹ Student-led free vision screening programs (SLFVSP) represent one component of safety-net eye care by offering routine vision services to un- and under-insured populations. The Consortium of Student-Led Eye Clinics (CSLEC) was launched in 2021 to promote collaboration among SLFVSPs to identify and address mutual challenges in preventative eye care service delivery. The operational and service delivery characteristics of vision screening programs affiliated with primary care Student Run Free Clinics (SRFCs), have been preliminarily characterized by Okaka et al.² and by some CSLEC members individually.^3–6 

To standardize operations and optimize service delivery, the CSLEC appraised service delivery capabilities and limitations of SLFVSPs in the United States. Although there is great heterogeneity among SLFVSP service models, one notable differentiator centers on the capability of offering vision screening services in more than one location. In this vein, SLFVSPs with the flexibility to host vision screening services in various spaces, (like churches, community centers, or libraries) across a defined geographic territory are distinct from a program that exclusively offers services in one, predetermined location. Whereas theoretical advantages can be assigned to either model, it is uncertain what differences, if any, exist between screening services that self-identify as offering mobile services versus those that do not, or non-mobile screening services. Comparing performance outcomes and capabilities, therefore, between these two SLFVSP service delivery models may offer preliminary data for new and burgeoning programs to decide on an operational direction. 

Understanding the baseline characteristics of SLFVSPs is necessary for optimizing service delivery and improving workflow strategies that can render these vision screening programs as legitimate bridges to appropriate vision care utilization in un- and under-insured populations. The purpose of this study is to further identify strengths and weaknesses at the operational level to support recommendations for improvements to eye care delivery. 

Methods

A survey was developed by a team of cross-institutional CSLEC members, including medical students and ophthalmology faculty, to distribute to leaders of SLFVSPs. The questionnaire included elements from a survey of American Medical College (AAMC) member institutions.⁷ The final version of the survey contained 81 questions, including open-ended free response, multiple choice, and yes/no formats. Institutional review board approval was obtained at Indiana University and all affiliate research institutions, including the University of Pittsburgh School of Medicine, University of Cincinnati College of Medicine, Medical College of Wisconsin School of Medicine, and Johns Hopkins University School of Medicine. The selection of questions presented in this analysis are depicted in the Supplementary Figure. 

The survey was communicated to potential participants through email via the Association for University Professors of Ophthalmology (AUPO)’s Medical Student Educators Listserv of approximately 133 US ophthalmology programs and as a general announcement on the Journal for Student-Run Clinics (JSRC) website. In addition, from an initial list of ophthalmology residency programs obtained from Doximity, a web search was conducted to identify programs that ran an SLFVSP and then develop a personalized contact list of associated student and faculty leaders. These efforts were supplemented by informal communications streams that occurred between medical students and faculty affiliated with the CSLEC to other SLFVSPs that may have not been initially found in this web-search step. 

Personal invitations were sent to student-led vision screening program contacts from 42 institutions via email; 4 rounds of appeals were made to non-respondent contacts over the course of the study period. Despite demonstrating an online presence, it was unclear how many SLFVSPs were truly operational; at the same time, it is also possible that SLFVSPs could be operational without an online presence. 

The survey was distributed by email from October 1, 2022 to February 24, 2023, using a secure online Google Form with results only accessible to the CSLEC research team. Questions regarding respondent institutional affiliation and contact information were included to ensure only one unique response per program and to facilitate future contact for CSLEC quality improvement initiatives. 

The collected data were systematically analyzed, including tabulation of multiple choice and binomial yes/no responses and extraction of key words and phrases of open-ended free responses. Questions that included quantitative responses that were binned were extracted using the bin median. Subgroup analyses of numerical results were conducted with the Mann Whitney U tests assuming unequal subgroup variances. Binomial questions were analyzed using Fisher's Exact Test. Mobile and non-mobile programs were identified in the analysis by self-reported characterization. All data extraction and analyses were cross-checked by at least one additional member of the research team to ensure the accuracy of the results. 

Results

Clinic Respondent Characteristics

There were 23 responses in total, of which 4 were submitted with less than 5% of the survey filled out. Two of the responses were duplicates for the same clinic. One institution hosted two distinct SLFVSPs and were therefore considered as independent entities. Of these respondents, there were 13 SLFVSPs affiliated with the CSLEC represented, including 2 institutions that were yet to be operational, such that we included 93% (13/14) of complete SLFVSP responses and 92% (12/13) of complete responses from CSLEC members. The final analysis included 16 SLFVSPs, from 15 institutions, of which 10 were CSLEC members. 

The geographic locations of the institutions with operational SLFVSPs included 33.3% in the Northeast (n = 5), 26.7% in the Midwest (n = 4), 20% in the West (n = 3), and 20% in the Southeast (n = 3). Among operational CSLEC institutions (n = 11), 33.3% were in the Midwest (n = 4), 33.3% were in the Northeast (n = 4), 8.3% were in the West (n = 1), and 8.3% were in the Southeast (n = 1); the 2 non-operational, CSLEC-member SLFVSPs were both in the Midwest. The reach of the survey extended to two additional institutions each from the West and Southeast regions, and one more institution from the Northeast region. Just over half of the responding clinics (53.3%, n = 8) were also affiliated with an SRFC that offered primary care services. One third of SLFVSPs (33.3%, n = 6) self-identified with a “mobile” descriptor, and five of those were associated with the CSLEC. Notably, there was an even split between mobile and non-mobile operational models within the CSLEC, with five SLFVSPs in each category. 

Demographic and Socioeconomic Characteristics of Populations Screened

There was a 62.5% response rate from 10 SLFVSPs for questions related to patient sex and race/ethnicity. Among the 11 SLFVSPs associated with CSLEC, there were 8 responses (72.7% response rate); among the remaining 5 SLFVSPs, there were 2 responses (40% response rate). At the same time, respondents did not consistently respond to each category within the question; for example, an SLFVSP may have indicated the proportion of Hispanic/Latino patients, however, they did not respond to any other racial or ethnic category or only responded to patient sex and not any other demographic category. In the Figure, the reported patient sex and race/ethnicity are summarized, illustrating the percentages of each characteristic. The percentage of patients less than 18 years old was reported by 9 SLFVSPs, including 8 which reported 0% to 10% and one which reported 31% to 40%. 

Figure.

 

The demographic characteristics of patient populations seeking services at SLFVSPs are depicted, including patient sex and race/ethnicity. “OTHER” refers to indicated race/ethnicity. The prevalence of patients that identified with three socioeconomic parameters: uninsured status, unhoused status, and income at the federal poverty line. As indicated in the figure legend, Mobile SLFVSPs are “red”; CSLEC members are depicted as a solid bar while the non CSLEC members are signified by downward slanting stripes. Similarly, non-mobile SLFVSPs are “blue”; CSLEC members are depicted as a solid bar while the non CSLEC members are signified by upward slanting stripes.

Participating clinics were largely open to the public generally, with only one clinic mandating uninsured status as requisite for receiving screening services, and two other clinics exclusively working with patients referred from another free clinic or the state's Department of Public Health. Few clinics reported socioeconomic parameters, however; the Figure also depicts the socioeconomic status parameters of patient populations screened at mobile and non-mobile SLFVSPs, including uninsured status, unhoused status, and household income under the federal poverty line. Over half of the patients identified as either uninsured or low-income, even though these qualifications were not generally required to participate in the free screening. There were no statistically significant differences in socioeconomic characteristics of patients receiving services from between mobile versus non-mobile clinics. 

Service Delivery Parameters

Faculty supervision was reported in 15 programs (94%), with only one site that lacked direct faculty supervision during screenings but later consulted faculty for subsequent follow-up for screened patients. On average, mobile clinics held 2.3 screening events per month versus 1.2 held by non-mobile clinics (P = 0.16). Further, mobile clinics saw a mean of 9 patients per screening event, whereas non-mobile clinics screened 17 individuals, on average (P = 0.36). There was no statistically significant difference in productivity between mobile and traditional clinics; whereas mobile clinics may have offered twice as many screenings per month, each screening served half as many patients as a traditional clinic, resulting in a similar number of patients served each month. 

Table 1 describes the scope of the screening measurements used by the SLFVSPs. A slit lamp examination was performed by 60% of clinics. Handheld tonometry was the most common method to acquire intraocular pressure (IOP; 73%), although the specific method tonometric device was not identified. Only two clinics did not perform any type of tonometry. Automated visual field testing was performed by 33% of clinics, although the type was also not specified. Nine SLFVSPs reported offering prescription glasses to patients, however, only eight programs reported the inclusion of manual or autorefraction as a part of their services. It is unclear if these programs offer vouchers, used glasses, or another option. 

Table 1.

 

Screening Services Offered by Clinics, Along With Measurement Modality, are Listed

 

Most clinics (63%) evaluated the retina and/or optic nerve via direct ophthalmoscopy, whereas fundus photography, optical coherence tomography (OCT), and dilated fundus examinations were performed only by a limited number of clinics. Among clinics that offered a dilated eye examination (n = 8), seven programs reported using multiple mydriatic drops. Tropicamide (50%, n = 8) and phenylephrine (44%, n = 7) were the most used. Whereas two programs exclusively screened for previous dilated eye examinations, the remaining six programs performed an anterior segment examination prior to dilation, with only one program specifying the use of gonioscopy. 

There is no statistically significant difference in the number of services offered by mobile and non-mobile clinics. Although many clinics offered comprehensive ophthalmic screening protocols (n = 12), others focused on one or two specific areas. For example, two programs exclusively reported only assessing IOP with handheld tonometry and two others exclusively focused on identifying diabetic retinopathy. 

Diagnoses and Follow-Up Care

SLFVSPs reported the most common leading diagnoses in their patient populations through a free-text response. Diabetic retinopathy and cataracts were reported by nine clinics as the most common leading diagnoses. Table 2 depicts the percentage of SLFVSPs that included each leading diagnosis as the most frequently identified. The diagnoses obtained between mobile and traditional clinics were statistically comparable with one exception; a larger percentage of non-mobile clinics reported cataracts as a common diagnosis than in mobile (P = 0.035). 

Table 2.

 

The Number of SLFVSPs Reporting Each Ocular Pathology as a Common, Leading Diagnosis is Reported Aggregately, With a Distinction Between the Mobile and Non-Mobile Service Delivery Models

 

Follow-up from student-led vision screening programs involved disparate implementation systems and outcomes. Follow-up appointments were directly scheduled by four mobile clinics and six non-mobile clinics, whereas referrals were sent by two mobile and four non-mobile clinics. Follow-up attendance rates were heterogeneous (Table 3). For both mobile and non-mobile SLFVSPs, direct scheduling of follow-up care was associated with a greater rate of successful follow-up attendance (66.5% attendance rate compared to 20%, respectively). 

Table 3.

 

The Mean of the Reported Appointment Attendance Rates for SLFVSPs Associated are Depicted as percentage

 

Table 4 shows the perceived barriers to successful follow-up after screening. Transportation was the most cited barrier (89%) to obtaining follow-up care for both mobile and non-mobile clinics. Among non-mobile clinics, housing was found to be a more significant barrier (P = 0.048) compared to mobile clinics. Although not statistically significant, it also appears that trust is not comparable between the 2 service models, with 0% of non-mobile clinics reporting this as a barrier for follow-up appointment attendance and 60% of mobile clinics endorsing trust as a problem. 

Table 4.

 

Perceived Barriers to Appointment Attendance Reported by SLFVSPs

 

Discussion

Free vision screenings play an important role in identifying eye pathologies in the community and referring patients for follow-up examinations and treatment, if indicated.³ In this study, we conducted a survey of student-led vision screening programs that ultimately recruited insights from five additional SLFVSPs beyond the CSLEC, and which added geographic diversity in this sample. Our study expands upon that of Okaka et al., which surveyed 14 vision screening programs affiliated with SRFCs, to also include SLFVSP that do not necessarily operate in conjunction with a primary care oriented SRFC. Our study did not include optometry schools, whose participation in student led initiatives is still unknown, and an area for future investigation. Despite our efforts at personalized outreach, we were unable to solicit a response from many programs, perhaps due to shifting leadership, resident-led or post-graduate trainee-led presence, or inactivity despite an online presence. It is likely that our sample is skewed toward SLFVSPs with established faculty oversight and this study unlikely represents the breadth of SLFVSPs that do not have a significant online presence. 

Apart from a limited response to the overall survey, there were specific questions that lacked responses proportional to the total number of respondents. The highest response rates pertained to clinical care, with a 94% response rate for the types of screening services offered and an 88% response rate for the most common diagnoses. At the same time, the lowest response rates pertained to patient demographics (63%), with incomplete responses even between different categories (not indicating proportion for all racial and ethnic groups listed), barriers to follow-up care (56%), and appointment attendance rates (43%). Because there were no required survey question responses to submit the survey, these non-respondent answers may reflect a possibility that individual SLFVSPs do not maintain records pertaining to all the questions in this survey. Further, some SLFVSPs may maintain more meticulous records than others related to patient care and operations or may not be able to easily prioritize consistent and robust characterization of demographics and follow-up care needs. Therefore, implementation of a formal data collection template for SLFVSPs may improve future efforts at nationwide data collection. Moreover, although the survey item describing most identified conditions received the highest response rate, the parameters by which these diagnoses were identified or characterized is unknown; as a case in point, it is unknown if cataracts were visually significant or whether diabetic retinopathy was staged. Although this non-response bias limits the ability to extrapolate findings from this survey data to represent national data, this survey is, to our knowledge, the most comprehensive questionnaire in the scope SLFVSP operations. 

A particular important finding was that the overall follow-up appointment attendance rate was about 51%, with greater adherence in programs that directly scheduled appointments for patients versus those who exclusively offered a referral. The Hoffberger program, one community vision screening program in Maryland with a combination of lay-person and technician volunteers, reported a no-show rate to follow-up appointments of about 50%, with little improvement gained by reducing barriers, such as cost and transportation.⁸ Two decades later, based on the respondents in this specific cohort, there appears to be little improvement in follow-up appointment attendance; whereas there was a limited response rate to this item, there is a strong possibility that the success of follow-up care from SLFVSPs is largely unknown. At the same time, the Manhattan Vision Screening and Follow-up Study in Vulnerable Populations (NYC-SIGHT) reported an 85% follow-up appointment attendance rate, likely achieved through scheduling appointments at a familiar location, already known to the participants. At the same time, 95% of the patients in the NYC-SIGHT cohort were reported to be insured, with 62% receiving Medicare coverage, rendering direct comparisons difficult.⁹ Although the results of this survey are not representative of a national sample, there is opportunity in exploring the effects of directly scheduling of follow-up appointments in known community eye centers. At least in the context of SLFVSPs, optimizing a structured protocol to follow-up care can be an important action step, regardless of the choice of between mobile versus non-mobile approaches. 

Analysis of barriers to follow-up care showed that transportation was the most significant barrier to follow-up. This was true for both mobile and non-mobile clinics. Elam et al. found transportation as one of the frequently identified barriers in the utilization of eye care in high-risk individuals.¹ This obstacle is not exclusive to vision screening and has been a significant problem for other student-run free clinics.¹⁰ Whereas mobile program models allow greater access to patients by bringing the clinic to the patient, they show similar rates of transportation barriers for follow-up, which frequently requires patients to visit an academic eye care center that is far from where the screening was performed. Addressing follow-up barriers will be a crucial part in improving the quality of student-led vision screening programs. 

Further, housing insecurity was more frequently reported as a barrier in non-mobile clinics, whereas trust was more frequently a barrier in mobile clinics, which may not have established connections with the community. In fact, non-mobile clinics self-reported a higher percentage of unhoused patients, suggesting that SLFVSPs which are established near or with a homeless shelter may see a regular patient population consisting of unhoused individuals. As noted in the Hoffberger program analysis, follow-up for individuals that are unhoused can be complicated by a lack of a telephone number and address, and integrating vision services in homeless shelters, for example, could be one possible solution.⁸ One CSLEC program (JHU) has implemented a strategy of partnering with local community-based grassroots service organizations to co-sponsor each screening event, as a method to enhance trust building with the community. At a larger scale, transportation and literacy continue to also serve as major barriers for follow-up appointment attendance, reinforcing that poor follow-up attendance rates are a multifactorial problem, which merits the rigorous study of additional barriers to follow-up care. 

As discussed previously, even though SLFVSPs reported the most frequent, leading diagnoses, the true prevalence of these conditions is unknown because definitive follow-up examinations were not universally implemented. Other studies conducted by student-led vision screening clinics demonstrate that refractive error, diabetic retinopathy, cataracts, and glaucoma are frequently identified, and, inherently, these diagnoses require definitive follow-up evaluation treatment, reiterating the importance of a robust mechanism for continuity of care, and arguably, definitions of indications for referral. 

In fact, there is little clarity on defining diagnoses themselves. Both mobile and non-mobile SLFVSPs demonstrated use of the slit lamp examination and, yet, a statistically significant difference was noted in the rate of cataract diagnosis; this difference may be explained by differences in patient population or the criteria of reporting cataracts, with some clinics only reporting visually significant cataracts as opposed to any other type of cataract. This point is underscored by the inclusion of both glaucoma and glaucoma suspect among respondents; there is little clarity as to how abnormal findings identified in the vision screening are reported or translated into a leading diagnosis without subsequent definitive care. Furthermore, certain SLFVSPs only offered restricted services, such as screening for diabetic retinopathy exclusively or only measuring patient IOPs, which may skew the aggregate reporting of most common leading diagnoses in this cohort. Furthermore, only a subset of SLFVSPs offer prescription glasses; refractive error screening through pinhole testing using a Snellen chart can be one step, for example, to better identify and address refractive error, even in the absence of access to highly specialized refractive devices.¹¹ Thus, there is a dire need to standardize the services offered and consolidate criteria for identifying suspect versus definitive diagnoses. 

During our analysis we recognized that mobile clinics and non-mobile clinics serve a similar number of patients in each month. A limiting factor in our analysis was the designation of mobile and non-mobile clinics using self-identifier. More standardized criteria may better characterize these categories, however, if establishing a permanent, single-site clinic proves challenging, mobile clinics could serve as a comparable alternative that can help overcome barriers to accessing ophthalmic care. 

Future Directions

As with any other survey tool, this effort is limited by non-respondents and recall bias. Given that most responses were from CSLEC members, predominantly reflecting conditions in the Midwest and Northeast, the results of our study will be disseminated to CSLEC members along with recommended actions to take accordingly. The CSLEC seeks to develop a standard template for data stewardship practices, to consider factors like patient demographics and disease prevalence, in real-time. Furthermore, standardizing the methodology for vision screenings across clinics will improve inter-clinic reliability and assure that SLFVSPs are sufficient in and of themselves to bridge vision care utilization. An uninsured patient who exclusively receives an IOP measurement, for example, misses other essential components of vision screening care. The process of standardization can support all CSLEC member, and non-member, SLFVSPs to develop the infrastructure needed to assure quality and define what constitutes an impactful community screening event. 

As a part of standardizing services, best-practices should also be established, such as identifying optimal approaches to visual field testing, handheld tonometry devices, ophthalmic imaging, and dilated eye examination recommendations. If SLFVSPs approach screening similarly, a more concerted effort can be undertaken for population health studies nationally to better estimate the prevalence of conditions in at-risk populations. In addition, streamlining protocols for follow-up care to include scheduling direct appointments for patients may help mitigate some difficulties with appointment attendance. Further pilot interventions can be introduced to address other barriers to ameliorate this long-lasting problem of poor follow-up; for instance, transportation vouchers could be offered to attend follow-up appointments. It is not enough for patients to be screened; definitive management is often required for vision threatening disease. 

Our objective is to systematically improve eye care delivery models to ensure adequate screening methodologies and enhance access to eye care, both medically and surgically. 

Conclusions

This study highlights the results of a survey that evaluated the delivery methods of vision screening programs led by students in both mobile and non-mobile SLFVSPs primarily affiliated with the CSLEC, and by five additional programs. This work expands upon other assessments of community vision screening programs, designed to identify vision-threatening disease in asymptomatic individuals within economically marginalized populations. We sought to report on baseline characteristics of service delivery characteristics, patient demographics, and follow-up care in SLFVSPs. As a part of this survey, we also solicited information about the operational models of SLFVSPs and will report the findings of these data in subsequent reports soon. 

Acknowledgments

The authors offer their gratitude to the Journal of Student Run Free Clinic Editorial Fellowship, which through the Macy Jr. Foundation, offered an initial platform to create the CSLEC. We thank our faculty mentors and the innumerable contributions of student leaders and volunteers who make vision screening programs possible. 

Supported by the Macy Jr. Foundation: Editorial Fellowship Program with the Journal of Student Run Clinics (N.D. and M.S.). 

Disclosure: N. Devanathan, None; M. Scheive, None; B.S. Nawash, None; A. Selvam, None; A. Murphy, None; M. Morrow, None; S. Anant, None; N. Chen, None; E.A. Martin, None; J.S. Kruger, None; C.-W.R. Yung, None; T.V. Johnson, None