Community-Based Trial of a Peripheral Prism Visual Field Expansion Device for Hemianopia

Objective  To determine the functional utility for general mobility of peripheral prism glasses, a novel visual field expansion device for hemianopia, in a large-scale, community-based, multicenter study with long-term follow-up.

Methods  Forty-three participants with homonymous hemianopia were fitted with temporary press-on Fresnel peripheral prism segments of 40 prism diopters. Follow-up questionnaires evaluating functional benefits for mobility were administered in the office at week 6. Participants who continued wearing the prisms were interviewed again by telephone after a median of 12 months. Primary outcome measures included clinical success (a clinical decision to continue wear) and 5-point ratings of prism helpfulness for obstacle avoidance when walking.

Results  Thirty-two participants (74%) continued prism wear at week 6, and 20 (47%) were still wearing the prisms after 12 months (median time, 8 hours per day). These participants rated the prism glasses as very helpful for obstacle avoidance and reported significant benefits for obstacle avoidance in a variety of mobility situations. Success rates varied among clinic groups (27%-81%), with higher rates at the clinics that fitted more patients.

Conclusion  Our results demonstrate the functional utility of peripheral prism glasses as a general mobility aid for patients with hemianopia.

Homonymous visual field defects (HVFDs) have a prevalence of 0.8% in the general population older than 49 years.¹ They are caused by lesions in the postchiasmal visual pathways, primarily due to strokes and, to a lesser extent, trauma and tumors.² Patients with HVFDs have difficulty detecting obstacles on the side of the loss, resulting in impaired mobility. The main vision-based rehabilitation strategies include provision of optical aids (eg, prismatic corrections to expand or relocate the visual field)³; visual search training to improve the efficiency of visual exploration on the side of the loss^4-7; and visual restoration training to recover lost visual function.⁸ Herein we report, to our knowledge, the first multicenter evaluation of a novel prismatic device to expand visual fields for patients with HVFDs.

The introduction of press-on Fresnel prisms in the 1970s provided a temporary and inexpensive method of applying prismatic corrections for patients with HVFDs.^9-12 Low- to moderate-powered prism segments (12-20 prism diopters [Δ]) were fitted mainly as binocular or monocular sector corrections (covering only part of the lens), providing field relocation or expansion of less than 10°.^10,13-21 Neither method provides visual field expansion effective in all positions of gaze.²² Binocular sector prisms do not provide field expansion, and they only relocate (shift) images to a functional part of the field when the gaze is directed into the prism. Monocular sector prisms can provide visual field expansion, but only when the gaze is directed into the prism, and the central diplopia accompanying the field expansion may be disorienting to thepatient.^3,23

In 2000, Peli²² described a novel prismatic device to expand the visual fields of patients with HVFDs. High-power press-on prism segments (40Δ) are placed across the whole width of the spectacle lens above and below the pupil area on the side of the field loss (Figure 1). Visual field expansion of about 20°, which is effective at all lateral positions of gaze, is provided via peripheral diplopia (Figure 2). Patients are taught to view through the central, prism-free area of the spectacle lens (never looking into the prisms), so that central diplopia does not occur. A case-series report by Peli²² and a recent laboratory-based clinical evaluation²⁴ suggest that many patients find the peripheral prism glasses helpful for obstacle avoidance when walking. However, these studies were limited by small sample sizes, and the prisms were fitted by Peli and coworkers.^22,24

The purpose of this study was to conduct a multicenter evaluation of the peripheral prism glasses with long-term follow-up to determine patient acceptance and functional utility of the glasses for general mobility (walking) when fitted by community-based vision rehabilitation practitioners. Furthermore, to aid future development of a simple-fitting protocol, the minimum interprism separation that could be tolerated when walking was determined. We used procedures and measures that could be implemented within the normal schedule of a vision rehabilitation clinic and that reflected those typical in clinical practice, thus opening up the study and the novel peripheral prism technique to a wide group of practitioners.

The Schepens Eye Research Institute acted as the coordinating and data management center for the study. The tenets of the Declaration of Helsinki were followed, and the study was approved by the internal review board of the institute. Data were collected from April 30, 2004, to December 31, 2005.

Fifteen community-based vision rehabilitation practitioners at 18 clinics across the United States recruited and screened patients, fitted prisms, and performed follow-up evaluations. Most of the practitioners were recruited by means of an announcement about the study at the American Academy of Optometry 2003 Annual Meeting.

Participants with complete homonymous hemianopia, as determined by a recent (within the previous 3 months) visual field plot, and corrected monocular visual acuity of at least 20/50 in each eye were recruited. Only patients with no physical or cognitive impairments, balance problems, or other deficits that could impair their ability to walk or use the peripheral prism glasses were included. Based on the findings of case histories and medical records, patients with visual neglect, diagnosed dementia, or a history of seizures in the past 6 months were excluded. Visual field mapping extended to at least 50° from fixation in all directions and was performed using Goldmann perimetry (V-4-e target; Haag-Streit, Koeniz, Switzerland), Humphrey Field Analyzer 120-point full-field screening test (Carl Zeiss Meditec, Dublin, California), or similar tests, depending on the equipment available at each clinic. Before screening data were collected, the nature of the study was explained and informed consent was obtained from all participants. To ensure that study inclusion criteria were uniformly applied (in particular, that all participants met the criteria for complete homonymous hemianopia), screening data and visual field plots were sent to the study coordinator (A.R.B.), who determined eligibility.

Study procedures aimed to ensure that all participants were treated equally, regardless of the clinic they attended. Detailed written protocols and data sheets were provided to each practitioner by the study coordinator. After each assessment, data sheets were sent to the study coordinator for immediate review. This day-to-day monitoring ensured protocol adherence and speedy remediation of protocol deviations. Figure 3 summarizes the study visit schedule.

The complete study protocol for prism fitting and training is available on the data sharing page of the Vision Rehabilitation Laboratory Web site (http://www.eri.harvard.edu/faculty/peli/index.html). Fitting procedures were modified from the method proposed by Peli.²² In brief, upper and lower 40Δ press-on Fresnel prism segments (3M Press-On Optics; 3M Health Care, St Paul, Minnesota) were fitted to 1 lens of participants' spectacles. The prisms are intended as a mobility aid; therefore, they were fitted to single-vision distance, bifocal, or progressive addition lenses. If the participant did not have suitable spectacles, they were provided with study glasses at no cost. The prisms were fitted on the side of the field loss (eg, left eye for left hemianopia) with the base in the direction of the field loss (base-out prismatic effect), so that objects from the nonseeing side were imaged (and hence detected) on the functional side of the retina. The upper prism was fitted first (week 0) and worn at home for 2 weeks before the lower prism was fitted (week 2), providing a graduated introduction to the use of the prisms. If the prism segments were fitted to bifocal or progressive addition lenses, a small semicircular aperture was cut from the bottom part of the lower segment to provide sufficient area for short-duration reading through the near-vision correction (Figure 1).

The prism-fitting procedure was designed to determine the minimum amount of interprism separation that could be tolerated. Toleration was defined as comfortable single central vision with no change in head posture between walking without and with the prisms. The starting point for fitting the upper prism was to place the lower edge of the segment 6 mm above the pupil center. The participant then walked around the clinic with the prism at this height. Toleration of the position was determined by observations of head posture before and after fitting (eg, elevation of head posture with the prism indicated that it was set too low), and by asking the patient whether he or she noticed central diplopia or noticed the prism interfering with central vision. The prism position was adjusted in a staircase fashion using 2-mm followed by 1-mm steps to determine the lowest tolerated position (closest to the pupil center). The prism was moved down toward the pupil center until it was below the lowest tolerated position, such that either it interfered with central vision or head posture was elevated. It was then moved back up to find the lowest position at which it first did not interfere with central vision or alter head posture. A similar procedure was followed for fitting the lower prism, starting with the upper edge of the segment 6 mm below the pupil center and adjusting to determine the highest tolerated position (closest to pupil center).

Participants were taught to view through the central prism-free area of the spectacle lens at all times and to turn the head and eyes to fixate objects of interest that were initially detected from the prism image in peripheral vision. A simple reach-and-touch training exercise was used to familiarize participants with the relationship between the apparent and real positions of objects detected from the prism image; this exercise was also encouraged for home training. Participants were given verbal and written instructions about how to use the prism glasses and were encouraged to wear them as much as possible each day. They were advised not to use the peripheral prism glasses for driving or prolonged reading; if necessary, a separate pair of reading glasses was provided at no cost.

The fitting positions of the upper and lower prisms were reviewed at the start of the week 2 and week 6 visits (Figure 3), respectively. If necessary, the position was adjusted and another 2 weeks of prism wear was provided before progressing to the next part of the study.

After both prism segments had been worn for 4 weeks, an in-office follow-up interview was conducted (week 6). A clinical decision was made concerning whether to continue prism wear. The criteria to continue wear were (1) the prisms were helpful for mobility (eg, helpful for obstacle avoidance when walking outdoors); (2) the patient wanted to continue wear; and (3) the practitioner deemed that it was clinically appropriate. Participants' experiences of wearing the prisms were evaluated through a series of questions, including 5-point ratings of prism helpfulness for detecting obstacles on the “blind” side in time to avoid them when walking; 5-point ratings of vision comfort when wearing the prisms (because the prisms might, for example, cause visual discomfort due to overhead glare, or due to central diplopia if not used correctly); and open-ended questions about mobility situations where the prisms were very helpful and about any difficulties encountered.

At least 6 months after the week 6 visit, a long-term telephone interview was conducted with all participants for whom the clinical decision was to continue wear. For participants who were still wearing prisms, the interview included all of the questions asked at the short-term follow-up, supplemented by an additional question asking whether they would be willing to pay $600 for permanent prism glasses. For participants who had discontinued wearing the prisms, only questions to ascertain the reasons for discontinuing were asked. Participants who continued to wear press-on prisms at the week 6 interview and press-on or permanent prisms (as described in the “Permanent Prism Glasses” subsection of the “Methods” section) at the long-term interview were allowed to keep them at no cost.

The primary outcome measures were clinical success, defined as a clinical decision to continue wear, and ratings of prism helpfulness for obstacle avoidance at the short- and long-terminterviews.

While the study was in progress, Chadwick Optical, Inc, developed a rigid form of the Fresnel prism segment that could be embedded into a plastic spectacle lens. This provided a permanent peripheral prism spectacle correction (Figure 4) as an alternative to the temporary press-on prisms, which might have important advantages for long-term wear (better durability, better optical quality, and no need for replacement every 3 months). For these reasons, when the permanent prisms were first available in April 2005, 15 of the 18 study participants who were wearing temporary press-on prisms at that time, and who had worn them for at least 2 months after the week 6 visit, were provided (at no cost) with permanent 40Δ peripheral prism glasses. We were interested in long-term use of the prism glasses; therefore, our selection was biased toward patients whom practitioners deemed most likely to continue long-term wear. Because there were insufficient funds to provide permanent prisms to all participants, those who completed 2 months of wear after April 2005 (7 participants) were not fitted with the permanent form of the prisms. Permanent prism glasses were provided about 5 to 6 months after the week 6 visit and about 5 to 6 months before the long-term interview. The prisms were fitted at the same interprism separation as the press-on prisms and provided a similar visual field expansion effect.

For statistical analyses, participants were grouped according to final status, consisting of discontinued prism wear at or before week 6, discontinued prism wear after week 6, and continued long-term prism wear. Differences in fitting positions, wearing times, and prism ratings among participants grouped by final status were evaluated at week 6. For participants who discontinued prism wear before week 6, data included in these analyses were from the last visit before week 6. Because several clinics fitted only a small number of patients, the following 2-way grouping of clinics was used (Table 1): group A (each practitioner fitted ≥ 8 participants) and group B (each practitioner fitted ≤ 5 participants). None of the continuous variables conformed to a normal distribution; therefore, nonparametric statistics were used for all between-group comparisons. P < .05 was taken to indicate statistical significance.

Sixty potential participants were screened. Fifty met the study criteria and were enrolled; 7 withdrew before the upper prism fitting and 43 were fitted with prisms (Figure 3). Reasons for early withdrawal included concerns about falling (1 participant), fitting with prisms outside of the study (1), and unknown reasons (5). Prisms were fitted to single-vision distance lenses (24 participants [56%]) or bifocal lenses (18 [42%]). Only 1 participant (2%) had prisms fitted to progressive addition lenses.

For 32 of the 43 participants fitted with prisms (74%), the clinical decision at the week 6 visit was to continue wearing the prisms, and 20 (47%) were still wearing them 12 months later at the long-term interview (Figure 3 and Figure 5). Of the 15 participants who were fitted with permanent prism glasses, 11 (73%) were still wearing them at the long-term interview, 2 (13%) had reverted to press-on prisms (owing to a problem with an early prototype of the bifocal permanent prism glasses), and 2 (13%) had discontinued wear. By comparison, 7 of the 10 participants who were not fitted with permanent prisms (70%) were still wearing press-on prisms at the long-term interview (Figure 3). The main reasons for discontinuing prism wear were difficulties in adapting to the prism images, including confusion of images or sudden appearance of images causing anxiety (7 of the 22 participants who discontinued [32%]), no perceived benefit (3 participants [14%]), and deteriorating general health (3 participants [14%]).

Long-term success rates were significantly higher at clinics where 8 or more patients were fitted than at clinics where 5 or fewer patients were fitted (81% vs 27%; χ²₁ = 12; P = .001; Table 1). Evaluating clinical variables (summarized in Table 2) that might affect long-term success rates was not an aim of this study; a much larger sample would have been needed. Spontaneous improvement in hemianopia, most likely to occur within the first 3 months after the onset of hemianopia,²⁵ could affect success rates, because patients who recover lost visual field might discontinue wearing prism glasses. In our sample, 5 participants had hemianopia for less than 6 months, but only 1 discontinued wear. Therefore, it is unlikely that spontaneous improvement affected success rates in this study.

There were no differences in final upper or lower prism-fitting positions between long-term wearers and participants who discontinued prism wear (Kruskal-Wallis test, χ²₂ < 3.0; P > .2; Figure 6), and there were no differences in fitting positions between clinic groups (Mann-Whitney test, z < −0.7; P > .5). The median interprism separation for the 32 participants who continued wear at week 6 was 11 mm (interquartile range [IQR], 8-11 mm). The positioning of the lower prism was more critical because it was more likely to interfere with central vision when walking. Seven participants had the lower prism position changed after the visit when it was first fitted (median, 2 mm away from the pupil center), whereas only 4 participants had the upper prism position changed (median, 1.25 mm away from the pupil center).

At the week 6 follow-up, participants who continued long-term prism wear reported longer daily wearing times (Mann-Whitney test, z = −1.6; P = .1; Figure 7A), higher vision comfort ratings (Mann-Whitney test, z = −3.0; P = .003; Figure 7B), and higher obstacle avoidance ratings (Mann-Whitney test, z = −4.3; P < .001; Figure 7C) than did participants who discontinued prism wear at week 6. Minor differences were also apparent between long-term wearers and participants who discontinued wear after week 6; that is, daily wearing times were shorter and obstacle avoidance ratings were lower in the latter group, suggesting that these participants already (at week 6) found the prisms less useful than the long-term wearers did (Mann-Whitney test, z = −1.8; P = .07; Figure 7A and C). Participants who discontinued wear at week 6 gave significantly lower obstacle avoidance ratings than did participants who discontinued wear after week 6 (Mann-Whitney test, z = −3.5; P < .001; Figure 7C), but daily wearing times were similar (Mann-Whitney test, z = −0.6; P = .6; Figure 7A).

At the long-term interview, the 20 participants who continued prism wear reported median daily wearing times of 8 hours (IQR, 4-13 hours) and rated the prism glasses as very helpful for obstacle avoidance (median rating, 5; IQR, 4-5), with high levels of vision comfort (median rating, 4; IQR, 3-5). The ratings were not significantly different from those at week 6 (Wilcoxon signed rank test, z < −0.6; P > .6), and there were no differences in ratings between those participants wearing press-on and permanent prisms (Mann-Whitney test, z < −0.5; P > .6). Long-term wearers reported that the prisms were particularly helpful when shopping in malls and stores and moving in crowded and unfamiliar areas (17 participants [85%]). As expected, the main situation in which the glasses were not worn was during performance of near-vision tasks, including reading and using the computer (9 participants [45%]). A minority of long-term wearers reported difficulties when using the prism glasses (8 participants [40%]), most commonly problems with steps or curbs (2 participants [10%]) and reading (3 participants [15%]). Of the 19 long-term wearers who answered the question about whether they would be willing to pay $600 for permanent prism glasses, a significantly higher proportion of permanent prism wearers than press-on wearers responded in the affirmative (11 participants [100%] vs 5 participants [63%]; χ²₁ = 4.9; P = .03).

The results of this multicenter evaluation demonstrate the utility of the peripheral prism glasses as a mobility aid for patients with HVFDs (specifically complete hemianopia). Almost half of all participants were still wearing the prism glasses after 12 months, typically for 8 hours per day, reporting significant benefits for obstacle avoidance in a variety of mobility situations. The minimum interprism separation that could be tolerated when walking was determined for each participant, a time-consuming process that would not be practical in a busy clinic. Currently, we are evaluating a simplified fitting protocol using a standard interprism separation of 12 mm (the 90th percentile of the data from participants who continued wear at week 6).

The overall 6-week and 12-month success rates in this study (74% and 47%, respectively) were similar to those reported in early, single-center evaluations of the peripheral prism glasses (when prisms were fitted by Peli and coworkers).^22,24 However, long-term success rates varied widely between clinics (27%-81%), with higher rates at clinics where more participants were fitted. All participants were treated equally irrespective of which clinic they attended, and final prism-fitting positions did not differ between clinics, indicating that practitioners adhered to the standardized fitting protocols of the study. Nevertheless, the results suggest that confidence in fitting and training patients to use the prisms may be an important factor in determining success. The early withdrawal rates (before the prisms were fitted) were lower in clinics with higher success rates (Table 1). A limitation of the study was the large number of clinics (n = 9) that fitted only 1 or 2 patients compared with only 2 clinics that fitted 8 or more patients. Although this was a less-than-ideal situation for a multicenter study, our aim was to introduce the concept of fitting the prisms to as many practitioners as possible.

The potential advantages of the permanent prisms and the feel-good factor of being one of the first people to evaluate the new permanent prisms could have contributed to higher success rates among participants fitted with permanent prisms than among those not fitted with permanent prisms. However, there was little evidence in the study results of higher success rates in the former group, and comparisons were limited by small sample sizes. Although 40Δ press-on and permanent prisms provide similar visual field expansion effects, permanent prisms have several advantages compared with press-on prisms for long-term wear. The initial optical quality is better and does not deteriorate with time, the durability is far superior, and the prisms do not have to be replaced every 3 months. It is encouraging that most of the long-term wearers (both those fitted and those not fitted with permanent prisms) indicated that they would be willing to pay $600 for permanent prism glasses; this provides an indication of the perceived importance of the benefits of the glasses in their everyday lives. Of the 12 patients described in the early series by Peli,²² 4 purchased the permanent prism glasses when they became available and 1 ordered a replacement pair when his prescription changed after cataract surgery. The Massachusetts Medicaid program has subsequently preapproved payment for the permanent prism glasses for 2 patients.

The success rates for the peripheral prism glasses were similar to or better than those reported for alternative monocular sector prism designs (eg, 20%-60%).^17,18,20,21 However, evaluations of these alternative prism designs were limited by small samples (≤ 10 patients),^18,21 or by the lack of clarity and consistency in reporting methods and results.²⁰ Furthermore, success rates in these studies might have been overestimated because monocular sector prism glasses could have been worn without experiencing central diplopia.^17,18,20,21 Because a monocular sector prism covers less than half of the spectacle lens on the side of the visual field loss, it would have been possible for patients to wear the glasses yet avoid central diplopia by never looking through the prism (and never experiencing the intended field relocation effect). By comparison, the peripheral prism design offers the advantage of visual field expansion that is present all of the time and in all lateral positions of gaze. However, if the peripheral diplopia is bothersome for the patient, the only relief is to remove the glasses. The typical field expansion (20°) with the 40Δ peripheral prisms is about double that reported for monocular sector designs²⁰ and is 4 times greater than the average 5° field recovery reported for vision restoration therapy,^8,26 a much more time-intensive, expensive, and controversial rehabilitation method with doubtful efficacy.^27-31 Furthermore, the peripheral prism glasses could be used to provide additional visual field expansion to supplement visual restoration or visual search training.

The main factor leading to discontinuation of prism wear was confusion of images or anxiety due to the sudden appearance of images, suggesting that these participants did not fully comprehend how to use the prisms. Such patients may have benefited from further training. Other minor reported difficulties included difficulties with reading (the inconvenience for bifocal users of having only a small aperture in the lower prism segment that could only be used for short-duration reading; other glasses had to be used for prolonged reading) and problems with steps or curbs. The latter difficulty results from the lower prism segment extending across the entire inferior region of the spectacle lens. If the wearer looks down to negotiate descending steps and happens to glance through the lower prism, central diplopia will result. To alleviate this, press-on prisms could be cut to provide a prism-free area below the lower segment (as is the case with the permanent prism glasses; Figure 4).

Peripheral prism glasses provide a promising and relatively inexpensive treatment for HVFDs that can be successfully fitted by community-based practitioners. Based on the participants' reports and acceptance of the device, this study provides evidence of the functional utility of the peripheral prism glasses to aid patients with hemianopia with their general mobility. However, objective measures of functional performance with and without prisms and a control or comparison treatment were not included. To provide a more rigorous evaluation of the efficacy of the peripheral prism glasses,³¹ we are now conducting a randomized, controlled, multicenter crossover trial using real and sham prisms.

Correspondence: Alex R. Bowers, PhD, Schepens Eye Research Institute, 20 Staniford St, Boston, MA 02114 (alex.bowers@schepens.harvard.edu).

Submitted for Publication: February 28, 2007; final revision received October 4, 2007; accepted October 10, 2007.

Author Contributions: Dr Bowers had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Financial Disclosure: Dr Peli has financial interest in a patent related to the peripheral prism glasses (assigned to Schepens Eye Research Institute) and was a paid consultant to Chadwick Optical, Inc, on the design of the permanent prisms. Ms Keeney has licensed that patent for Chadwick Optical, Inc.

Funding/Support: This study was supported in part by grants R43-EY014723 (Ms Keeney) and R24-EY12890 (Dr Peli) from the National Institutes of Health.

Role of the Sponsor: The National Institutes of Health had no role in study design, data collection, data analysis, or report writing.

Additional Contributions: Doris Apfelbaum, BA, maintained the study database. Practitioners who fitted the peripheral prisms and collected data included JeanAstorino, OD (Blind and Vision Rehabilitation Services of Pittsburgh, Bridgeville, Pennsylvania), Alfred Dick, OD (Byron Optometry, London, Ontario), Bruce Gilliland, OD (SE Retina Associates, Knoxville, Tennessee), Robert Goulding, OD (The Sight Center, Toledo, Ohio), Erica Hacker, OD (Blind and Vision Rehabilitation Services of Pittsburgh, Pittsburgh, Pennsylvania), M. Richard (Scott) Hearing, OD (Visual Health Jupiter Eye Center, Jupiter, Florida), Jill Pauly Holler, OD (The Sight Center, Toledo), Rebecca Kammer, OD (Southern California College of Optometry, Fullerton), George Kremer, MA, and Dennis Cobler, OD (Association for the Blind and Visually Impaired, Grand Rapids, Michigan), David Lewerenz, OD (Eyecare Associates of Salina, Salina, Kansas), Tracy Matchinski, OD (The Chicago Lighthouse, Chicago, Illinois), James L. Nedrow, OD (Oculi Vision Rehabilitation, Lincoln, Nebraska), Cheryl Reed, OD (United Disability Services, Akron, Ohio), and Laura Windsor, OD, and Richard Windsor, OD (Low Vision Centers of Indiana/Eye Associates Group, Hartford City).