ERIC EJ1149518: The Experience of a Randomized Clinical Trial of Closed-Circuit Television versus Eccentric Viewing Training for People with Age-Related Macular Degeneration

2017·0.44 MB·English

by  ERIC

The Experience of a Randomized
Clinical Trial of Closed-Circuit
Television Versus Eccentric Viewing
Training for People with
Age-Related Macular Degeneration
Susan J. Leat, Francis Fengqin Si, Deborah Gold, Dawn Pickering,
Keith Gordon, and William Hodge
Structured abstract: Introduction: In addition to optical devices, closed-circuit
televisions (CCTVs) and eccentric viewing training are both recognized interven-
tions to improve reading performance in individuals with vision loss secondary to
age-related macular degeneration. Both are relatively expensive, however, either in
the cost of the device or in the amount of time personnel need to provide training.
In this randomized trial, we compared the effectiveness of these two interventions.
Methods: Participants with age-related macular degeneration and visual acuity
between 6/48 (20/160) and 6/120 (20/400) first received basic low vision care,
including optical devices. At the subsequent baseline visit, they undertook a battery
ofmeasuresincludinglogMARvisualacuity;readingspeedandaccuracyfortextin
1.3M and 1M fonts; reading information on medicine bottles, utility bills, and food
packages; the NEI-VFQ; the Geriatric Depression Scale; and a reading inventory
questionnaire.TheywerethenrandomizedtoeitherobtainingaCCTVforhomeuse
oreccentricviewingtrainingoverthefollowingsixweeks.Results:Recruitmentwas
moredifficultthanexpectedforthispopulation.Of145patientsreferred,29metthe
inclusion-exclusioncriteria,14werewillingtoenroll,and10completedthetrial.For
the primary outcome (reading speed for 1.3M print), there was a significant im-
provementbetweenbaselineandoutcomefortheCCTVgroup(p(cid:2)0.005),butnot
for the eccentric viewing training group (p (cid:2) 0.28), and the CCTV group showed
significantlygreaterchange(p(cid:2)0.04).Therewasanonsignificantimprovementin
reading speed for 1M text and a decrease in the amount of time taken to read utility
bill information in the CCTV group. There was a significant improvement in near
visual acuity with current glasses with eccentric viewing training. The other mea-
suresdidnotreachstatisticalsignificance.Discussion:Randomizedclinicaltrialsfor
low vision rehabilitation, particularly in the elderly population with vision loss, are
challenging, but such trials are important for the allocation of resources. This trial
showed early indications of more impact on reading performance from CCTV than
eccentric viewing training.
354
JournalofVisualImpairment&Blindness,July-August2017 ©2017AFB,AllRightsReserved
I (Goodrich & Kirby, 2001; Peterson et
n Western countries, age-related macular
al., 2003).
degeneration is the most frequent cause
Those with more advanced age-related
of severe visual impairment (Klein, Lee,
macular degeneration develop a central
Gangnon,&Klein,2013;Quartilhoetal.,
scotoma, which necessitates the develop-
2016) and reading is the most common
ment of a preferred retinal locus for fix-
goal—it is the first or second rehabilita-
ation, instead of the dysfunctional ana-
tion goal for 96% of this population
tomical fovea. Although preferred retinal
(Elliottetal.,1997).Peoplewithage-related
locus development occurs naturally with
macular degeneration usually require
time for the majority of patients (Cross-
magnification in order to be able to read
land, Culham, Kabanarou, & Rubin,
standard print. Magnification can be pro-
2005; Fletcher & Schuchard, 1997), it is
videdeitherbyopticaldevicesor,formod-
thought that in many cases the naturally
erate and advanced vision loss, electronic
developedpreferredretinallocusisnotin
video enhancement systems or video
the ideal position (Fine & Rubin, 1999;
magnifiers, commonly called closed-
Fletcher & Schuchard, 1997; Nilsson,
circuit televisions (CCTVs). The signif-
1990; Petre, Hazel, Fine, & Rubin, 2000).
icant advantages of CCTV over optical
Therefore, eccentric viewing training
magnifiers are that it provides high levels
is commonly recommended (Owsley,
of magnification with a greater field of
McGwin, Lee, Wasserman, & Searcey,
view (compared to the equivalent optical
2009; Stelmack, Massof, & Stelmack,
device); allows reading at a more typical
2004) with several purposes in mind: to
viewing distance of about 40–50 cm
develop the more consistent and efficient
(15.8(cid:3)–19.7(cid:3));andallowsbinocularview-
use of a preferred retinal locus; to speed
ing (which higher-powered optical mag-
up the process of developing a preferred
nifiers cannot) (Peterson, Wolffsohn, Ru-
retinal locus; and to optimize the pre-
binstein, & Lowe, 2003). There is a
ferred retinal locus position (Gaffney,
general consensus among studies that
Margrain, Bunce, & Binns, 2014).
reading speeds are faster with CCTVs
Both of these interventions for more
compared to optical magnifiers and that
advancedvisionlossarerelativelyexpensive.
the duration of reading is increased
Currently, handheld and tabletop CCTVs
range in price from $500 to $3,000. The
FundedbyLawsonResearchInstituteinLon-
amount of time involved in eccentric
don, Ontario (Lawson IRF-063-08 and SJHC
viewing training varies between approxi-
Foundation-Jarmain Family Fund). Our thanks
mately 1 and 14 hours (Gaffney et al.,
go to Shampa Bose for coordinating the study
and Julia Baryla (University of Western On- 2014; Stelmack et al., 2004), with an es-
tario),JaneChen,SusanNieto,SuzanneDecary timated average of 4 hours (Stelmack et
and Roxanne Hazell at CNIB (formerly Cana- al.,2004),anditiscostlyifundertakenby
dian National Institute for the Blind), Toronto,
an occupational therapist or low vision
andLisaHamm,BettyLeeson,SherryMalcho
therapist,sincetheirfeesfor4orupto14
atCNIB,London,Ontario,forhelpinrecruiting
hours of work needs to be paid by the
participantsandcollectingdata.AlexMaoaided
with data analysis and Optelec contributed to- government, private insurance, or the
wards the cost of CCTVs. patients themselves. Interestingly, we do
355
©2017AFB,AllRightsReserved JournalofVisualImpairment&Blindness,July-August2017
not know the dose effect for eccentric ing training (specifically, that it would im-
viewing training—that is, it is not known prove reading speed by at least 10 words
how much is sufficient to gain the near per minute for 1.3M print). Secondary
maximumimprovementandatwhatpoint outcomes were: reading accuracy for
further training does not yield further 1.3Mreadingspeed,andaccuracyfor1M
benefit (Gaffney et al., 2014). There are print, and reading performance for utility
very few randomized clinical trials of and telephone bills and a medicine bottle
the effectiveness of eccentric viewing label.Thefollowinginstruments(question-
training,andevenfewerincomparisonto naires) were also used to measure out-
other interventions (Gaffney et al., 2014;
comes: a reading behavior inventory, the
Hamade, Hodge, Rakibuz-Zaman, &
VFQ-25 plus questions 3 and A4, and the
Malvankar-Mehta, 2016).
Geriatric Depression Scale.
Depending on what is funded or af-
fordable for patients, some clinics may Methods
spend more resources on eccentric
This study was a prospective, random-
viewing training while others spend
ized,parallel-armedclinicaltrial.Theset-
more on electronic devices. In the prov-
ting was CNIB (formerly Canadian Na-
ince of Ontario, Canada, there may be a
tional Institute for the Blind), Toronto,
tendency to use resources for CCTVs,
Canada, which draws patients from a
since these are funded through the As-
broad population base of six million in the
sistive Devices Program (ADP) for eli-
Greater Toronto area in southern Ontario.
gible individuals (Ontario Ministry of
We aimed for an initial sample of 30
Health and Long Term Care, 2008).
participants. Inclusion and exclusion cri-
In an ideal world, one might want to
teria are listed in Table 1. The better end
include both CCTV and eccentric view-
of the visual acuity range was selected so
ing training, but given the reality of lim-
that prospective participants would have
ited resources, agencies need to consider
beenlikelytohavecentralvisionlossand
whichofthesetreatmentsorinterventions
require CCTV or eccentric viewing train-
would have the bigger impact. Assuming
ing, and the poorer limit was chosen be-
that resources are not limitless, it is im-
cause individuals with poorer vision than
portant to know which interventions are
the range specified would have been un-
more effective in order to improve read-
likely to benefit sufficiently from the op-
ing for people with age-related macular
tical devices that would be provided to
degeneration.Toanswerthisquestion,we
initiated a randomized clinical trial of those in the eccentric viewing training
CCTVprovisionversuseccentricviewing group of the study.
training for people with moderate to ad- The study received ethics clearance
vanced age-related macular degeneration through the Western University Research
who had already received a low vision Ethics Board and followed the tenets of
assessment and appropriate optical de- the Declaration of Helsinki on Ethical
vices. The primary hypothesis was that a Principles for Medical Research Involv-
CCTVwouldprovidegreaterincreasesin ing Human Subjects. All participants
reading speed compared to eccentric view- gave written informed consent. This
356
JournalofVisualImpairment&Blindness,July-August2017 ©2017AFB,AllRightsReserved
Table1
Inclusionandexclusioncriteria.
Initialinclusioncriteria Relaxedinclusioncriteria
DiagnosisofAMDreducingvisualacuityto Forreferraltothestudy,thelowerlimitofVAwas
between20/160and20/400(logMAR changedto20/630,soastoallowforthe
0.9–1.3)inthebettereye possibilityofimprovementwithnewglasses
First-timepatientatCNIB Unchanged
Agedover50years Unchanged
NopreviousEVTorCCTV Unchanged
AbletoreadEnglish Unchanged
AhelperathometohelpwithEVT Unchanged
Exclusioncriteria Relaxedexclusioncriteria
Noanticipatedoculartreatmentincluding Noanticipatedoculartreatmentexceptforanti-
anti-VEGFtreatmentoverthecourseof VEGFtreatmentoverthecourseofthestudy
thestudy
Readingnotbeingagoal Unchanged
Minimental(Folstein,etal.,1975)score(cid:4) Unchanged
22(thecopydesigntaskwasenlargedtoa
8.5(cid:5)11page)
Otheroculardiseasereducingcentralvision Unchanged
AMD(cid:2)age-relatedmaculardegeneration;EVT(cid:2)eccentricviewingtraining;CCTV(cid:2)closed-circuit
television.
clinical trial was registered with Clinical- goal, whichever was smaller), and incor-
Trials.gov (2017). porating a 2X acuity reserve (Lovie-
The study protocol is shown in Figure Kitchin, 2011). The final device was the
1. After patients had received a standard one that gave maximum reading fluency
low vision assessment from the low vi- forthetargetprintsize.Demonstrationof
sion therapist at CNIB and met the initial reading stands, line guides, illumination,
eligibility criteria (diagnosis, reading as a and fit-overs for glare was also included.
goal,Englishlanguageability,andnoob- If the optimum aid for reading was a
viouscognitivedifficulties),theywerein- handheld or stand magnifier, this device
vited to participate. The low vision as- was provided through the CNIB. The
sessment included: case history and goal client was directed to go to his or
assessment; distance visual acuity with her eye care practitioner for provision of
theEarlyTreatmentDiabeticRetinopathy spectacle-mounted microscopes (prism
Study (ETDRS) chart; near visual acuity half eye or full field microscope). Fit-
with continuous text print; contrast sensi- overs for glare were available through
tivity assessment; determination of mag- CNIB, and prescription tints were pro-
nification; a trial of optical devices; and vided through the eye care practitioner.
modification of magnification to deter- To increase recruitment, we asked optom-
mine the optimum device and magnifica- etristsinTorontowhowereknowntooffer
tion. The magnification was based on the low vision services to refer potential par-
nearreadingacuityandthegoalprintsize ticipants. In these cases, the initial
(which was 1.3M or the patient’s own low vision assessment and provision of
357
©2017AFB,AllRightsReserved JournalofVisualImpairment&Blindness,July-August2017
Figure 1. Study protocol. VA (cid:2) visual acuity; OH (cid:2) ocular health; MMSE (cid:2) Mini Mental
StatusExam;VFQ(cid:2)VFQ-25;RBI(cid:2)ReadingBehaviorInventory;GDS(cid:2)GeriatricDepres-
sion Scale; EVT (cid:2) eccentric viewing training; CCTV (cid:2) closed-circuit television.
358
JournalofVisualImpairment&Blindness,July-August2017 ©2017AFB,AllRightsReserved
optical devices were undertaken by the randomized to baseline or outcome be-
optometrist. tween participants. For this task, par-
ticipants were allowed to use their pre-
BASELINEVISIT ferred optical devices and were allowed
After a six-week period to allow partici- to adjust the lighting (both were re-
pants time to visit their eye care practi- corded). They were asked to read out
tioners, obtain microscopes, and practice loud as fast as possible, but without
withstandardlowvisiondevices,thepar- sacrificing accuracy. After a demonstra-
ticipants attended CNIB for the baseline tion of the task with a different text,
assessment. An eye care practitioner per- participantswereaskedtoreadthepara-
formed an eye examination to determine graphs. The participant was handed the
eligibility regarding the eye condition or text, and timing with a stopwatch
conditions and ocular history. Then the started when reading began. All errors
following were performed and recorded: were recorded. The number of correct
ETDRS distance visual acuity for both words per minute and the number of
eyes with by-letter scoring, near reading errors per 100 words were calculated.
acuitywiththecurrentspectaclesforboth Three other reading tasks were included,
eyes with the Lighthouse chart, reading which were modeled on Dougherty et
acuity with the current optical aid, the al. (2009). These were: reading the
type of optical aids, Mini Mental Status amount due on a utility bill, microwave
Exam (MMSE), Geriatric Depression Scale, cooking time that was indicated on a
and a check for fulfillment of other in- food package, and the patient’s name
clusion criteria. and dosage (number of tablets per day)
Reading speed and accuracy were on a medicine bottle. The amount of
assessed with elementary school fifth time taken and the accuracy were re-
grade–leveltextsthatwereapproximately corded for each. As for the text reading,
150 words long in either 1.3M or 1M participants were allowed to use their
print. Print that is 1.3M is equivalent to preferred devices and lighting. They
12-pointfont,whichiscommonlyusedin chose their preferred lighting and de-
letters and documents, whereas 1M print vice using a demonstration version of
represents average newsprint. Four pas- each task. There were two versions of
sages were selected that had been previ- each task, which were randomized be-
ously validated in the lab of Dr. Leat. tween participants to baseline or out-
The paragraphs did not contain un- come visit. There was an upper time
common proper nouns or any quoted limit of two minutes for each of these
speech. Out of eight candidate para- tasks, after which time they were re-
graphs, four paragraphs were selected cordedasincorrectifthepatienthadnot
and matched in pairstotheclosestFlesch- responded.
Kincaid grade and the closest (almost Participants were asked verbally about
identical) average reading speeds. Two their reading patterns using Questions 1
were assigned to the 1.3M print and and 3 from the Reading Behavior Inven-
two to the 1M print. The order of the tory (Goodrich, Kirby, Wood, & Peters,
two paragraphs for each print size was 2006).TheNEI-VFQ25(Mangioneetal.,
359
©2017AFB,AllRightsReserved JournalofVisualImpairment&Blindness,July-August2017
CCTV intervention
The 22(cid:3) Clearview tabletop CCTV was used for all participants in the study.
CCTV assessment (within one week of randomization). This was based on the standard
Assistive Devices Program (ADP) assessment procedure, briefly described as follows:
After determining the participant’s reading goals, the minimum magnification to
read thegoal print was assessed, the magnification was recorded, this magnification
was multiplied by 5X, and then the participant was allowed to increase or decrease
magnification until they could comfortably and fluently read the goal print. This
magnification was recorded and the preferred polarity was determined. Last, the
participant was trained to use the CCTV controls.
The CCTV was delivered and set up at the patient’s house.
TheparticipantwasaskedtousetheCCTVforaminimumoftwosessionsof10minutes
each day during a six-week period.
During the trial period, the participant received two phone calls with standard questions
to check the setup and their use of the CCTV.
CCTV (cid:2) closed-circuit television.
Box1
2001) was administered using standard OUTCOMEVISIT
procedure, with the additional instruction The outcome assessment was the same
“if you use a low vision device for an
as the baseline, with the exception that
activity, please answer the questions as
participants in the CCTV group were
thoughyouwereusingit.”Theadditional
asked to use the CCTV for the reading
questions A3 and A4 were included, giv-
texts and those in the eccentric viewing
ing a total of 27 questions. Last, the Ge-
training group were asked to use their
riatric Depression Scale-15 was adminis-
trained preferred retinal locus. For the
tered (Yesavage et al., 1982).
otherreadingtasks(theutilitybill,med-
Afterthesetasks,participantswereran-
icine bottle, and food package), partic-
domized into either the CCTV or the ec-
ipants were allowed to use whichever
centric viewing training group. Random-
device, viewing strategies, and lighting
ization was stratified by presenting visual
that they preferred—that is, those in the
acuity (group 1: 0.90–1.00 logMAR;
CCTV group could choose to use their
group 2: 1.00–1.20; group 3: 1.20–1.3).
optical device or the CCTV. After the
The intervention protocols are shown in
outcome visit, both groups were offered
Boxes 1 and 2. The duration of both in-
the alternative treatment.
terventions was six weeks. The eccentric
viewingtrainingwasundertakenbyalow
vision therapist and the protocol included DATAANALYSIS
components which are commonly in- The reading speed measurements were
cludedbylowvisiontherapists(Stelmack calculatedinlog correctwordspermin-
10
et al., 2004). ute (log CWMP) for analysis. The main
360
JournalofVisualImpairment&Blindness,July-August2017 ©2017AFB,AllRightsReserved
Eccentric viewing training intervention
First EVT training visit (within one week of randomization, duration 1.5 to 2 hours)
• For the better eye, the central scotoma was plotted with a tangent screen.
• Training started with blind-spot awareness (Fletcher & Schuchard, 1997) using the
face method.
• The likely direction of eccentric viewing was determined from the tangent screen
plot, confirmed with the clock or face method. If the field loss was symmetric around
fixation, the participant was asked to move their gaze upwards (since this is one of
the most common and effective directions for EVT (Fletcher & Schuchard, 1997;
Petre et al. 2000). If there was an asymmetric scotoma, the participant was asked to
move their gaze to give the best horizontal area of intact visual field for reading,
extending to the right, and which was nearest to fovea. This gaze shift was
demonstrated with the Amsler chart or tangent screen.
• The direction was confirmed with the clock or face method. The participant was
asked to fixate on the center of the trainer’s face or a clock and to note which section
of the face or numbers on the clock were clearest. Fixation was moved away from
the clearest section. If there was no clear preference, the EVT position would be
chosen to be upwards (as above).
• Using Quillman-type exercises, a print size two times larger than the participant’s
near visual acuity (with their bifocal glasses) was chosen. The participant was trained
to move their fixation in the direction of the determined EVT. The amount of
improvement in near visual acuity for print with EVT was determined.
• Quillman-type exercises in a range of print sizes were provided (including threshold
and 2X smaller and 2X larger than threshold) for both with and without the optical
magnifier.
• Last, the EVT was demonstrated with the participant’s own optical aid.
• For the home training, the participant was asked to practice with an observer and
simply give feedback regarding their level of accuracy.
• They were asked to practice for at least two sessions of 10 minutes each per day for
three weeks.
Theparticipantreceivedonephonecallwithstandardquestionsbetweensessions1and2
to check their compliance and use of an observer.
Second EVT visit (three weeks later, duration 1.5 to 2 hours)
• After reviewing the participant’s progress, steady-eye strategy (Gaffney et al., 2014)
was demonstrated, during which the participant maintains his or her fixation in the
EVT position, rather than making a fixation movement to each word and then to the
EVT position.
• A new range of exercise print sizes was determined and provided. The participant
was asked to practice for another three weeks.
EVT (cid:2) eccentric viewing training; CCTV (cid:2) closed-circuit television.
Box2
361
©2017AFB,AllRightsReserved JournalofVisualImpairment&Blindness,July-August2017
outcomes, which are the reading speeds ideal). We relaxed them as indicated in
and accuracies, were analyzed by two- Table 1.
waytwo-samplet-testsbetweenthepre- After 18 months, to recruit more par-
and post-results for the CCTV and ec- ticipants, a decision was made to move
centric viewing training groups. Paired the study to CNIB, London, Ontario,
t-tests were also undertaken between another large city where CNIB is the
the pre- and post-measures for each largest provider of low vision services.
group separately. Eventually, after 2 years of recruit-
ment, a decision was made to close the
Results study.
Ofthe10participantswhofinallycom-
In this study, we aimed for 30 partici-
pleted the study, 4 were randomized to
pants. After 8 months, only 3 had been
the eccentric viewing training group (av-
recruited, and only 7 by 1 year. We lost
erage age 82, 2 of whom were female)
potential participants at each step of the
and6totheCCTV(averageage83.5,3of
recruitmentprocess.Attheendof2years,
whom were female).
of 145 patients referred to CNIB in that
At baseline, there were no significant
time frame, 35 charts were reviewed as
differences in distance visual acuity,
potential participants, 29 met the inclusion-
reading speeds, reading accuracy, the
exclusion criteria (the others either did
amount of time taken for reading tasks,
not meet the visual acuity requirement,
MMSE, Geriatric Depression Scale, or
hadcommunicationdifficultiessecondary
the VFQ (composite score or any sub-
to English as a second language, or had a scales) between the two groups (p (cid:6)
lack of support to perform the eccentric
0.05). However, there was a significant
viewingtrainingathome).Ofthese29,14
difference in near logMAR visual acu-
were willing to participate (the others
ity with the current glasses (eccentric
were not willing for reasons including viewing training group (cid:2) 1.42, CCTV
unwell health, feeling there would be group(cid:2)1.3,p(cid:2)0.035),whereahigher
travel difficulties (despite being offered a
logMAR score represents poorer visual
taxi service), having too many doctor’s
acuity.
appointments, feeling that they could not Figure 2 shows the primary outcome
ask their child to take them to more ap- variable (reading speed for 1.3M text).
pointments,believingthattheywouldnot Within the CCTV group, there was a sig-
benefit, lacking the patience to do daily nificantimprovementattheoutcomevisit
training, or being uninterested. At the compared to the baseline (paired t-test,
next visit (the baseline visit), we also lost p (cid:2) 0.005), but not within the eccentric
participants whose visual acuity was im- viewing training group (p (cid:2) 0.28). A
proved to better than the inclusion crite- two-sample t-test of the changes (pre-
rion after they received new spectacles post)showedasignificantdifferencebe-
from their eye care practitioner whom tween the groups (p (cid:2) 0.04). Figure 3
they had seen between visits. Early in the shows the secondary outcome of read-
recruitment process, we realized that the ing speed for 1M text. There was no
inclusioncriteriaweretoostrict(although significant change in the reading speed
362
JournalofVisualImpairment&Blindness,July-August2017 ©2017AFB,AllRightsReserved
Figure 2. Primary outcome: Reading speed for 1.3M text (average (cid:7) SD).
within the eccentric viewing training or within the eccentric viewing training
the CCTV group, although this almost group, and there was no significant dif-
reached significance within the CCTV ference in the change (pre-post) be-
group (p (cid:2) 0.051). The between-groups tween the groups. It is noteworthy that
difference in change (pre-post) did not the large variability in the amount of
quite reach significance (p (cid:2) 0.089). time taken at baseline is mostly due to
Similar analyses for accuracy showed one individual who read the bill very
no significant differences. quickly.
Figure 4 shows the results for the Figure 5 shows the changes in near
amount of time taken to read the utility visual acuity. There was a significant
bill. There was, however, a significant improvement in near visual acuity (with
improvement in the amount of time current eyeglasses) within the eccentric
taken within the CCTV group, but not viewing training group, but no signifi-
Figure 3. Secondary outcome: Reading speed for 1M text (average (cid:7) SD).
363
©2017AFB,AllRightsReserved JournalofVisualImpairment&Blindness,July-August2017