14751313, 2011, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1475-1313.2010.00816.x by INASP/HINARI - SERBIA, Wiley Online Library on [28/10/2022].

See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Ophthalmic & Physiological Optics ISSN 0275-5408

Do best practice guidelines improve the legibility of

pharmacy labels for the visually impaired?
Keziah Latham1,2, Sam Waller3 and James Schaitel2
1
Vision and Eye Research Unit, Anglia Ruskin University, Cambridge, UK, 2Department of Vision and Hearing Sciences, Anglia Ruskin University,
Cambridge, UK, and 3Department of Engineering, Cambridge University, Cambridge, UK

Citation information: Latham K, Waller S & Schaitel J. Do best practice guidelines improve the legibility of pharmacy labels for the visually
impaired? Ophthalmic Physiol Opt 2011, 31, 275–282. doi: 10.1111/j.1475-1313.2010.00816.x

Keywords: activities of daily living, drug Abstract

labelling, legibility, low vision, visual
impairment simulation Purpose: Reading pharmacy labels on medications is a challenging task for visu-
ally impaired people. Design for Patient Safety (DfPS) best practice guidelines
Correspondence: Keziah Latham exist on the presentation of information on pharmacy labels, but it is unclear
E-mail address: [email protected] to what extent current labels follow the guidance. It is also unclear whether
labels produced to DfPS guidelines are more accessible to patients with
Received: 19 August 2010; Accepted: 6
impaired vision.
December 2010
Methods: Twenty-four sample labels were obtained from six different pharmacy
chains. Experimental labels were constructed reflecting a typical pharmacy
label, an ideal label constructed to DfPS guidelines, and a large print label. 20
normally-sighted subjects read labels under habitual conditions (mean VA
)0.14 log MAR, Snellen equivalent 6/4.4), and under two conditions of simu-
lated visual impairment (mean VA +0.41 and +0.69 logMAR, Snellen equiva-
lents 6/15.4 and 6/29.4). Outcome measures were speed and accuracy of label
reading.
Results: Median font size for the primary directions on the sample labels was 9.5
point (range 8–10 point), rather than the recommended minimum size of 12
point. In the mild visual impairment condition, using the ideal label improved
accurate reading speed by 58% over the typical label, and using the large print
label improved accurate reading speed by about 100%. In the moderate visual
impairment condition, 65% of subjects were able to see sufficient of the direc-
tions to be able to take the medication appropriately with the ideal label, and
80% with the large print label, as compared to 20% with the typical label.
Conclusions: DfPS best practice guidelines were not fully met by any of the
pharmacy labels in this sample. With unimpaired vision, label design had little
impact on legibility. However, the results provide evidence that preparing phar-
macy labels according to DfPS guidelines improves their relative legibility in
simulated visual impairment. These findings need extending to those with
actual visual impairment, but the current results strengthen the argument for
conformance to DfPS guidelines.

In a recent study,5 it was noted that pharmacy labels

Introduction
constructed to guidelines used in previous research4 were
Reading medicine labels is a challenging visual task that is considered by subjects to be very much easier to read
important to be able to do accurately.1,2 Its importance than their own medication labels. It is known that text
has been recognised in vision research by authors assess- legibility is dependent on many factors including size,11,12
ing functional vision3–7 and by its inclusion in vision- font,13 and contrast,14 but although details of what should
related quality of life questionnaires.8–10 be printed on a dispensed medicinal product are specified

Ophthalmic & Physiological Optics 31 (2011) 275–282 ª 2011 The College of Optometrists 275
 14751313, 2011, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1475-1313.2010.00816.x by INASP/HINARI - SERBIA, Wiley Online Library on [28/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Legibility of pharmacy labels K Latham et al.

in UK law,15 the legislation is silent with regard to the

Simulators
font or text size to be used on dispensing labels (personal
communication: V Ramgoolam, Royal Pharmaceutical In order to objectively examine how design parameters
Society of Great Britain). Guidelines do exist, however, as for printed labels affect their legibility for people with
to what constitutes clear print16 and effective pharmacy vision impairments, visual impairment was produced
label design.17,18 using simulators designed by the University of Cambridge
The purpose of this study was to examine to what Engineering Design Centre.20 The simulators were con-
extent pharmacy labels follow the existing best practice structed from Peltor OX2000 polycarbonate overspec
guidelines of the Design for Patient Safety (DfPS).17 We safety spectacles. The frame was adapted by adding pro-
then examined the influence of label design on the rela- jecting mounts to the sides so that binocular filters (max-
tive legibility of pharmacy labels by examining how well imum dimensions 190 · 85 mm) could be placed in
they could be read by subjects with simulated visual individual cells in front of the eyes, similar to the use of
impairment under standardised conditions. The purpose a trial frame (Figure 1). The simulators use a variable
of this experiment was to determine whether labels pro- number of translucent filters to provide a known and
duced to DfPS guidelines have the potential to be more progressive reduction of the contrast sensitivity function.
accessible to patients with impaired vision. Here we use one and two filters to produce ‘mild’ and
‘moderate’ visual impairment.
Methods
Visual function
Subjects
Twenty visually normal subjects participated. Mean age Vision tests were conducted binocularly with habitual
was 22.4 ± 3.6 years; 16 female, 4 male. Mean habitual bin- correction (simulator overspecs worn over correction, but
ocular distance visual acuity (DVA) was )0.14 ± 0.07 log- with no simulators in place) and with one and two visual
MAR, Snellen equivalent 6/4.4. Two subjects took one impairment simulators in addition to the habitual correc-
prescription medication regularly, the remainder took tion. DVA was assessed using an illuminated 3 m ETDRS
none. All subjects had a Mini Mental State Examination logMAR chart21 scored letter by letter. Contrast sensitivity
(MMSE) score19 of at least 29 out of 30, indicating that (CS) was assessed using Mars charts22 at 50 cm. Near
they were cognitively unimpaired. Informed consent was vision was assessed with an MNRead chart23 viewed at
obtained from all participants and the tenets of the Decla- 40 cm. Reading acuity, critical print size (CPS, the small-
ration of Helsinki were followed. Ethical approval was est print size supporting maximum reading speed), and
obtained from Anglia Ruskin University Ethics Committee. maximum sentence reading speed were calculated accord-
ing to the chart instructions.24

Pharmacy medicine labels

Six pharmacies (three major chains, two independents
and one supermarket) provided standard patient labels
for four medications. DfPS guidelines indicate that most
attention should be paid to the primary directions on
how to take the medicine: these directions should be in
the largest font size, and at least 12 point is recom-
mended. None of the labels in this sample met this crite-
rion: median font size for the primary directions was 9.5
point (range 8–10 point), similar to the size of the other
directions (median 9 point, range 8–10 point), and the
pharmacy name was consistently in a larger font (range
10–11 point). Median font sizes and the range of sizes
seen on the sample labels are given in Table 1. Other sig-
nificant deviations from DfPS guidelines in the sample
were that all labels used centre justified rather than the
Figure 1. The simulator spectacles used to produce ‘mild’ (one simu- recommended left justified text, none of the labels used
lator placed in the frame in front of the eyes) and ‘moderate’ (two bold to highlight the primary directions, and most labels
simulators in place) visual impairment. incorporated a brand label in addition to the pharmacy

276 Ophthalmic & Physiological Optics 31 (2011) 275–282 ª 2011 The College of Optometrists
 14751313, 2011, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1475-1313.2010.00816.x by INASP/HINARI - SERBIA, Wiley Online Library on [28/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
K Latham et al. Legibility of pharmacy labels

Table 1. Print characteristics of the three types of medicine labels used. All drug and patient information was printed using a Postscript-enabled
1200 dpi thermal inkjet printer onto existing actual pharmacy labels of 100 gsm weight which were pre-printed with the Pharmacy name, address
and ‘Keep out of reach’ notice. Print sizes are given in N point as determined from Postscript Arial font samples

Item Typical Ideal Large print

Label size 70 · 35 mm 70 · 35 mm 70 · 35 mm multiples

Font Arial Arial Arial
Justification Centre Left Left
Line spacing Single within; double between Single within; increased gap between 1.5 within and between
Medicine name/dose 9 pt (range 8–9 pt); uppercase; 10 pt; mixed case; bold 12 pt; mixed case; bold
not underlined
Primary directions 9 pt (range 8–10 pt); mixed case; 12 pt; mixed case; caps and bold 14 pt; mixed case; caps
caps highlighting highlighting and bold highlighting
Highlighted directions 9 pt (range 8–10 pt); capitals 12 pt bold capitals 14 pt bold capitals
Additional directions 9 pt (range 8–10 pt); mixed case 8 pt; mixed case 12 pt; mixed case
Patient name 9 pt (range 8–9 pt); plain text 7 pt; bold 12 pt
Date format Numbers only Numbers and letters Numbers and letters
Date size 8 pt (range 8–9 pt) 7 pt 12 pt

name. Conformity to DfPS guidelines was seen in other print characteristics of each label, and the sample label in
areas, in that all labels were of standard size and used each design is shown in Figure 2.
good quality printing. In addition, most used an Arial Variants of each label were created such that each
style (sans serif) font, and capitals to highlight the subject would read a different label for each of the label
primary directions. types (typical, ideal, large print) and visual conditions
Three types of medicine label were then constructed. (habitual, ‘mild’ impairment, ‘moderate’ impairment).
The ‘typical’ label reflects the average parameters of the Product ‘names’ were standardised by using five letter,
high street pharmacy labels (median for font sizes, mode one syllable nouns of high frequency (271–607 instances
for text design such as use of capitals or bold). The ‘ideal’ per million) taken from the British National Corpus.25
label reflects the design characteristics of the Design for
Patient Safety (DfPS) guidelines.17 The ‘large print’ label
Assessment of pharmacy labels
is designed to the characteristics specified for large print
in the Design for Patient Safety guidelines, and those of Trial labels (Figure 2) were presented to participants to
the ‘See it Right’ guidelines.16 Table 1 summarises the familiarise them with the task.3 Test labels were then

(a) (b)

24 THING 60 MG TABLETS 24 Thing 60 mg tablets

Take ONE tablet TWICE a day Take ONE tablet TWICE a day
Avoid consumption of grapefruit
Avoid consumption of grapefruit
Ms Jane Patel 12/08/2009 Ms Jane Patel 12 Aug 2009

(c)

24 Thing 60 mg tablets Avoid consumption of grapefruit

Take ONE tablet TWICE a Ms Jane Patel 12 Aug 2009

day

Figure 2. (a) ‘typical’ trial label. (b) ‘ideal’ trial label. (c) ‘large print’ trial label. The phrase ‘Take ONE tablet TWICE a day’ is described as the ‘pri-
mary directions’, with ‘ONE’ and ‘TWICE’ considered as the ‘highlighted directions’. The phrase ‘avoid consumption of grapefruit’ is the additional
directions.

Ophthalmic & Physiological Optics 31 (2011) 275–282 ª 2011 The College of Optometrists 277
 14751313, 2011, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1475-1313.2010.00816.x by INASP/HINARI - SERBIA, Wiley Online Library on [28/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Legibility of pharmacy labels K Latham et al.

Table 2. Clinical visual parameters of the subjects (n = 20) under the three visual conditions assessed. Most parameters were normally distributed
(Komolgorov–Smirnoff > 0.05), and for these parameters mean ± S.D. are given. For the non-normally distributed parameters, median (25–75%
interquartile range) are given. DVA, distance visual acuity; CS, contrast sensitivity; CPS, critical print size

Near Acuity CPS (Arial N Reading

DVA Near Acuity (Arial N point point speed
(logMAR) CS (logCS) (logMAR) equivalent) CPS (logMAR) equivalent) (wpm)

Habitual )0.14 ± 0.07 1.72 (1.68–1.76) )0.14 ± 0.08 2.3 0.00 (0.00–0.10) 3.1 215 ± 30
Mild impairment 0.41 ± 0.06 1.48 (1.44–1.54) 0.39 ± 0.07 7.6 0.60 (0.50–0.60) 12.4 215 ± 32
(one simulator)
Moderate impairment 0.69 ± 0.05 1.25 ± 0.07 0.66 ± 0.09 14.2 0.95 ± 0.10 27.8 206 ± 34
(two simulators)

Print sizes for the MNRead chart are specified in both logMAR and the equivalent N point size, for comparison to the label font sizes. Conversion
of print sizes between logMAR (MNRead chart, Times New Roman font) and N point (labels, Arial font) was calculated as follows. Legge42 speci-
fies that the relationship between logMAR (40 cm working distance) and x-height is: x-height (cm) = 10logMAR · 40/687.5
The size of a lower case Arial letter ‘x’ was measured with an engineer’s rule for a range of Postscript Arial font sizes. The relationship between
font size in N point and x-height was determined to be: x-height (cm) = (1.867 · N point)/100. Thus, N point = [(10logMAR · 40/687.5] · 100)/
1.867

presented to subjects in randomised order in the centre

160
of an A4 sheet placed on a reading easel at a working dis-
tance of 40 cm. Overhead fluorescent lighting and a day-
Reading speed (word per minute correct)
140
light task lamp were used to provide luminance of
100 cdm)2. The subject’s task was to read aloud the med- 120 Typical
ication and instructions, which was recorded. Errors and
Ideal
time taken were assessed by two examiners independently. 100 Large print

80
Results
Table 2 shows the clinical visual parameters of the sub- 60
jects under habitual conditions, and with the two levels of
simulated visual impairment. Both visual acuity and con- 40
trast sensitivity were reduced by the simulators, but not
reading speed. Variability in measurements did not 20
increase as the number of simulators increased, indicating
a consistent effect of the simulators across all participants. 0
Habitual 1 simulator 2 simulator
Figure 3 shows the accurate reading speed for the three Visual condition
label types under the three visual conditions. The influ-
ence of label design on reading speed in words per min- Figure 3. Mean reading speed in words per minute correct for each
label and visual condition. Error bars show standard deviation. Open
ute correct was highly significant (F1.9, 36.5 = 32.8,
bar: typical label; dotted bar: ideal label; striped bar: large print label.
p < 0.0001) as was the influence of vision (F1.5, 29.1 = 266,
p < 0.0001), and the interaction between label design and
vision (F2.4, 45.8 = 9.9, p < 0.0001; all repeated measures with mild visual impairment, reading speed was depen-
anova using Greenhouse-Glasser correction). dent on label design (Figure 3). Mean reading speed for
With unimpaired (habitual) vision, information was the typical label was 53 wpm correct, while the ideal label
accessed accurately (no errors recorded) and fluently was read at 84 wpm correct, a 58% improvement. The
(>80 wpm11) regardless of label design. With mild visual large print label was read at 107 wpm correct, a further
impairment (one simulator), accuracy of reading the criti- 27% improvement on the ideal label, and twice the speed
cal information on the label remained essentially intact, of reading a typical label.
with 90% of observers able to read sufficient of the pri- With moderate visual impairment (two simulators),
mary directions to be able to appropriately take the medi- reading speed was slow (<25 wpm correct) with all label
cation with the typical label, 100% with the ideal label, designs (Figure 3). Accuracy varied with label design as
and 95% with the large print label (Figure 4). However, shown in Figure 5. The typical label was read poorly, with

278 Ophthalmic & Physiological Optics 31 (2011) 275–282 ª 2011 The College of Optometrists
 14751313, 2011, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1475-1313.2010.00816.x by INASP/HINARI - SERBIA, Wiley Online Library on [28/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
K Latham et al. Legibility of pharmacy labels

100 100
90 90

80 80
70 Able to read medicine
70

% of observers
name
% of observers

Able to read medicine 60

60 Able to read highlighted
name
directions
50 Able to read highlighted 50
Able to take medication
directions
40 appropriately
40 Able to take medication
appropriately 30
30
20
20
10
10
0
0 Typical label Ideal label Large print label
Typical label Ideal label Large print label
Figure 5. Accuracy of label reading in the moderate visual impair-
Figure 4. Accuracy of label reading in the mild visual impairment ment (two simulator) condition for the three label designs. The medi-
(one simulator) condition for the three label designs. The medicine cine name and highlighted directions are illustrated in Figure 2, and
name and highlighted directions each had to be read accurately for had to be read accurately for the observer to be judged successful. In
the observer to be judged successful in these categories. In order be order to be judged able to take the medication appropriately, observ-
judged able to take the medication appropriately, observers had to ers had to read sufficient of the primary directions to take the correct
read sufficient of the primary directions to take the correct number of number of tablets at the appropriate frequency. Font sizes of relevant
tablets at the appropriate frequency. In the trial labels illustrated in parts of the labels are as follows: Typical label – medicine name and
Figure 2, the medicine name is ‘Thing’, the highlighted directions are highlighted directions: 9 point capitals; remainder of primary direc-
‘ONE’ and ‘TWICE’, and the primary directions are the phrase ‘Take tions: 9 point lower case. Ideal label – medicine name: 10 point bold
ONE tablet TWICE a day’. Font sizes of relevant parts of the labels are lower case; highlighted directions: 12 point bold capitals; remainder
described with Figure 5. of primary directions: 12 point lower case. Large print label – medi-
cine name: 12 point bold lower case; highlighted directions: 14 point
only 20% of observers able to see sufficient of the direc- bold capitals; remainder of primary directions: 14 point lower case.
tions to take the medication appropriately and none able
to read the name of the medicine. The ideal label fared mately 440 000 people in the UK have DVA of this level
better in that 65% of observers were able to read the pri- (6/18–6/60), with a further 200 000 having vision worse
mary directions sufficiently to take the medication appro- than 6/60.31
priately, and 80% of observers could read this Label design had little impact on performance with no
information with the large print label. visual impairment (habitual condition). All label fonts
were larger than subjects’ mean critical print size
(0.01 logMAR is equivalent to approximately 3 point
Discussion
Arial font), accuracy is almost perfect and reading speed
The high street pharmacy labels assessed did not conform was limited by non-visual factors such as speed of verbali-
to DfPS guidelines (Table 1). In particular, the font size sation.
of the primary directions for taking the medication were With ‘mild’ visual impairment (one simulator), accu-
smaller (median 9.5 point, range 8–10 point) than sug- racy remained largely intact across all label types, but
gested (12 point). reading was slower with the typical label. Accuracy was
Clinical visual performance of the observers in the intact because print size was at or larger than near acuity
habitual condition (Table 2) relates closely to published (8 point from Table 2). Reading speed was slow for the
normative values.26–28 One simulator (‘mild’ impairment) typical label because the font size (largely 9 point, see
reduced DVA to the Snellen equivalent of 6/15.4, at about Table 1) was smaller than the critical print size (11.3
the limit for driving in the UK,29,30 reduced CS to a level point from Table 2). The font sizes on the ideal and large
consistent with ‘noticeable loss’,11 and reduced near acu- print labels (largely 10–14 point, see Table 1) were closer
ity to the equivalent of N8 at 40 cm. Approximately to the average critical print size required for reading at
1.14 million people in the UK have DVA of this level (6/ maximum speed. Although accessing information accu-
12-6/18).31 Two simulators (‘moderate’ impairment) rately is the most important aspect of reading medicine
reduced DVA to 6/29.4, not poor enough for registration labels, reading speed indicates the ease with which infor-
as sight impaired.32 CS was reduced further than with mation can be accessed, and thus the relative legibility of
one simulator, but still consistent with ‘noticeable loss’, the label. A label that can be read faster and more easily
and near acuity was equivalent to N14 at 40 cm. Approxi- is less likely to require compensatory strategies to be

Ophthalmic & Physiological Optics 31 (2011) 275–282 ª 2011 The College of Optometrists 279
 14751313, 2011, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1475-1313.2010.00816.x by INASP/HINARI - SERBIA, Wiley Online Library on [28/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Legibility of pharmacy labels K Latham et al.

employed in order to read it, or can be read by people label, they will not necessarily understand it,34 and there
with more significant levels of visual impairment. is little evidence to date that more legible labels improve
With ‘moderate’ visual impairment (two simulator adherence to medication2,34 or give measurable improve-
condition), performance on the label reading task was ments in health outcomes.18 However, accessing the
badly affected. Few subjects could read any information printed information on a pharmacy label is the first
on the typical label as the print size (nine point, see barrier to a patient being able to take their medications
Table 1) was smaller than the 14 point average reading correctly, and making such information accessible to as
acuity (Table 2). Ideal and large print labels were much wide a proportion of the public as possible is essential.
better in terms of allowing access to the primary direc- Providing information in a format that can be read by
tions for taking the medication, as the print size was clo- visually impaired people is one requirement of the Dis-
ser to the average acuity threshold: 65% could read the ability Discrimination Act,35 and is highlighted by the
12 point directions on the ideal label, and 80% the 14 RNIB ‘Losing Patients’ campaign36 to provide health
point directions on the large print label. However, few information, including pharmacy medication labelling, in
observers could read the medication name on any of the accessible formats.
labels (0% for the typical label; 5% for the ideal label; Note that neither the ‘mild’ nor ‘moderate’ visual
20% for the large print label). Being unable to read medi- impairment conditions simulate vision poor enough for
cation names could lead to confusion between multiple registration as visually impaired, yet performance on this
medications for those who take more than one. task of daily living deteriorates noticeably under both
The use of capitals helped to emphasise important conditions. In agreement with other authors,37 the
information in print presented at around the acuity limit results suggest that low vision rehabilitation services to
in the moderate impairment condition: 90% of observers assist in providing compensatory strategies are of benefit
could read the highlighted directions on the ideal label in to patients with vision loss that are ineligible for regis-
12 point bold capitals, whereas only 20% could read the tration as visually impaired under current UK guide-
medicine name on the large print label in 12 point bold lines.
lower case (Figure 5). These findings are consistent with The study design used a standardized methodology,
those of Arditi & Cho33 who showed, also with Arial font, including a fixed working distance of 40 cm and simu-
that thresholds for upper case text were 0.1 log units lated visual impairment in young normally sighted
smaller than those for lower case text in normally sighted observers. The advantage of such a methodology is that it
observers and those with low vision. Such a difference fits allows the assessment of the relative legibility of different
with the present findings. Acuity for lowercase MNRead label designs, without increased noise in the data from
text in the moderate impairment condition was 0.66 log- observers with different pathologies, ages, types of visual
MAR (N14.2 equivalent), and so N12 lowercase text on impairment or working distances. Labels which are rela-
the labels could not generally be read (20% success). tively more legible under standard conditions should
However, if acuity for uppercase text is assumed to be 0.1 require fewer compensatory strategies to be employed for
logMAR better, then uppercase acuity would be around them to be accessed by visually impaired people. How-
0.56 logMAR (N11.3 equivalent), which would be suffi- ever, these results should be considered as a first step as
cient to read N12. Such text was indeed read with 90% there are limitations in having used such a standardized
success. methodology. These limitations are now discussed.
The large print label was easier (faster and more accu- By using simulated visual impairment, there are limita-
rate) to read than the ideal label, but has the disadvantage tions to the application of the findings. The findings will
of requiring the information to be tiled over two or more be relevant across a variety of real visual conditions that
labels, depending on the additional directions required. reduce the contrast sensitivity function including cata-
Since it is only recommended to leave space for a single racts, uncorrected presbyopia, and reduced retinal sensi-
35 · 70 mm label on medication packaging,17 multiple tivity, but cannot necessarily be generalised to people
tiled labels may not fit or could obscure important areas with a central field obstruction, such as caused by macu-
of the original packaging. DfPS guidelines suggest placing lar degeneration. Reading with a central scotoma places
an ‘ideal’ label on the packaging, and if a patient requires additional restrictions on performance11,38,39 which are
it, to place more easily accessible information, such as not considered in these results. Since macular degenera-
large print labels, on an additional A4 sheet. tion is the commonest cause of registerable visual impair-
Being able to read a medicine label is a necessary, but ment in the UK40 accounting for approximately 55% of
not necessarily sufficient, step for someone to be able to registrations, further work is needed to clarify how poor
understand the directions on their medication and take it legibility of pharmacy labels affects this specific and sig-
appropriately. Even if a person can read their pharmacy nificant group of people.

280 Ophthalmic & Physiological Optics 31 (2011) 275–282 ª 2011 The College of Optometrists
 14751313, 2011, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1475-1313.2010.00816.x by INASP/HINARI - SERBIA, Wiley Online Library on [28/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
K Latham et al. Legibility of pharmacy labels

By using a fixed working distance, this study did not References

consider the impact of compensatory strategies on label
1. Shrank WH, Agnew-Blais J, Choudhry NK et al. The vari-
legibility, including the simple strategy of reducing work-
ability and quality of medication container labels. Arch
ing distance to increase angular print size. Whether the Intern Med 2007; 167: 1760–1765.
differences in label legibility seen in this study could be 2. Shrank WH, Gleason PP, Canning C et al. Can improved
overcome by using working distances closer than 40 cm medication labeling influence adherence to chronic medi-
cannot be seen from these results. As most people with cations? An evaluation of the Target pharmacy label. J Gen
visual impairment are older and therefore presbyopic, Intern Med 2009; 24: 570–578.
enlarging print by moving it closer may make it easier to 3. Dougherty BE, Martin SR, Kelly CB et al. Development of
read, or may make it more blurred and difficult to read if a battery of functional tests for low vision. Optom Vis Sci
they do not have suitable refractive correction for the 2009; 86: 955–963.
working distance. Further work with older patients is 4. Haymes SA, Johnston AW & Heyes AD. The development
therefore needed to clarify the generalizability of the pres- of the Melbourne low-vision ADL index: a measure of
ent findings. Note however, that low vision rehabilitation vision disability. Invest Ophthalmol Vis Sci 2001; 42: 1215–
(provision of spectacles, optical LVAs and training) has 1225.
been shown to improve the ability to read medicine 5. Latham K & Usherwood C. Assessing visual activities of
labels.3,41 daily living in the visually impaired. Ophthalmic Physiol
Opt 2010; 30: 55–65.
6. Owsley C, McGwin G, Sloane ME, Stalvey BT & Wells J.
Conclusions Timed instrumental activities of daily living tasks: relation-
The results of this study suggest that following DfPS ship to visual function in older adults. Optom Vis Sci
guidelines improves the relative legibility of medicine 2001; 78: 350–359.
7. Windham BG, Griswold ME, Fried LP et al. Impaired
labels for those with simulated impaired contrast sensitiv-
vision and the ability to take medications. J Am Geriatr
ity function. In particular, the font size of the primary
Soc 2005; 53: 1179–1190.
directions on a label should be at least 12 point Arial,
8. Mangione CM, Berry S, Spritzer K et al. Identifying the
rather than the current 8–10 point. ‘Ideal’ labels follow-
content area for the 51-item National Eye Institute visual
ing DfPS guidelines allowed people with simulated mildly function questionnaire. Arch Ophthalmol 1998; 116: 227–
impaired vision to read labels 57% faster than with a typ- 233.
ical current label under standardised conditions. For peo- 9. Massof RW, Ahmadian L, Grover LL et al. The Activity
ple with simulated moderate visual impairment, following Inventory: an adaptive visual function questionnaire.
DfPS guidelines improved the legibility of the primary Optom Vis Sci 2007; 84: 763–774.
directions on medicines from being accessible to 20% of 10. Wolffsohn JS & Cochrane AL. Design of the low vision
subjects with a typical label to 65% of subjects with an quality-of-life questionnaire (LVQOL) and measuring the
ideal label, under standardised conditions. However, with outcome of low-vision rehabilitation. Am J Ophthalmol
this level of vision even large print labels were difficult 2000; 130: 793–802.
to access fully without compensatory strategies. Provision 11. Whittaker SG & Lovie-Kitchen J. Visual requirements for
of information in alternative ways should be consid- reading. Optom Vis Sci 1993; 70: 54–65.
ered, and not just for those with registerable visual 12. Legge GE, Pelli DG, Rubin GS & Schleske MM. Psycho-
impairment. physics of reading. I. Normal vision. Vision Res 1985; 25:
In conclusion, although the findings need extending to 239–252.
older patients and those with actual visual impairment, 13. Mansfield JS, Legge GE & Bane MC. Psychophysics of
the results suggest that pharmacies should be recom- reading. XV. Font effects in normal and low vision. Invest
mended to adopt the DfPS guidelines in their labelling Ophthalmol Vis Sci 1996; 37: 1492–1501.
practices to assist the 1.8 million people in the UK with 14. Rubin GS & Legge GE. Psychophysics of reading. VI.
The role of contrast in low vision. Vision Res 1989; 29:
impaired vision to access the information on their phar-
79–91.
macy medicine labels more easily.
15. The Medicines for Human Use (Marketing Authorisations
Etc.) Regulations 1994. Available from http://www.
Acknowledgements legislation.gov.uk/uksi/1994/3144/contents/made, accessed
5/1/11.
JS supported by a Nuffield Vacation Scholarship. With 16. RNIB. See It Right: Making Information Accessible for
thanks to Prof Gordon Legge and Dr J Stephen Mansfield People with Sight Problems. RNIB: Peterborough, 2006.
for advice on font sizes.

Ophthalmic & Physiological Optics 31 (2011) 275–282 ª 2011 The College of Optometrists 281
 14751313, 2011, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1475-1313.2010.00816.x by INASP/HINARI - SERBIA, Wiley Online Library on [28/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Legibility of pharmacy labels K Latham et al.

17. National Patient Safety Agency. Design for patient safety: a assoc-optometrists.org/services/visual/visual_1010160976.
guide to the design of dispensed medicines. 2007. Avail- html, accessed 30/3/10.
able online at: http://www.nrls.npsa.nhs.uk/resources/ 30. Drasdo N & Haggerty C. A comparison of the British
collections/design-for-patient-safety. number-plate and Snellen vision tests for car drivers.
18. Shrank WH, Avorn J, Rolon C & Shekelle P. Effect of con- Ophthalmic Physiol Opt 1980; 1: 39–54.
tent and format of prescription drug labels on readability, 31. Access Economics. Future Sight Loss UK (1): The Eco-
understanding, and medication use: a systematic review. nomic Impact of Partial Sight and Blindness in the UK
Ann Pharmacother 2007; 41: 783–801. Adult Population. RNIB: Peterborough, 2009.
19. Folstein M, Folstein S & McHugh P. Mini-mental state: 32. Jackson AJ. Epidemiology of low vision. In: Low Vision
a practical method for grading the cognitive state of Manual (Jackson AJ & Wolffsohn JS, editors), Elsevier
patients for the clinician. J Psychiatr Res 1975; 313: Butterworth Heinemann: Edinburgh, 2007; pp. 1–26.
1419–1420. 33. Arditi A & Cho J. Letter case and text legibility in normal
20. Goodman-Deane J, Waller SD & Clarkson PJ. Simulating and low vision. Vision Res 2007; 47: 2499–2505.
impairment. In: Proceedings of (re)Actor3, the Third Inter- 34. Moisan J, Gaudet M, Gregoire J-P & Bouchard RJ. Non-
national Conference on Digital Live Art (Bryan-Kinns N, compliance with drug treatment and reading difficulties
Lloyd T & Sheridan J. editors), Liverpool, 2008. pp. 21–22. with regard to prescription labelling among seniors.
Available at: http://www.digitalliveart.com/reactor3.pdf, Gerontology 2002; 48: 44–51.
accessed 10/1/11. 35. Disability Discrimination Act. 1995. Available at: http://
21. Ferris FL, Kassoff A, Bresnick GH & Bailey IL. New visual www.legislation.gov.uk/ukpga/1995/50/contents, accessed
acuity charts for clinical research. Am J Ophthalmol 1982; 7/11/10.
94: 92–96. 36. RNIB. Losing patients. 2010. Available from: http://
22. Arditi A. Improving the design of the letter contrast sensi- www.rnib.org.uk/accesstohealth, accessed 7/11/10.
tivity test. Invest Ophthalmol Vis Sci 2005; 46: 2225–2229. 37. Ryan B & Margrain T. Registration for people with visual
23. Legge GE, Luebker A & Ahn SJ. Printed cards for measur- impairment: fit for purpose? Br J Ophthalmol 2010:
ing low-vision reading speed. Vision Res 1995; 35: 1939– doi:10.1136/bjo.2009.178558.
1944. 38. Legge GE, Rubin GS, Pelli DG & Schleske MM. Psycho-
24. Mansfield JS & Legge GE. MNRead acuity charts. Available physics of reading. II. Low vision. Vision Res 1985; 25:
from: http://vision.psych.umn.edu./groups/gellab/ 253–266.
MNREAD, accessed 13/3/08. 39. Legge GE, Ross JA, Isenberg LM & LaMay JM. Psycho-
25. British National Corpus. Available from: http://ucrel.lancs. physics of reading. XII. Clinical predictors of low-vision
ac.uk/bncfreq/flists.html, accessed 9/7/09. reading speed. Inv Ophthalmol Vis Sci 1992; 33: 677–687.
26. Dougherty BE, Flom RE & Bullimore MA. An evaluation 40. Bunce C & Wormald R. Leading causes of certification for
of the Mars letter contrast sensitivity test. Optom Vis Sci blindness and partial sight in England & Wales. BMC
2005; 82: 970–975. Public Health 2006; 6: 58.
27. Elliot DB, Yang KCH & Whitaker D. Visual acuity changes 41. Markowitz SN, Kent CK, Schuchard RA & Fletcher DC.
throughout adulthood in normal, healthy eyes: seeing Ability to read medication labels improved by participa-
beyond 6/6. Optom Vis Sci 1995; 72: 186–191. tion in a low vision rehabilitation program. J Vis Impair
28. Subramanian A & Pardhan S. The repeatability of Blind 2008; 102: 774–777.
MNREAD acuity charts and variability at different test 42. Legge GE. Psychophysics of Reading in Normal and Low
distances. Optom Vis Sci 2006; 83: 572–576. Vision. Lawrence Erlbaum Associates: Mahwah, New
29. Association of Optometrists. Vision standards: motor Jersey, 2007.
vehicle drivers, 2006. Available from: http://www.

282 Ophthalmic & Physiological Optics 31 (2011) 275–282 ª 2011 The College of Optometrists