See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/279726828

How to interpret visual fields

Article in Practical Neurology · July 2015

DOI: 10.1136/practneurol-2015-001155 · Source: PubMed

CITATIONS READS

16 222

2 authors, including:

Gordon Terence Plant

University College London
526 PUBLICATIONS 19,849 CITATIONS

SEE PROFILE

All content following this page was uploaded by Gordon Terence Plant on 21 September 2022.

The user has requested enhancement of the downloaded file.

 HOW TO UNDERSTAND IT

How to interpret visual fields

Pract Neurol: first published as 10.1136/practneurol-2015-001155 on 3 July 2015. Downloaded from http://pn.bmj.com/ on September 21, 2022 by guest. Protected by copyright.
Sui H Wong,1,2 Gordon T Plant1,2,3

1
Department of Neuro- INTRODUCTION therefore, rod photoreceptors do not con-
ophthalmology, Moorfields Eye Imagine you are assessing a patient with tribute to the findings.
Hospital, London, UK
2
Medical Eye Unit, St Thomas’ visual difficulties or optic disc swelling. 2. The normal field of vision extends to
Hospital, London, UK After a bedside visual field examination approximately 60° nasally, 90° temporally,
3
National Hospital for Neurology with waggling fingers and even a red 60° superiorly and 70° inferiorly.
and Neurosurgery, London, UK
hatpin, you decide that there is an abnor- 3. The blind spot indicates the location of
Correspondence to mality. After requesting quantified visual the optic nerve head—an area with no
Dr Sui H Wong, Consultant in field tests, the patient returns with a photoreceptors—in the temporal part of
Neuro-ophthalmology, black and white printout with numbers the visual field.
Moorfields Eye Hospital,
162 City Road, London (eg, Humphrey fields) or coloured lines 4. Anything obstructing the travel of light
EC1V 2PD, UK; on a sheet (eg, Goldmann fields). Where towards the retina may affect the field tests,
[email protected] is the report you ask? There is none! for example, lens opacity (cataract), ptosis
Static perimetry uses flashing stationary (if not taped away from the pupil) or the
Accepted 31 May 2015
Published Online First lights. This can be automated (eg, evenly rim of a correcting lens (test artefact)
3 July 2015 spaced points on a grid) or manual (eg,
as a small part of Goldmann test: detailed Goldmann field test
later). The Humphrey field analyser is by During a Goldmann field test, the patient
far the most commonly used for auto- positions their eye opposite the centre of a
mated static perimetry, although there are white hemispherical bowl (figure 2). The
also other machines such as Octopus and patient fixates upon the central target
Henson. Later, we describe in detail the 33 cm away, while the examiner sits
interpretation of Humphrey perimetry. opposite viewing through an eyepiece to
Kinetic perimetry uses a moving illumi- ensure good fixation throughout the test.
Editor’s choice nated target and is done either manually The examiner moves an illuminated white
Scan to access more
free content (eg, Goldmann) or on an automated target from the periphery towards the
machine (eg, Octopus). Goldmann centre, and the patient presses a buzzer to
machines are no longer manufactured, indicate when they first see the target.
being slowly replaced by Octopus This is repeated from different directions
machines. Nevertheless, Goldmann —allowing the examiner to plot the
remains the most commonly used kinetic patient’s field of vision—using targets
perimetry, and so we use this here to illus- varying in size and brightness. The exam-
trate interpretation of kinetic fields. The iner plots the blind spot and the edges of
principles for interpreting Goldmann also scotomas in a similar way, with the patient
apply to results from Octopus machines. pressing the buzzer to indicate when they
It is beyond the scope of this paper to first see the light target moving from a
cover the neuroanatomical localisation of blind to a seeing area. The examiner also
visual field defects. Instead we recom- performs static testing—involving the
mend two excellent recent reviews.1 2 brief appearance of the stationary light
Skilled interpretation of visual field tests target—in the four quadrants within the
requires a good grasp and application of central 20° or so, marking a tick on the
this prior knowledge. chart when the patient sees the target and
a cross if they do not.
Useful aspects of eye anatomy The target sizes are labelled with three
1. The fovea is the area of greatest visual sen- alphanumeric digits, for example, ‘V4e’.
sitivity, where the cone photoreceptor 1. The first digit is a Roman numeral (I–V),
density is at its highest. The visual sensitiv- indicating the size of the target, for example,
ity slopes off further from the fovea. This V is equivalent to a target diameter of
drop in sensitivity can be visualised as a 9.03 mm. With every drop in number (eg,
To cite: Wong SH, Plant GT. hill, with the fovea is at the peak (figure 1). from V to IV) the diameter halves.
Pract Neurol 2015;15: Conventional perimetry is carried out 2. The second digit is an Arabic number
374–381. under photopic (well lit) conditions, and (1–4), indicating the brightness of the

374 Wong SH, Plant GT. Pract Neurol 2015;15:374–381. doi:10.1136/practneurol-2015-001155

 HOW TO UNDERSTAND IT

Pract Neurol: first published as 10.1136/practneurol-2015-001155 on 3 July 2015. Downloaded from http://pn.bmj.com/ on September 21, 2022 by guest. Protected by copyright.
Figure 1 Normal hill of vision.

stimulus: the larger the number the higher the examiners test smaller or wider visual angles;
luminance. however, the wider the visual angle tested, the more
3. The third digit is a letter (a–e), indicating a finer calibration coarse the grid, and hence the greater the likelihood
of luminance. ‘4e’ is equivalent to 10-decibel (dB) bright- of missing small scotomas. The 24-2 assesses the
ness; each consecutive drop in number represents a 5 dB central 24° with a 54-point grid; 10-2 assesses the
change and each drop in letter represents a 1 dB change. central 10° with a 68-point grid; and 30-2 assesses the
By convention, the examiner maps three isopters: central 30° with a 76-point grid.
lines of equal sensitivity to targets of a specified size The examiner plots the hill of vision based upon
and luminance. The first isopter, mapping the farthest the threshold for detecting different target luminance;
peripheral vision, requires the largest and brightest as visual sensitivity improves towards the fovea, so the
target ‘V4e’. Another isopter is mapped in the central detection threshold for the target decreases. Unlike
30° of vision, and a third isopter is intermediate Goldmann, the target size stays the same during the
between these two. The isopter lines therefore show test, with a default size equivalent to Goldmann size
the margins of different visual sensitivity, analogous to III targets. It is rare to need a different default size.
the contour lines of a map marking different eleva- The Swedish interactive threshold algorithm (SITA)
tions. This allows us to visualise the hill of vision. The is the most commonly used test algorithm,3 designed
base of the hill represents the area at the periphery to reduce the time to complete a test; a short test dur-
with least visual sensitivity, detecting only the largest ation limits the likelihood of errors from patient
and brightest target. As we move up towards the peak fatigue. SITA starts by determining the visual stimula-
of the hill, the visual sensitivity increases and the tion thresholds at the four quadrants. If the patient
patient sees smaller and dimmer targets. sees the initial stimulus, the examiner reduces its
brightness to the level where it is no longer seen.
Humphrey field test Conversely, if the patient does not see the stimulus, its
The same principles apply to the Humphrey test as to brightness is increased to find this threshold. The
the Goldmann test, but instead with static light stimula- examiner adjusts the initial brightness at adjacent
tion. The machine can also be programmed to perform points according to the threshold of its neighbouring
kinetic tests though we have no experience with this. point. During the test, the examiner retests some loca-
The illuminated targets appear for 200 ms at prede- tions to determine reliability (see false-negative errors,
termined locations on a grid. Humphrey tests are below). At completion, the computer generates a stat-
widely used in glaucoma clinics, the most common set istical analysis, which is compared to an age-matched
up being to test the central 24° (‘24-2’ setting). Some normal population.

Wong SH, Plant GT. Pract Neurol 2015;15:374–381. doi:10.1136/practneurol-2015-001155 375

 HOW TO UNDERSTAND IT

Pract Neurol: first published as 10.1136/practneurol-2015-001155 on 3 July 2015. Downloaded from http://pn.bmj.com/ on September 21, 2022 by guest. Protected by copyright.
Figure 2 Goldmann machine. The patient’s eye is positioned at the centre of a white hemispheric bowl, with the examiner looking
through an eyepiece to ensure good fixation. A white light (indicated by yellow arrow in (A) is brought in from the periphery into the
patient’s field of vision. The examiner does this by controlling connecting levers (indicated by orange arrows in A and B). The patient
presses a buzzer when the light target is seen (blue arrow).

HUMPHREY OR GOLDMANN? A. Is the isopter smooth, as expected for a normal

The choice may depend upon local availability. The hill of vision?
Humphrey is slightly less operator-dependent than the B. Is there restriction? Examples would be a nasal
Goldmann and has the advantage of numbers to indi- step in papilloedema or an altitudinal defect in
cate reliability of the test. The Goldmann tests periph- anterior ischaemic optic neuropathy.
eral fields better, may be more patient-friendly for C. Are the isopters spaced, as expected for the
those who are hesitant on the Humphrey, and is par- normal hill of vision? (1) A tiny central field with
ticularly useful for central scotoma, as it is easier to ‘stacked’ isopters—very close to one another as in
manage fixation losses. As a rule of thumb, when a steep hill—usually denotes functional overlay
monitoring disease, it is sensible to use the same test (figure 7); however, patients with genuine retinal
as was used previously. Both tests can complement and striate cortex lesions may also have stacked
each other, confirming deficit patterns when in doubt. isopters. (2) Isopter lines that cross always indicate
unreliable test: isopters cannot cross since this
Interpreting the Goldmann field test would indicate two different sensitivities at one
The key to interpreting Goldmann visual fields is to location. (3) Spiralling isopters suggest functional
keep in mind the normal hill of vision (figure 1) and visual loss and indicate a steady decline in sensitiv-
how it compares with the patient’s results. The skill is in ity during the test.
identifying patterns and observing any change with D. Are there scotomas? It is important to correlate
repeated tests. This may require experience to be adept, this with the patient’s symptoms and clinical
though the following checklist may help (figure 3): (bedside) examination.
1. Patient name and date of test: a good habit always to 4. Is the blind spot size enlarged? This is particularly rele-
check the test belongs to your patient! vant in papilloedema (figure 5). The normal blind spot
2. What is the largest peripheral field (V4e)? This can vary size is oval, roughly 10° in diameter, and located 10–20°
according to age and test response. It normally extends temporally from the central fixation point.
to approximately 60° nasally, 90° temporally, 60° super- 5. Is the central field affected? Was static testing done (indi-
iorly and 70° inferiorly. Thus, the superior aspect of the cated with a tick when the patient saw the target)?
field is usually less sensitive than the inferior field, 6. Is any defect monocular or binocular, when comparing
though ptosis could also artefactually reduce it. the fields for each eye? If binocular, is the defect hom-
3. Is there any distortion to the ‘contours’? (Contours are onymous or heteronymous?
the smaller isopters corresponding to targets that are Any comments written about patient fixation or
either smaller or dimmer or both). attention also help.

376 Wong SH, Plant GT. Pract Neurol 2015;15:374–381. doi:10.1136/practneurol-2015-001155

 HOW TO UNDERSTAND IT

Pract Neurol: first published as 10.1136/practneurol-2015-001155 on 3 July 2015. Downloaded from http://pn.bmj.com/ on September 21, 2022 by guest. Protected by copyright.
Figure 3 Interpreting the Goldmann visual field.The chart is viewed from the perspective of the patient looking into the test bowl,
as if patient is looking into the paper. Suggested checklist to review the Goldmann fields systematically (see text for details):
1. Patient name and identification number, date of test.
2. The largest isopter, that is, peripheral field.
3. The other isopters—any distortion to the ‘contours’ of the hill of vision? Any scotomas?
4. Blind spot.
5. Central vision.
6. If there is an abnormality, is it monocular or binocular? If binocular, is it homonymous or heteronymous?
7. Other, for example, comments about fixation or attention.
This is an example of normal Goldmann fields. In contrast, this patient did not perform well on the Humphrey visual fields, with poor
reliability and cloverleaf pattern (figure 4).

▸ Small pupil size, ptosis and incorrect positioning of a corresponding increase in the points tested
correcting lens may affect the peripheral field. (76-point grid for 30°); thus, this is a longer test
▸ Inadequate correction of refraction error for the viewing with the risk of more patient errors.
distance (33 cm) may affect the central field. b. Was it a threshold or a screening test?
Screening tests use suprathreshold targets of single
Interpreting the Humphrey field test
luminance and in the past were particularly useful
We suggest the following framework to interpret
because full threshold tests were time consuming.
Humphrey test results (figure 4), structured to answer
However, SITA threshold tests have superseded
three questions:
these, reducing test times (equivalent to the time
1. Is this the correct test?
taken for screening tests) without losing sensitivity.
A. Name and patient number: confirm that the output
belongs to your patient! 2. Can I rely on this test?
B. Date of test: is this the output of interest? that is, A. False-positive errors
timing in relation to symptoms. False-positive errors identify ‘trigger happy’ patients
C. To which eye does this output correspond? Correlate who respond in the absence of light stimulus. They
the results with the history and clinical examination. are calibrated according to the patient’s overall
Beware of fields that are mounted incorrectly: the responses, therefore detecting when responses occur
conventional way of mounting is to place the left too soon after presenting a stimulus is. A false-positive
chart on the right and vice versa, ie, as if the patient is rate of >15% compromises test results.4
looking into the chart. B. False-negative errors
D. What test was performed? This is particularly import- A false negative is the failure to respond to a relatively
ant when comparing to any previous tests. bright suprathreshold target in a region that previously
a. What degree of visual angle was tested? responded to fainter stimuli. A high false-negative
Most commonly set to ‘24-2’ (central 24° tested index may indicate hesitation or inattentiveness,
with a 54-point grid). A smaller field with higher though a true scotoma may also give false-negative
concentration of points gives further details of the results. However, in a true scotoma, the false-negative
foveal region. For example, ‘10-2’ assesses the error rate is low for the contralateral (normal) eye.4
central 10° with a 68-point grid. ‘30-2’ is similar to False-negative error may reduce with repeated testing
‘24-2’ but with an additional 6° and with a as the patient gets used to the testing procedure.

Wong SH, Plant GT. Pract Neurol 2015;15:374–381. doi:10.1136/practneurol-2015-001155 377

 HOW TO UNDERSTAND IT

Pract Neurol: first published as 10.1136/practneurol-2015-001155 on 3 July 2015. Downloaded from http://pn.bmj.com/ on September 21, 2022 by guest. Protected by copyright.
Figure 4 Interpreting the Humphrey visual field. The charts are viewed from the perspective of the patient looking into the test
bowl, as if patient is looking into the paper. Suggested checklist to systematically review Humphrey visual fields (see text for details):

1. Is this the correct test? 2. Can I rely on this test? 3. Is the test normal?
A. Patient name and identification number A. False-positive errors A. Visual sensitivity map
B. Date of test B. False-negative errors B. Total deviation map
C. Left or right eye? C. Fixation-loss index C. Pattern deviation map
D. Test performed D. Gaze-tracking graph
degree of visual angle tested
test protocol: threshold or screening
This patient’s test was unreliable: high fixation loss index (and comment from technician, ‘patient advised several times for both eyes’
(suggesting poor compliance), gaze-tracking graph also showed eye movements (indicated by upward spike from baseline) and high
false-negative errors, up to 20% in the left eye. The grey scale visual sensitivity map suggests a ‘clover leaf’ type pattern (figure 9).
This provided the impression that the patient had difficulty with the Humphrey test itself. Clinical examination including visual acuity,
colour vision, pupillary examination and visual field to confrontation to red pin was normal. The patient’s Goldmann visual field test
was normal (figure 3).

C. Fixation-loss index 3. Is the test normal?

Fixation loss is tested by presenting a stimulus at the Three maps are generated with numbers and pictorial
blind spot. If the patient sees this stimulus, it indicates representations:
loss of fixation. Values of >20% can compromise the A. Visual sensitivity map
test.4 However, this number could be artefactually ele- The numbers indicate the threshold of stimulus inten-
vated if the blind spot was inaccurately located, or in sity detected in decibel (dB), with zero corresponding
‘trigger-happy’ patients. Tracking of the gaze (below) to the brightest intensity. Typical normal values cen-
is better for assessing fixation loss. trally are around 30 dB. Values of 40 dB should not
D. Gaze-tracking graph appear in standard test conditions but could occur in
The eyes are tracked using video. The gaze tracking patients with high false-positive errors. The visual
graph shows an upward spike when the eyes move and sensitivity may improve with repeat testing as patients
a downward spike when the eyes blink. become more familiar with it. The grey scale map is a

378 Wong SH, Plant GT. Pract Neurol 2015;15:374–381. doi:10.1136/practneurol-2015-001155

 HOW TO UNDERSTAND IT

Pract Neurol: first published as 10.1136/practneurol-2015-001155 on 3 July 2015. Downloaded from http://pn.bmj.com/ on September 21, 2022 by guest. Protected by copyright.
Figure 5 Goldmann visual field from papilloedema. This patient has papilloedema from idiopathic intracranial hypertension.
Goldmann fields show (1) an enlarged blind spot and (2) inferonasal field restriction.

Figure 6 Goldmann visual fields of a patient with right optic neuropathy. All isopters are restricted but with preserved ‘contours’ of
the hill of vision, giving the appearance of a ‘sunken hill’. Compare this with figure 7 showing stacked isopters in a patient with
functional visual loss.

Figure 7 Goldmann visual fields of a patient with ‘stacked isopters’. This patient has functional overlay of a previous episode of
mild optic neuritis affecting the right eye. Compare this with figure 6 of another patient with optic neuropathy. These ‘stacked
isopters’ would represent a hill vision that is too steep to be physiological, that is, the close ‘contours’ here appear like a cliff drop.
Clinical examination with a red target confirmed the presence of a tubular field (figure 8), with the size of visual field remaining
unchanged when examined at 1 and 4 m. This is not keeping with the optics of light, whereby at a constant visual angle, the size of
the field would appear larger the further away, that is, when examined at 4 m (with a proportionately larger target for acuity), the
size of field to confrontation should be larger than on examination at 1 m.

Wong SH, Plant GT. Pract Neurol 2015;15:374–381. doi:10.1136/practneurol-2015-001155 379

 HOW TO UNDERSTAND IT

Pract Neurol: first published as 10.1136/practneurol-2015-001155 on 3 July 2015. Downloaded from http://pn.bmj.com/ on September 21, 2022 by guest. Protected by copyright.
Figure 8 Tunnel vision: functional (ie, tubular field) versus physiological. The optics of light is such that at a constant visual angle,
the size of the field appears larger when further away. When examined at 4 m (with a proportionately larger target for acuity),
the size of field to confrontation should be larger than on examination at 1 m. Thus, a ‘tubular’ field, where the size of field is
unchanged, suggests functional overlay.

Figure 9 Cloverleaf pattern on Humphrey visual fields. This artefactual visual field defect results from a reduced response rate as the
test progresses. The Swedish interactive threshold algorithm (SITA) threshold test starts by determining the initial brightness in the four
quadrants using the four points indicated by the arrows. Therefore, if a patient’s response deteriorates as the test progresses, for
example, because of reduced concentration, the visual field shows a cloverleaf pattern, where the thresholds are low at the four points
initially tested and higher for the surrounding points. This pattern commonly occurs in non-organic visual loss and is equivalent to
‘spiralling’ on the Goldmann. There will also be a high rate of negative errors. (A) Shows an example where the patient stops
responding very early in the test, giving an extreme example of the cloverleaf pattern; (B) shows another example of cloverleaf pattern.

visual representation of the numbers, with darker population. The numbers indicate the difference com-
areas indicating poorer sensitivity to stimuli. pared to the mean, that is, a negative value indicates
B. Total deviation map less visual sensitivity compared to the mean popula-
This shows the deviations of the patient’s visual sensi- tion. The probability plot gives a visual representation
tivity compared to an age-matched normal of statistical analysis (t test) of this deviation from the

380 Wong SH, Plant GT. Pract Neurol 2015;15:374–381. doi:10.1136/practneurol-2015-001155

 HOW TO UNDERSTAND IT
mean; the larger departure from the mean, the darker CONCLUSION

Pract Neurol: first published as 10.1136/practneurol-2015-001155 on 3 July 2015. Downloaded from http://pn.bmj.com/ on September 21, 2022 by guest. Protected by copyright.
the symbol. We present these simplified checklists to help neurolo-
C. Pattern deviation map gists to interpret Humphrey and Goldmann visual
This shows the deviation of the pattern from a normal fields. We emphasise the importance of correlating these
visual hill, where the peak is at the fovea. The visual field outputs with careful patient history and clin-
numeric values show any departure from the mean of ical examination. Increased exposure to perimetry and
an age-matched population, and as above, the prob- its application in the clinical setting will help build up
ability plot is a visual representation of statistical ana- skills in its interpretation. For readers interested in dee-
lysis indicating the extent of departure from mean. pening their understanding of fields and its nuances, we
The pattern deviation adjusts for any shifts in overall suggest further reading from the reference list.4–6
sensitivity: for example, a patient with cataract might
have a smaller or ‘sunken’ hill but with normal
contour patterns. Key points
By statistical chance, patients may have a few scat-
tered dark symbols on the probability map, which ▸ Perimetry results give a pictorial representation of the
may not be of concern. Instead, look for patterns, for patient’s ‘hill of vision’; keep the normal hill in mind
example, whether these are around the blind spot, when reviewing these tests.
which might indicate a true enlargement. It is import- ▸ Correlate perimetry results with the clinical history
ant to correlate the test results with the history and and examination (including examination to confronta-
clinical examination. tion), as the tests often have artefacts.
The visual sensitivity, total deviation and pattern ▸ Watch out for patient performance effect, for
deviation maps should be viewed together for any dis- example, high false-positive or false-negative errors,
crepancies. It is worth noting the following scenarios: cloverleaf pattern (static perimetry) or spiralling of
▸ Abnormal grey scale on stimulus intensity map but normal fields (kinetic fields).
probability plots: lid partially obscuring the superior field. ▸ Perimetry results change if anything obstructs the
▸ Abnormal total deviation but normal pattern deviation: travel of light towards the retina (eg cataract).
cataract, small pupils, incorrect correction for refractive ▸ Static and kinetic perimetry complement one
error. another; consider the other if the first is unexpectedly
▸ Abnormal pattern deviation but normal total deviation: normal or abnormal.
a test with high false-positive (‘trigger happy’) patient.
Additional information that may help, especially
when comparing with previous tests, include pupil Contributors SHW wrote the first draft of the manuscript; GTP
diameter (is there a wide variation between tests?), lens reviewed and made revisions to the manuscript.
modification (was the same correction used?), time Competing interests None declared.
taken to do the test (was this particularly long?). The Provenance and peer review Commissioned; externally peer
global indices show the mean deviations, which can reviewed. This paper was reviewed by Mark Lawden, Leicester,
help to monitor progression, especially in glaucoma. UK.
Three summary indices appear on the printout4: REFERENCES
1. The visual field index is a staging index designed to cor- 1 Cooper SA, Metcalfe RA. Assess and interpret the visual fields
respond to ganglion cell loss, that is, 100% represents at the bedside. Pract Neurol 2009;9:324–34.
normal fields and 0% represents blind fields. 2 Hickman SJ. Neurological visual field defects.
2. The mean deviation represents the degree of departure Neuro-ophthalmology 2011;35:242–50.
of the whole field’s average values, from age-adjusted 3 Bengtsson B, Olsson J, Heijl A, et al. A new generation of
normal values. algorithms for computerized threshold perimetry, SITA. Acta
3. The pattern SD represents irregularities within the field, for Ophthalmol Scand 1997;75:368–75.
4 Heijl A, Patella VM, Bengtsoon B. The field analyzer primer:
example, of localised field defects. This can be small in com-
effective perimetry. 2012, Carl Zeiss Meditec.
pletely normal patients or in those with complete blindness.
5 Barton JJS, Benatar M. Field of vision: a manual and atlas of
The visual field index and the mean deviation can perimetry. Current Clinical Neurology series. New Jersey:
help to identify progression; the visual field index may Humana Press, 2003.
be less prone to artefacts from cataract. These values 6 Carl Zeiss Meditec, Inc. Humphrey Field Analyzer Manual
may help to monitor progression, but with caution, Book II-i series system software version 5.1. 2012, Carl Zeiss
since artefacts and test reliability can affect them. Meditec.

Wong SH, Plant GT. Pract Neurol 2015;15:374–381. doi:10.1136/practneurol-2015-001155 381

 Miscellaneous

Correction: How to interpret visual

fields
Wong SH, Plant GT. How to interpret visual fields. Practical Neurology 2015;15:374-381.

Since the publication of the article, the authors noted an error in the figure legend of figure 6.
Where it is written ‘Goldmann visual fields of a patient with right optic neuropathy’ it should be
‘Goldmann visual fields of a patient with left optic neuropathy’.

©
©Author(s)
Author(s)(or(ortheir employer(s))
their 2021.
employer(s)) No commercial
2020. re-­use.
No commercial See rights
re-use. and permissions.
See rights Published
and permissions. by BMJ. by BMJ.
Published

PractNeurol
Pract Neurol2021;21:e1.
2020;0:1. doi:10.1136/practneurol-2015-001155corr1
doi:10.1136/practneurol-2015-001155corr1

Pract Neurol August 2021 Vol 21 No 4 e1

View publication stats