Comparison of Exophthalmos Measurements: Hertel Exophthalmometer

versus Orbital Parameters in 2-Dimensional Computed Tomography

Kyung Jun Choi, MD, Min Joung Lee, MD
From the Department of Ophthalmology, Hallym University Sacred Heart Hospital, Hallym University
College of Medicine, Anyang, South Korea.

Presented by :
Muhammad Alfin Kamal

Supervised by :
Dr. dr. Trilaksana Nugroho, M.Kes, FISCM, Sp.M (K)
 Introduction

affected by the However, selection of a proper base may require skill and
tightness of the clinical experience and can act a factor of disagreement
septum and the between clinicians
vascular flow of the
orbit

improved with a standard protocol such as keeping the same

base between measurements and using the same
volume of orbital exophthalmometer
contents and the
morphologic of the
bony orbit
low accuracy and poor reproducibility

the most widely used instrument is the Hertel

exophthalmometer
Proptosis
Proptosis is a particularly critical sign for grading the severity
of disease and for assessing the course of thyroid-associated
orbitopathy (TAO).
 What the alternative?
CT-3D
Accurately reflect complex orbital morphology
and enables volumetric analysis
limitation:
• need for specific software,
• the intricate nature of the procedure
• the length of time it takes to perform
VS CT-2D
• Relatively rapid
• Simple
• Easily assessable procedure
• Has been used to measure orbital volume and to
characterize orbital morphology
But there has been no unified standard or easily
available tool to measure proptosis using 2D-CT
 Campi et al.
Recently introduced a novel method to calculate the degree of
proptosis with 2D-CT images

The reliability and clinical usefulness of this method for

measuring proptosis has not been proven yet

They modified the measurement protocol using the ImageJ program and simplified the calculation

Purpose To compare the reliability of Hertel exophthalmometry and CT parameters.

Purpose
Two observers carried out both examinations in a blinded manner,and intraclass correlation
coefficients (ICCs) were analyzed

In addition, we analyzed the correlation between Hertel exophthal- mometry and CT parameters as
measured by a single oculoplasty specialist and the association between CT parameters and the
activity and severity of TAO
 methods

Type of studies • Retrospective, observational case series

• Was performed at the Hallym University Sacred Heart

Place and Periode Hospital
• From August 1, 2011, to December 31,2013
The treatment modalities used for thyroid disease
smoking history,
Diabetes
hypertension, and presence of other autoimmune disease
Patients

Inclusion : Case control Inclusion :

patients who had clinically 69 eyes of 41 31 eyes of 31 patients who underwent orbital
confirmed TAO and available CT for unilateral benign orbital
CT scans disease other than TAO and had
28 patients unaffected normal orbits
with bilateral TAO (control group).

13 with unilateral TAO

All patients in the TAO group and control group received comprehensive ophthalmic and orbital examinations at the initial visit.

Exclusion :
• Patients with other conditions that could change the morphology of the orbit, such as blow-out fracture or a history of any
orbital surgery, were excluded
• Patients with a history of ocular surgery other than cataract surgery were also excluded
 Clinical Examination patient visited the oculoplasty clinic
Ophthalmic examination data included :
• Best-corrected visual acuity
• Intra ocular pressure orbital examinations were usually performed twice
• Spherical equivalent
• Slitlamp biomicroscopy
• Fundus examination once by a designated resident (K.J.C., examiner 2)
Orbital examination including :
• Margin reflex distance 1,
• Measurement of proptosis by Hertel
exophthalmometry
second by a faculty member of the oculoplasty division (M.J.
• Extraocular muscle movement (EOM) L., examiner 1)

Patients with TAO

The degree of exophthalmos measured by the 2
ophthalmologists using the Hertel exophthalmometer was
CAS NOSPECS recorded.

scores ≥3 Assessed • The 2 observers used same exophthalmometer.

severity • All examinations were performed separately
active TAO • the examiners did not refer to each other’s testresults
• The base of exophthalmometry did not always coincide
between the 2 observers.
 Measurement of CT Parameters

We used the concept of Campi et al. to assess orbital parameters and modified the protocol using
ImageJ1.47 as follows
 Data Analysis

The absolute or relative interobserver difference in measurements of the same orbits was plotted against the mean
value using the Bland–Altman approach

Receiver operating characteristic (ROC) curves were obtained to compare the ability of the CT parameters and
Hertel exophthalmometry to diagnose TAO.

In the TAO group, the CT parameters and Hertel exophthalmometry were also com-pared for their ability to assess
the activity and severity of TAO.
 RESULTS were significantly higher in the
TAO group
Patients Characteristics

Were significantly different between the 2 groups

 Interobserver Consistency Rates of Proptosis Measurement
Methods

We investigated the rates of agreement for the exophthal- mos measurements targeting patients with TAO who under- went
orbital examinations by both observers (52 eyes of 30 patients).

The ICC was considered high for GA (0.984) and the GA/
OA ratio (0.993

The ICC was considered moderate for OA (0.531) and Hertel

exophthalmometry (0.569).
 Correlation of Hertel Exophthalmometer and CT Parameters

We analyzed the correlation between Hertel exophthalmometry and GA, OA, or GA/OA ratio in all patients (92
eyes of 74 patients).

GA was inversely correlated with the However, there was a negligible With regard to the GA/OA ratio, a
Hertel value (r = −0.740, p < 0.001; Fig. correlation between OA and the Hertel significant negative correlation with
3A). value(r = −0.08, p ¼ 0.401; Fig. 3B). the Hertel value was also observed (r =
−0.706, p o 0.001; Fig. 3C).
We then compared the ability of Hertel
exophthalmometry and the CT parameters to
differentiate between the TAO group and the
control group using ROC curves.

The Hertel exophthalmometer showed the largest

area under the ROC curve (AUROC) (0.813)
followed by GA (0.781), theGA/OA ratio (0.761),
and OA (0.552).Comparison of the AUROC
showed that the Hertel and OA,GA and OA, and
GA/OA and OA were statistically significant (all p
< 0.01),but other pairwise comparisons were not
significant (Fig. 4).
 Discussion
GA and the GA/OA ratio measured using 2D-CT images showed better interobserver
agreement than Hertel exophthalmometry.

examiner-related errors instrument-related errors patient-related errors

 In contrast,measurement of CT parameters may be less
affected by the examiner’s experience or by periorbital changes.

In this study, GA and the GA/OA ratio showed

very high interobserver agreement, over 0.9 (0.984 and 0.993, respectively).

OA showed moderate interobserver agreement (0.531)

When drawing the triangle,it is technically more difficult to setthe apical point than the anterior 2 points
 In our study, GA and the GA/OA ratio showed a strong negative correlation with
In our study, GA and the GA/OA ratio showed a strong negative correlation with
Hertel exophthalmometry
Hertel exophthalmometry
1. Measurement of proptosis using 2D-CT can be distorted by the patient’s head position.
2. Flexion and extension of the neck are associated with the cutting angle of the scan
3. Lateral bending or rotation of the neck can affect the symmetry between the right and left sides

Commented that the percentage of the OA occupied by the

Campi et.al
eyeball can reduce the measurement error.

However, it is more technically difficult to calculate the GA/OA

ratio because 3 points need to be drawn, and the apical point is
difficult to determine;

we believe that GA would be a better parameter to use in the

clinic due to its simplicity.

The measurement of GA will be difficult and in accurate in cases with eye ball elongation or posterior staphyloma, and
this may be one of limitations of our measurements.
 We analyzed the diagnostic power of Hertel
exophthalmometry and CT parameters to discriminate
between the TAO group and the control group using
ROC curves.

The Hertel exophthalmometer ,the GA/OA ratio, and our results cannot
GA showed fair diagnostic ability with an AUROC of our results cannot
be generalized
over 0.7. In our subject cohort, the cut of fvalue using be generalized
because of the
the Hertel exophthalmometer was 17mm with 64.4% because of the
small number of
sensitivity and 87.5% specificity. small number of
cases
cases
In addition,the proportion of female patients was lowerin
the control group than in theTAO group, although the
difference was not statistically significant.

Further prospective, large-scale study is warranted to

confirm the cut-off value.
 Conclusion

The results show that GA and the GA/OA ratio are reliable CT
The results showwith
parameters that aGA andICC
high the GA/OA
when ratio arewith
compared reliable CT
parameters withusing
Hertelexophthalmometry.By a highImageJ,
ICC when
GA compared with ratio can
and the GA/OA
Hertelexophthalmometry.By
be measured more asily, and using ImageJ,
these valuesGAcan
andprovide
the GA/OA ratio can
a useful
be measured morefor
reference asily, and these
evaluation of values can provide a useful
exophthalmos.
reference for evaluation of exophthalmos.
Critical Appraisal
 Comparison of Exophthalmos Measurements: Hertel Exophthalmometer versus
Orbital Parameters in 2-Dimensional Computed Tomography
Section 1: Internal validity
In this study this criterion is:
1.1 The study addresses an appropriate and clearly focused question.

Write the reason of your answer: Addressed

To compare the reliability of orbital parameters calculated using 2-dimensional computed
tomography (CT) and Hertel exophthalmometry when measuring exophthalmos in normal
subjects and in patients with thyroid-associated orbitopathy (TAO).
1.2 The assignment of subjects to treatment groups is randomised

Write the reason of your answer:

This journal mention that the study included 69 eyes of 41 patients who had clinically
confirmed TAO and available CT scans (TAO group: 28 patients with bilateral TAO,13 with Addressed
unilateral TAO), and 31 eyes of 31 patients who underwent orbital CT for unilateral benign
orbital disease other than TAO and had unaffected normal orbits(control group).
 it’s mean the group assignment not randomised , and with the consecutive sampling
1.3 An adequate method is used
Write the reason of your answer:
Retrospective, observational case control.

Retrospective : all case have already happened before the study

begins
observational case series :
Addressed
This journal mention that the study included 69 eyes of 41 patients
who had clinically confirmed TAO and available CT scans (TAO group:
28 patients with bilateral TAO,13 with unilateral TAO), and 31 eyes of
31 patients who underwent orbital CT for unilateral benign orbital
disease other than TAO and had unaffected normal orbits (control
group).
1.4 Subjects and investigators are kept ‘blind’ about treatment/diagnostic
allocation
Write the reason of your answer:
Addressed
Investigators are kept blind :
Two observers carried out both examinations in a blinded manner, and
intraclass correla- tion coefficients (ICCs) were analyzed.
1.5 The treatment and control groups are similar at the start of the trial
Write the reason of your answer:
The demographic and ophthalmic characteristics of the control group and the
TAO group are shown. Age, male-to-female ratio,visual acuity,and refraction
expressed by spherical equivalent were not significantly different between the 2 Addressed
groups.
However,intraocular pressure and exophthalmos measured by Hertel
exophthalmometry were significantly higher in the TAO group (p < 0.007 and p <
0.001, respectively). Among the CT parameters ,GA and the GA/OA ratio were
significantly different between the 2 groups(all p < 0.001).
1.6 All relevant outcomes are measured in a standard,
valid and reliable way
 
Write the reason of your answer:
The ICCs between the 2 observers (examiner 1 and
examiner 2) were calculated to assess the interobserver Addressed
consistency of Hertel exophthalmometry and the CT
parameters (GA, OA, and GA/OA ratio).
The absolute or relative interobserver difference in meas-
urements of the same orbits was plotted against the mean
value using the Bland–Altman approach
1.8 What percentage of the individuals or clusters recruited into each treatment arm of the study
dropped out before the study was completed?
No dropped out

1.9 All the subjects are analysed in the groups to which they were randomly allocated
 
Write the reason of your answer:
Yes

1.10 Where the study is carried out at more than one site, results are comparable for all sites
 
Write the reason of your answer:
No
Section 2: Overall assessment of the study

2.1 How well was the study done to minimise bias?  

With Exclusion criteria

2.2 If coded as +, or - what is the likely direction in  Measurement bias

which bias might affect the study results?

2.3 Are the results of this study directly applicable to  our results cannot be generalized because of the
the patient group targeted by this guideline? small number of cases
Section 3: Description of the study (the following information is required to complete evidence tables facilitating cross-study comparisons
3.1 Do we know who the study was funded by? [ ] Academic Institution
[ ] Healthcare Industry
The author didn’t declare the funding of this study [ ] Government [ ] NGO  [ ] Public funds 
[ ] Other
3.2 How many centres are patients recruited from?  1 centres
3.3 What is the social setting (ie type of environment in which they [ ]   Urban  [ ] Rural  [ ] Mixed
live) of patients in the study?

There is no information
3.4 What criteria are used to decide who should be INCLUDED in  Patients who had clinically confirmed TAO and available
the study? CT scans
   Patients who underwent orbital CT for unilateral benign
orbital disease other than TAO and had unaffected
normal orbits
3.5 What criteria are used to decide who should be EXCLUDED from • Patients with other conditions that could change the
the study? morphology of the orbit, such as blow-out fracture or a
  history of any orbital surgery.
• Patients with a history of ocular surgery other than
cataract surgery were also excluded
3.6 What intervention is investigated in the
study?
 orbital parameters in 2-dimensional computed
 
tomography

3.7 What comparisons are made in the study?

(ie what alternative treatments are used to
compare the intervention with?). Include
dosage where appropriate. Hertel exophthalmometer
 

3.8 What methods were used to randomise

patients, blind patients or investigators, and
to conceal the randomisation process from
investigators? Not Reported
 
Thank You
Thank You