Embedded Scribd iPaper - Requires Javascript and Flash Player
Research Articles
Refinement and Revalidation of the Equine Acute Abdominal
Pain Scale (EAAPS)
Sutton, G.A.* and Bar, L.
Koret School of Veterinary Medicine - Veterinary Teaching Hospital, The Hebrew University of Jerusalem, Robert H Smith
Faculty of Agriculture, Food and Environmental Sciences, POB 12, Rehovot, Israel 76100.
* Corresponding author: Dr. Gila Sutton, DVM, M.Sc, Ph.D, Koret School of Veterinary Medicine - Veterinary Teaching Hospital, The Hebrew University
of Jerusalem, Robert H Smith Faculty of Agriculture, Food and Environmental Sciences, POB 12, Rehovot, Israel 76100. Tel.: 972 3 968 8507,
Fax: 972 8 946 7940, Email address: gila.sutton@mail.huji.ac.il
AB S T RAC T
Assessment of pain is vital for colic treatment. The purpose of this study was to revalidate a refined version of
the behaviour-based Equine Acute Abdominal Pain Scale (EAAPS). Based on an earlier study, behaviours
in the scale were removed or replaced. Ten behaviours remained. For revalidation, forty films of horses with
colic were presented by computer-generated random order to two randomly-assigned groups of equine
veterinarians. One group (n=8) scored the severity of pain demonstrated in the films by utilizing a numerical
rating scale (NRS) and one group (n=7) with the refined version of the EAAPS. Intra-rater reliabilities of
the EAAPS and of the NRS were comparable based on Limits of Agreement. The inter-rater reliability of
the EAAPS was significantly improved compared to the NRS (NRS; Intraclass correlation (ICC) = 0.6
(95% Confidence Interval (CI); 0.5-0.8) and EAAPS; ICC = 0.88 (95%CI: 0.8-0.9)). Face validity was
71% (95% CI; 29-96) in support of the EAAPS. The two scales showed substantial convergent validity
(weighted kappa of 0.73 (95%CI; 0.58-0.88). The predictive validity of the EAAPS scale was similar to the
NRS (AUC of EAAPS; 0.75 versus NRS; 0.78 for mortality; AUC of EAAPS 0.76 versus NRS of 0.83 for
treatment modality) and the ability to discriminate between extreme groups of either control horses versus
cases or by extreme groups defined by NRS scores of 0-2 versus 3-5 was excellent (AUC 0.99 and 0.955,
respectively). In summary, revalidation of the refined EAAPS was necessary and was found to be highly
reliable and comparatively valid.
Keywords: Equine; Colic; Pain Assessment; Validation; Facial Validity.
INTRODUCTION
Colic, a major cause of death in horses (1), is a condition
characterised by pain but until recently, there has been no
standardised pain severity scale for use in these cases. All
of the earlier scales were designed for use in experimental situations, from the first scale published by Muir and
Robertson in 1985 (2) through a final modification of it by
Boatwright et al., 1996 (3). These early scales were never
validated. Beginning this century, however, there have been
a number of pain scales developed for post-operative pain
Israel Journal of Veterinary Medicine  Vol. 71 (1)  March 2016
March 2016.indb 15
(4, 5, 6) as well as a scale based on equine facial expression,
the “Horse Grimace Scale” (7). Some have undergone initial
validation (6, 7), however, all of these more recent scales have
been developed for post-operative pain and most involve
complex composite scales rather than simple clinical indices
(4-6, 8). In contrast, the Equine Acute Abdominal Pain
Scales (EAAPS) was designed specifically for evaluation of
the severity of acute colic pain for use in adult horses under
clinical conditions unrelated to surgery (see Appendix 1).
As a simple clinical index rather than a composite scale, it
Evaluation of Pain in Horses
15
17/03/2016 11:04:12
Research Articles
is based entirely on behaviours, and includes only a small
number of items, requiring no arithmetic calculations, in
order for it to be easy to use at horse farms as well as at a
referral clinic (9, 10).
Originally the EAAPS was developed as two contending behaviour-based scales; EAAPS-1 and EAAPS-2. They
each included 12 identical behaviours with scores from
1 to 5 assigned to each behaviour. The scales differed in
that EAAPS-1 had one score assigned to each behaviour
and the EAAPS-2 had one or two scores per behaviour
depending on the intensity of the behaviour demonstrated
(9).
The EAAPS scales were subsequently validated (10),
based on principles of clinimetrics, by demonstrating
reliability and validity (11, 12). Reliability is the extent
to which a scale yields the same results on repeated trials
and validity conveys whether the scale measures what it
aims to measure. Ideally, validity involves comparison to
a gold standard but when there is no gold standard, as for
the measurement of pain, alternate types of validity are
required, such as face validity and various types of construct
validity including convergent, discriminative and predictive
validities (13).
Following the validation study comparing and contrasting the two EAAPS scales to each other and to a global pain
scale, the EAAPS-1 scale was chosen due to its superior
reliability (10). Then a revised version of the EAAPS was
constructed by making changes based on the two studies. The
number of behaviours was reduced to ten. Three behaviours;
depression, weight shifting and collapse were removed due
to their relatively poor performance regarding agreement
of observers as to the presence of these signs in film clips
(9, 10). Lip curling (Flehman) was introduced, as it had
shown good agreement between observers (9) and the scores
were adjusted in order to reduce the number of behaviours
assigned a score of 3, since it appeared to be an overly common score for the EAAPS-1 in comparison with the other
scales (9, 10).
The overall purpose of this study was to validate the
revised version of the EAAPS scale for its clinimetric properties. When scales measure hypothetical constructs, such as
pain, validation is an ongoing task (13). The specific objectives
were to assess the inter- and intra-rater reliabilities and three
general types of validities: face validity, construct validity and
predictive validity of the scale.
16
Sutton, G.A.
March 2016.indb 16
MATERIALS AND METHODS
Study design
An evaluation group of equine practitioners viewed films of
horses exhibiting signs of colic (or of control horses) via a
specially designed website (see below). The films were shown
in random order by the website program. Each participant
viewed a total of 41 films and scored the films using one of
two scales; the revised EAAPS scale or a global numerical
rating scale (NRS) as a control group.
Films
Films clips were chosen purposively from films previously prepared (9), from 28 cases of colic, over 1 year of age, (cases) presented to the Koret School of Veterinary Medicine, Veterinary
Teaching Hospital, Equine Department. Written approval
was obtained from the Internal Ethics Review Committee of
the Veterinary Teaching Hospital prior to filming. An attempt
was made to select films representing the entire breadth of
pain spectrum and those in which behaviours were clearly
demonstrated. Six cases provided two films each giving a total
of 34 digital film clips of horses with colic. Six films of control
horses, not suffering from colic, who were hospitalised for
fertility treatment, were added (filmed by LB) to raise the
total number of films from 34 to 40. The median length of the
film clips was 27 seconds (interquartile range; 19-46 seconds).
One of the 34 films of colic cases was randomly selected to be
shown twice for assessing intra-rater reliability.
Participants
Eighteen equine practitioners were recruited by general electronic mail sent to 59 local equine practitioners in Israel. Ten
letters were returned by the postmaster. Of those not returned,
18 (36%) responded and agreed to participate. All participants
were asked questions regarding themselves, their veterinary
education, specialization training and experience treating
horses. The assessors were assigned to one of two independent groups (n=9); a control group and a test group, by block
randomization using the EXCEL program. The participants
were not trained to use the EAAPS since the descriptors are
commonly used in the scientific literature, the participants
were equine practitioners and the descriptors were defined
(Appendix 1b). The untrained participants of the test group
scored the severity of pain the horses were showing in the
films using the revised EAAPS scale and those of the control
Israel Journal of Veterinary Medicine  Vol. 71 (1)  March 2016
17/03/2016 11:04:12
Research Articles
group provided a global assessment of pain by using a 6-point
numerical rating scale (NRS) in which ‘0’ indicated ‘no pain’, ‘1’
- ‘mild pain’, ‘2’ - ‘mild to moderate pain’, ‘3’ - ‘moderate pain’,
‘4’ - ‘moderate to severe pain’, and ‘5’ indicating ‘severe pain’.
Face Validity
Face validity of the EAAPS scale was evaluated by asking the
participants of the test groups, upon completion of the scoring, to what degree they agreed with the following statement:
‘The EAAPS scale is a valid scale for the assessment of acute
abdominal pain in the adult horse’, using four response options (strongly agree, agree, disagree, strongly disagree) and by
asking whether the overall impression was that the EAAPS
scale made clinical sense with a yes/no response option.
Behaviour descriptor evaluation
Two new behaviour descriptors; “lip-curling”(14) (“Flehman”)
(15) and “crouching”(16), had not been previously evaluated
for inter-rater agreement. Therefore, participants in the test
group were asked to indicate whether these behaviours were
demonstrated in each film. The response option was dichotomous (Yes/No).
Website Program
For the entire process, the assessors viewed, managed and
scored the films using a structured, web-based course management system (Moodle) modified for this study (http://
muddle.cs.huji.ac.il/mu11). Access to the website required
a username and password assigned individually to each
participant.
Sample size calculation
Sample size was calculated based on intra-class correlation
coefficients (ICC) for the reliability analysis. At an alpha
of 0.05 and power of 80%, 40 films would be needed with
at least 5 observations per film in order to differentiate an
expected ICC of 0.8 from an ICC of 0.67, as obtained, respectively, for the EAAPS-1 scale and the NRS scale in the
earlier study (10). The calculation was made in WinPEPI as
well (WinPEPI (version 2.91), PairsEtc., Sample Size S6,
copyright J.H. Abramson).
Statistical analysis
Point estimates and Fisher’s 95% confidence intervals (95%
CI) were calculated (WinPEPI ( www.brixtonhealth.com/
Israel Journal of Veterinary Medicine  Vol. 71 (1)  March 2016
March 2016.indb 17
pepi5windows.html) ). Spearman’s ranked correlation (rho)
was used to evaluate association between age of horse and
median scores of films. Non-parametric independent t-tests
and Kruskall-Wallis were used to evaluate associations
between sex/breed and median scores of films. Inter-rater
reliability was tested by intraclass correlation (ICC), utilising
the two-way random effects model, for absolute agreement
and single measures (McGraw and Wong A1 Model) (17,
18) (SPSS 18, IBM, USA). Interpretation of the ICC was
as follows; values above 0.75 were regarded as excellent reliability, 0.4 as good reliability (19). Intra-rater reliability was
evaluated by Limits of Agreement (LOA) (20) between two
scores given to identical films. Individual behaviour descriptors were evaluated for bias and for inter-rater agreement by
multirater, multicategory kappa coefficient (Software by Mr.
William Sears, Ontario Veterinary College, Guelph, Ontario,
Canada). Interpretation of the kappa coefficient was based
on Landis and Koch, 1977 (21).
Face validity was evaluated as frequency of endorsement
with 95% confidence intervals for each response option
(either; strongly agree, agree, disagree, strongly disagree, or;
yes, no) (WinPEPI). The responses of the two questions
were compared for reliability. Several constructs were used
to evaluate Construct validity; convergent, discriminant and
predictive. Convergent validity of the EAAPS scale in comparison to the global assessment of pain (NRS) was assessed
by agreement between rounded median scores of the two
scales as expressed by weighted kappas (http://www.vassarstats.net/kappa.html). Discriminant validity was expressed
as the ability of the median EAAPS scores to discriminate
between two types of extreme groups. First, colic versus control horses by Fisher’s Exact Test (SPSS 18) and second, by
ROC curve to predict severe versus mild pain as defined by
the NRS scores. Severity of pain demonstrated in each film
was assessed by median NRS scores of 0, 1 or 2 indicating
mild pain, compared to scores of 3, 4 or 5 indicating severe
pain. The cut-off value of 3 or more for the NRS scale as
indicating severe pain was chosen based on the ROC curve
of the mean NRS scores to predict death (area under the cure
(AUC) = 0.784; 95%CI 0.633-0.934) or treatment outcome
of surgery or euthanasia versus no treatment or medical treatment (AUC=0.833; 95% CI 0.705-0.962). Predictive validity
of each scale was evaluated by comparing the frequency of
median EAAPS scores to mortality outcome (alive or dead)
and to treatment modality (controls, medically treated, surEvaluation of Pain in Horses
17
17/03/2016 11:04:12
Research Articles
gically treated or euthanatized) by Fisher’s Exact and chisquare for trend tests (SPSS 18). The predictive validity of the
scales was also assessed by ROC curve for treatment modality
(no treatment or medical treatment versus surgical treatment
or euthanasia) (SPSS 18). The frequency distribution of the
rounded median scores (0 to 5) over all of the films for each
scale (EAAPS and NRS) were compared by Fisher’s exact
test (SPSS 18).
Figure 1: Frequency distribution of the severity of pain over the films
(n=40) based on the median of the numerical rating scale (NRS) scores
of each film.
RESULTS
Films
Characteristics of the horse population of the films can be
found in Table 1. No statistically significant association was
found between the median EAAPS or NRS scores and sex,
breed or age of the horses, nor between scores and participants. The correlation between the NRS scores and age was
statistically significant (P < 0.05) but small to moderate in
strength (Rho = -0.364). The frequency distribution of the
median NRS scores for the 40 films can be seen in Figure 1.
Reliability
The response rate was 83% (15/18) (7 in the test group, 8 in
the control group). The participants were general practitioners except for two; a board-certified theriogenologist and
The EAAPS scale demonstrated superior inter-rater reliability in comparison to the NRS scale (P < 0.05). The ICC of
the EAAPS scale was 0.88 (95%CI: 0.83, 0.93) compared to
the NRS; 0.60 (95%CI: 0.51, 0.79). The intra-rater reliability
of the EAAPS scale was similar to the NRS based on limits
of agreement (LOA) (Figure 2a & 2b). Both the EAAPS
scale and the NRS scale demonstrated no difference in the
scores of identical films in 5 of 7 (71%) and in 6 of 8 (75%)
cases, respectively.
Table 1: Characteristics of the horse population of the films (n=40);
median (IQR=interquartile range) or frequency (n) and percentage.
Table 2: Demographic characteristics of participants (n=15);
the number in each category (percentage)
Participants
Characteristic
Age (years)
Heart rate (beats
per minute)
Sex
Breed
Treatment
Mortality
18
an anaesthetist. Additional characteristics of the participants
appear in Table 2.
Sutton, G.A.
March 2016.indb 18
Mares
Stallions
Geldings
Arabians
Quarterhorses
Grade
Others
Medical
Surgical
Euthanatised
Controls
Died
Median
6
IQR
(3-11)
44
(40-60)
Frequency (n)
27
4
9
17
10
5
8
16
16
2
6
7
Percentage
68%
10%
22%
42%
25%
13%
20%
40%
40%
5%
15%
18%
Characteristic
Sex
Male
Female
Clinical training level
Internship
Residency
Neither
Percentage of practice equine
>=90%
<90%
Number of years in practice
Median 10 years (IQR*=12)
Range 2-31 years
<5 years
Participate in research
Number (%)
9 (60)
6 (40)
8 (53)
1 (7)
6 (40)
12 (80)
3 (20)
5 (33)
6 (40)
* IQR = interquartile range
Israel Journal of Veterinary Medicine  Vol. 71 (1)  March 2016
17/03/2016 11:04:12
Research Articles
(a) EAAPS scale
(b) NRS scale
Figure 2: Intra-rater reliability for each of the scales as shown by 95% Limits of Agreement (LOA); (a) EAAPS scale, and (b) NRS scale.
(a) The point marked by the letter ‘a’ denotes 5 points in the identical location (total n=7 points) and each asterisk in
(b) denotes 3 points in the identical location (total n=8 points)
Previously Unevaluated Behaviours
Agreement between observers regarding “lip curling” demonstrated substantial agreement with a kappa coefficient of
0.77 (95%CI; 0.60-0.95) with insignificant bias. “Crouching”
demonstrated fair agreement with a kappa of 0.25 (95%CI;
0.04-0.47) with significant bias (21).
Validity
Face validity was demonstrated by 5 of 7 (71%; 95%CI:
29-96) equine practitioners agreeing or strongly agreeing
that the EAAPS scale was a valid scale for the assessment of
acute abdominal pain in adult horses. The same practitioners
replied in the affirmative that the EAAPS scale is valid in
response to the second question. Convergent validity of the
EAAPS scale was substantial when compared to the NRS
(weighted kappa 0.73; 95%CI 0.58-0.88). As evidence of
discriminant validity, the EAAPS scale highly discriminated
between extreme groups, as defined either by the severity
of the NRS scores or by comparing the scores of cases to
controls. When compared to severity of pain demonstrated
in the film, based on the NRS score, the AUC for the
EAAPS scale to predict severe pain was 0.955 (95%CI:
0.87-1.0). When comparing cases to controls, 5/6 films of
control horses received a median EAAPS score of 0 and 1/6
received a median score of 1, while all cases of colic received
median scores greater than zero (P < 0.00001) (Figure 3).
Predictive validity was demonstrated as a significant associaIsrael Journal of Veterinary Medicine  Vol. 71 (1)  March 2016
March 2016.indb 19
tion between the frequency of pain scores in each treatment
modality group (none, medical, surgical or euthanasia (P
= 0.001 Fisher’s Exact; P = 0.00001 chi-square for trend)
but not between the frequency of pain scores and death (P
= 0.214 Fisher’s Exact) (Figures 4a & 4b). The AUC of
the ROC curve for the outcomes of mortality or treatment
modality for the EAAPS scales were comparable to the NRS
scale (Table 3).
Figure 3: Discriminant reliability (extreme groups): Comparing
median EAAPS scores given to cases of colic versus control horses
without colic (P < 0.00001).
Evaluation of Pain in Horses
19
17/03/2016 11:04:13
Research Articles
Scale comparison
The Equine Acute Abdominal Pain Scale (EAAPS) is the
only pain scale developed and validated for acute abdominal
pain prior to colic surgery. Revalidation is important since
pain is not biologically measureable and validation depends
on contructs or mini-theories that may not be true under
differing circumstances and when changes are made (13).
This validation was also carried out on a different population of observers than the previous validation as the equine
practitioners who participated in this study were from Israel
and in the previous study they were from Europe and the
United States of America (10).
The better of two EAAPS, which had been vigorously
constructed and validated, and had shown excellent reliability and adequate validity, had, nonetheless, had weaknesses identified in the previous validation study (9, 10).
Alterations were made in order to remedy the weaknesses,
and the revised version underwent revalidation in this study.
(a) Treatment modality
(b) Mortality
The frequency distribution of the rounded median scores
over all of the films varied significantly with the scale used
to assess them (Fisher’s Exact P < 0.001). The agreement
was stronger at the extremes of no pain (0) and severe pain
(5) and weaker over the mid-range where the EAAPS scale
scored more films as 1 and 2 and less as 3 and 4 compared to
the NRS scale (Figure 5).
DISCUSSION
Figure 4: Predictive validity: Association of median EAAPS scores with; (a) Treatment modality, and (b) Mortality
Table 3: Predictive validity based on ROC curves (AUC (Area
under the curve), 95% CI (confidence interval), cut-off values with
corresponding sensitivity and specificity)
Mortality
EAAPS
NRS
Treatment
EAAPS
NRS
20
AUC
(95%CI)
0.75
(0.54-0.96)
0.78
(0.63-0.93)
0.76
(0.62-0.91)
0.83
(0.7-0.96)
Sutton, G.A.
March 2016.indb 20
Cut-off
Value
Sensitivity
Specificity
3.93
71%
85%
3.36
86%
68%
2.92
78%
65%
2.64
94%
70%
Figure 5: Comparison of the frequency (%) of films assigned each score
(0-5) (median over all the observers) by each of the scales.
Israel Journal of Veterinary Medicine  Vol. 71 (1)  March 2016
17/03/2016 11:04:14
Research Articles
The revised version of the scale is an improvement because
there are fewer behaviours (10 versus 12) which is preferable
for clinimetric scales (22, 23), it includes a new behaviour that
demonstrated substantial agreement (21) between observers
(lip curling), and excluded behaviours that had demonstrated
poor agreement between observers (depression, weight shifting and collapse). In this study, “crouching”, a term that was
introduced in order to provide an alternative descriptor
for “attempting to lie down”, surprisingly showed only fair
agreement between observers (21). This finding justifies the
removal of this term from later versions of the EAAPS since
the agreement was not strong, although perhaps the observers
in this study, not being native English speakers, were not
clear about the meaning of the term. This theory is supported
by the significant bias which may indicate that only certain
people understood the word. Lip curling, which was added
in a similar manner in order to explain Flehman , which itself
showed excellent agreement in the earlier study (9), may be
a more commonly used term (14), and therefore, may have
been more universally recognised among the observers which
may explain the substantial agreement and lack of bias it
obtained in this study.
As in the previous validation, the inter-rater reliability
of the new version was very high. It is considered excellent
(19) and compares favourably to pain scales developed in
young children aged 5 or younger (24) and in critically ill
and cognitively impaired children (25) as well as in comparison to the earlier version of the EAAPS (EAAPS-1)
(10).
In this study, the intra-rater reliability of the EAAPS
scale was not better than the numerical rating scale (NRS),
with a difference in one observer and in one score only. In
future studies, more than one repetition of the film will be
used for evaluating intra-rater reliability, particularly if the
number of observers is relatively low, as in this study.
The revised version of the EAAPS scale showed adequate validity in comparison to the NRS as did the earlier
EAAPS-1 scale (10). The face validity, demonstrating overall
endorsement of the scale, had excellent intra-rater reliability
in the responses since the two questions pertaining to face
validity produced identical results. This, however, may have
been influenced by the fact that the questions were placed in
adjacent positions in the questionnaire. Regarding construct
validity, all of the various types (convergent, predictive and
extreme groups) demonstrated higher point estimates than
Israel Journal of Veterinary Medicine  Vol. 71 (1)  March 2016
March 2016.indb 21
the earlier version, however since these were different studies,
they could not be statistically compared (10).
Based on the distribution of the films compared to the
NRS scale (Figure 5), the scale is an improvement since the
earlier version had a preponderance of films with the score
of 3 (10) and in this study, the distribution of the film scores
more closely paralleled those of the NRS scale, which is currently the accepted method that pain severity is assessed in
cases of colic.
Limitations of this study include technical aspects of
films that make it difficult at times to observe the behaviours
needed to score the level of pain the horse is demonstrating.
The film clips were short in order to enhance compliance of
the participants, however, perhaps too short to assess severity
of pain. On the other hand, pain is dynamic and perhaps
longer films would encompass different levels of pain in the
same film.
Future studies should evaluate the usability and feasibility
(26) of the scale in a prospective, real-time study.
CONCLUSIONS
In this study, a revised version of the EAAPS scale was
validated and shown to have excellent inter-rater reliability
and comparable validity but appeared improved since the
distribution of the scores appears closer to the distribution of
the global rating scale (NRS) than the earlier version. Future
studies are needed to demonstrate usefulness in the field in
a prospective study.
CONFLICT OF I NT EREST STAT EMENT
Neither of the authors of this paper has a financial or personal
relationship with other people or organizations that could inappropriately influence or bias the content of the paper.
ACKNOWLED GEMENTS
We would like to thank Ms. Tali Bdolach-Avram, Mr. William
Sears and Dr. Hillary Voet for statistical consultation, Chana
Slutzkin of the System Group of the School of Computer Science
and Engineering of the Hebrew University, for website management and Dvorah Weisman of the Computer Unit of the Hebrew
University, The Robert H. Smith Faculty of Agriculture, Food and
Environment for technical support. We would also like to thank
the equine practitioners who participated in the study as well as
Dr. Amir Steinman who made this project possible.
Evaluation of Pain in Horses
21
17/03/2016 11:04:14
Research Articles
REFERENCES
1. Traub-Dargatz, J.L., Kopral, C.A., Seitzinger, A.H., Garber, L.P.,
Forde, K. and White, N.A.: Estimate of the national incidence
of and operation-level risk factors for colic among horses in the
United States, spring 1998 to spring 1999. J. Am. Vet. Med. Assoc.
219: 67-71, 2001.
2. Muir, W.W. and Robertson, J.T.: Visceral analgesia: effects of xylazine, butorphanol, meperidine, and pentazocine in horses. Am.
J. Vet. Res. 46: 2081-2084, 1985.
3. Boatwright, C.E., Fubini, S.L., Grohn, Y.T. and Goossens, L.: A
comparison of N-butylscopolammonium bromide and butorphanol tartrate for analgesia using a balloon model of abdominal pain
in ponies. Can. J. Vet. Res. 60: 65-68, 1996.
4. Pritchett, L.C., Ulibarri, C., Roberts, C.R., Schneifer, R.K. and
Sellon, D.C.: Identification of potential physiological and behavioral indicators of post operative pain in horses after exploratory
celiotomy for colic. Appl. Anim. Behav. Sci. 80: 31-43, 2003.
5. Graubner, C., Gerber, V., Doherr, M. and Spadavecchia, C.: Clinical application and reliability of a post abdominal surgery pain
assessment scale (PASPAS) in horses. Vet. J. 188: 178-83, 2010.
6. van Loon, J.P., Jonckheer-Sheehy, V.S., Back, W., van Weeren,
P.R. and Hellebrekers, L.J.: Monitoring equine visceral pain with
a composite pain scale score and correlation with survival after
emergency gastrointestinal surgery. Vet. J. 200: 109-15, 2014.
7. Dalla Costa, E., Minero, M., Lebelt, D., Stucke, D., Canali, E. and
Leach, M.C.: Development of the Horse Grimace Scale (HGS)
as a pain assessment tool in horses undergoing routine castration.
PLoS One. 9:e92281, 2014.
8. Taffarel, M.O., Luna, S.P., de Oliveira, F.A., Cardoso, G.S.,
Alonso, J. de M., Pantoja, J.C., Brondani, J.T., Love, E., Taylor, P.,
White, K. and Murrell, J.C.: Refinement and partial validation of
the UNESP-Botucatu multidimensional composite pain scale for
assessing postoperative pain in horses. BMC Vet. Res. 11: 83, 2015.
9. Sutton, G.A., Dahan, R., Turner, D. and Paltiel, O.: A behaviourbased pain scale for horses with acute colic: scale construction.
Vet. J. 196: 394-401, 2013.
10. Sutton, G.A., Paltiel, O., Soffer, M. and Turner, D.:Validation of
two behaviour-based pain scales for horses with acute colic. Vet.
J. 197: 646-650, 2013.
11. de Vet, H.C., Terwee, C.B. and Bouter, L.M.: Current challenges
in clinimetrics. J. Clin. Epidemiol. 56: 1137-41, 2003.
12. Galea, M.P.: Introducing Clinimetrics. Aust. J. Physiother. 51:
139-140, 2005.
13. Streiner, D.L., Norman, G.R. and Cairney, J.: Health Measure-
22
Sutton, G.A.
March 2016.indb 22
ment Scales: a practical guide to their development and use. (5th
Ed.), Oxford University Press, Oxford, 2015.
14. Taylor, P.M., Pascoe, P.J. and Mama, K.R.: Diagnosing and treating pain in the horse. Where are we today? Vet. Clin. North Am:
Equine Pract. 18: 1-19, 2002.
15. Ashley, F.H., Waterman-Pearson, A.E. and Whay, H.R.: Behavioural assessment of pain in horses and donkeys: application to
clinical practice and future studies. Equine Vet J. 37: 565-75, 2005.
16. Mair, T.S. and Smith, L.J.: Survival and complication rates in 300
horses undergoing surgical treatment of colic. Part 1: Short-term
survival following a single laparotomy. Equine Vet. J. 37: 296-302,
2005.
17. McGraw, K.O. and Wong, S.P.: Forming inferences about certain
intraclass correlation coefficients. Psychol. Methods. 1: 30-46,
1996.
18. Weir, J.P.: Quantifying test-retest reliability using the intraclass
correlation coefficient and the SEM. J. Strength Cond. Res. 19:
231-240, 2005.
19. Shoukri, M.M. and Pause, C.A.: Models for reliability studies.
In: M.M. Shoukri and C.A. Pause, (Eds): Statistical Methods for
Health Sciences. CRC Press, Boca Raton. p. 27, 1999.
20. Bland, J.M. and Altman, D.G.: Statistical methods for assessing
agreement between two methods of clinical measurement. Lancet.
1: 307-310, 1986.
21. Landis, J.R. and Koch, G.G.: The measurement of observer agreement for categorical data. Biometrics. 33: 159–174, 1977.
22. Wright, J.G. and Feinstein, A.R.: A comparative contrast of clinimetric and psychometric methods for constructing indexes and
rating scales. J. Clin. Epidemiol. 45: 1201-1218, 1992.
23. Turner, D., Griffiths, A.M., Steinhart, A.H., Otley, A.R. and
Beaton, D.E.: Mathematical weighting of a clinimetric index
(Pediatric Ulcerative Colitis Activity Index) was superior to the
judgmental approach. J. Clin. Epidemiol. 62: 738-44, 2009.
24. de Jong, A.E., Bremer, M., Schouten, M., Tuinebreijer, W.E. and
Faber, A.W.: Reliability and validity of the pain observation scale
for young children and the visual analogue scale in children with
burns. Burns. 31: 198-204, 2005.
25. Voepel-Lewis, T., Zanotti, J., Dammeyer, J.A. and Merkel, S.:
Reliability and validity of the face, legs, activity, cry, consolability
behavioral tool in assessing acute pain in critically ill patients. Am.
J. Crit. Care. 19: 55-61, 2010.
26. Johnston, C.C.: Psychometric issues in the measurement of pain.
In: Finley, G.A. and McGrath, P.J. (Eds): Measurement of pain in
infants and children. IASP Press, Seattle, pp. 5-20, 1998.
Israel Journal of Veterinary Medicine  Vol. 71 (1)  March 2016
17/03/2016 11:04:14
Research Articles
Appendix 1:
a: The Revised Equine Acute Abdominal Pain Scale (EAAPS)
b: The Revised Equine Acute Abdominal Pain Scale (EAAPS) descriptions
Israel Journal of Veterinary Medicine  Vol. 71 (1)  March 2016
March 2016.indb 23
Evaluation of Pain in Horses
23
17/03/2016 11:04:15