Jospt 2014 4994

[ research report ]
ANGELA BLASIMANN, PT, MPTSc1-4 • SHARON WOOD DAUPHINEE, PT, PhD5 • J. BART STAAL, PT, PhD6
Translation, Cross-cultural Adaptation,

and Psychometric Properties of the
Downloaded from www.jospt.org at Bird Lib/OUHSC Rm 490 on December 17, 2014. For personal use only. No other uses without permission.
German Version of the Hip Disability

and Osteoarthritis Outcome Score
O
steoarthritis (OA) is the most common joint disorder which often leads to hip replacement.8
worldwide and primarily affects the knees, hips and spine.20 Very few measures are specifically de-
signed for patients with hip dysfunction.
Approximately 50% of the Swiss population aged over 65 years
The Hip disability and Osteoarthritis
and 80% aged over 75 years show radiological signs of OA,11 Outcome Score (HOOS)15 was designed
Copyright © 2014 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
to assess patients’ perceptions about their

hip. The HOOS is easy to complete and
TTSTUDY DESIGN: Clinical measurement. Universities Osteoarthritis Index. The HOOS was
provides an option to examine changes in
TTOBJECTIVES: To translate and cross-culturally also compared to the Medical Outcomes Study
pain, other symptoms, activities of daily
adapt the Hip disability and Osteoarthritis Out- 36-Item Short-Form Health Survey.
living, sport and recreation, as well as
come Score (HOOS) from English into German, TTRESULTS: Cronbach alpha values for all Ger-
and to study its psychometric properties in
hip-related quality of life (QoL). To our
man HOOS subscales were between .87 and .93.
patients after hip surgery. knowledge, there is no specific hip ques-
For test-retest reliability, the intraclass correlation
TTBACKGROUND: There is no specific hip coefficient for agreement was 0.85 for the total tionnaire in German that measures not
questionnaire in German that not only measures scores of the German HOOS. The Spearman rho only symptoms and function but also hip-
Journal of Orthopaedic & Sports Physical Therapy®
symptoms and function but also contains items for the Medical Outcomes Study 36-Item Short- related QoL.
about hip-related quality of life. Form Health Survey physical functioning subscale The objectives of the present study
TTMETHODS: The translation and cross-cultural compared to the sum of all HOOS subscales was were to translate the English HOOS into
adaptation involved forward translation, harmo- 0.71, and that for the Medical Outcomes Study conceptually equivalent German and to
nization, cognitive debriefing, back translation, 36-Item Short-Form Health Survey physical com- determine its validity and reliability. The
and comparison to the original HOOS following ponent summary was 0.97.
study protocol was approved by the Eth-
international guidelines. The German version was TTCONCLUSION: The German HOOS has demon- ics Committee of Berne, Switzerland.
tested in 51 Swiss inpatients 8 weeks after different strated adequate reliability and validity. Use of the
types of hip surgery, mainly total hip replacement.
The mean age of the participants was 62.5 years,
German HOOS is recommended for assessment
METHODS
of patients after hip surgery, with the proviso that
and the age range was from 27 to 87 years. Thirty
additional psychometric testing should be done
(58.8%) of the participants were women. Internal
in future research. J Orthop Sports Phys Ther
Translation Process
T
consistency and test-retest reliability were estimat- he translation of the HOOS
2014;44(12):989-997. Epub 13 November 2014.
ed using Cronbach alpha and intraclass correlation followed international guide-
doi:10.2519/jospt.2014.4994
coefficients for agreement. For construct validity,
TTKEY WORDS: HOOS, questionnaire, reliability,
lines.2,7,10 The preliminary German
total scores of the German HOOS were correlated
with those of the Western Ontario and McMaster validity version was tested by cognitively debrief-
ing 8 German-speaking, Swiss individu-
1
Department of Health, Berne University of Applied Sciences, Berne, Switzerland. 2Physiotherapy Salemspital, Berne, Switzerland. 3Physiotherapy Science Studies, Department
of Rheumatology and Institute of Physical Medicine, University Hospital Zurich, Zurich, Switzerland. 4Department of Epidemiology, Maastricht University, Maastricht, The
Netherlands. 5School of Physical and Occupational Therapy, McGill University, Montreal, Quebec, Canada. 6Scientific Institute for Quality of Healthcare, Radboud University
Nijmegen Medical Centre, Nijmegen, The Netherlands. This work was partly supported by a grant from the Physical Therapy Sciences Foundation, Zurich, Switzerland. This study
was approved by the Ethics Committee of Berne, Switzerland (reference number KEK-BE 153/07). The authors certify that they have no affiliations with or financial involvement
in any organization or entity with a direct financial interest in the subject matter or materials discussed in the article. Address correspondence to Angela Blasimann, Berne
University of Applied Sciences, Murtenstrasse 10, CH-3008 Berne, Switzerland. E-mail: angela.blasimann@bfh.ch; ablasimann@hotmail.com t Copyright ©2014 Journal of
Orthopaedic & Sports Physical Therapy®
journal of orthopaedic & sports physical therapy | volume 44 | number 12 | december 2014 | 989
44-12 Blasimann.indd 989 11/17/2014 4:46:43 PM

[ research report ]
als (age range, 25-87 years) with different sities Osteoarthritis Index (WOMAC) is scores, the scores for each HOOS sub-
levels of education. Minor alterations a patient-reported outcome measure for scale were added and the sum was cor-
were made. After the back translation, populations with hip and knee OA or ar- related to the total WOMAC scores. Prior
no further adjustments were needed throplasty.3,17 As the HOOS contains all to testing concurrent construct validity, a
(APPENDIX). the WOMAC questions in an unchanged scatterplot was drawn to examine linear-
form, WOMAC scores can be calculated ity and covariance of the data. Spearman
Patient Recruitment and Data Collection from the HOOS.13 The WOMAC was used correlation coefficients were used to as-
From August to December 2007, inpa- to evaluate concurrent construct validity. sess covariance between the HOOS and
tients of a private hospital in Switzerland The SF-36 The SF-36, a self-adminis- the WOMAC.
who met the following criteria and con- tered generic health status questionnaire, Before examining convergent con-
sented to participate were included in was applied to test for convergent construct validity, SF-36 scores for the PF
the study: hip problems (eg, after total struct validity.13 It includes 8 multi-item subscale and the PCS were calculated
hip replacement [THR]) or with OA of scales measuring physical functioning using a database sheet in Excel 2003
the hip (confirmed by radiography), and (PF), among other things. Physical com- (Microsoft Corporation, Redmond, WA).
adults (20 years and older) able to un- ponent summary (PCS) scores can be de- In addition, HOOS scores were trans-
derstand and fill out the questionnaires. rived from the items.6,9,19 formed with a formula provided in the
Exclusion criteria were dementia, a neu- HOOS user’s guide.15 Then, a scatterplot
rological diagnosis such as stroke, and Statistical Analysis was drawn, and the correlations between
total or partial hip replacement after a All statistical analyses were performed the HOOS and the PCS and the HOOS
femoral neck fracture. In total, 52 of 58 using SPSS Version 12.0.1 for Windows and the PF subscale of the SF-36 were
contacted subjects provided written in- (IBM Corporation, Armonk, NY), with a calculated using Spearman correlation
formed consent. level of significance of 5%. coefficients.
Eight weeks after hip surgery, all 52 Descriptive Data Descriptive statistics, Floor and Ceiling Effects A floor effect
individuals received the Medical Out- means and ranges or counts and percent, was defined as being present when a pa-
comes Study 36-Item Short-Form Health were used to display the patients’ demo- tient scored at the bottom of the HOOS
Survey (SF-36) and the German HOOS. graphic and clinical characteristics. scale (0 points out of 100), meaning that
Participants were asked to fill out both Reliability Internal consistency reliabil- he or she had no problem with the hip
questionnaires and to return the com- ity was estimated by calculating corrected anymore. The opposite was the ceiling
pleted questionnaires by mail. As soon item-to-total correlations and Cronbach effect, which occurred when an individ-
as the questionnaires were received, an- alpha for each subscale using data from ual achieved the maximum score (100
other copy of the HOOS was sent to the the first mailed HOOS. points), indicating extreme problems.
participants. The whole process took up Before conducting the test-retest re-
to 2 weeks. liability analysis, a 2-dimensional scat- RESULTS
terplot was drawn to look at the general
Instrumentation trend of the data. After assuming linearity, Subjects
O
The HOOS The 40-item HOOS contains the total scores of the 2 mailed German f the 52 patients who agreed to
5 subscales: pain (10 items), other symp- HOOS were analyzed using intraclass cor- participate and received the ques-
toms (5 items), function in activities of relation coefficients for agreement. Anoth- tionnaires, 1 did not respond, even
daily living (17 items), function in sport er way to show agreement is the method though the patient received 2 reminder
and recreation (4 items), and hip-related described by Bland and Altman,5 by which letters and 1 telephone call. All other pa-
QoL (4 items).12 To answer each ques- the limits of agreement equal the mean tients (mean age, 62.5 years; age range,
tion, patients were asked to mark 1 of the difference in scores of repeated measure- 27-87 years; women, 30 [58.8%]; men,
5 boxes scored from 0 to 4. By summing ments 1.96 SD of these differences. 21 [41.2%]) answered on time. The in-
the total score of each subscale and di- Validity Concurrent construct valid- cluded patients had different types of hip
viding by the possible maximum score for ity was examined using WOMAC scores surgery, such as conventional THR (32
the scale, the normalized score for each from the HOOS. The sum of items P4, patients [62.7%]), THR with minimal in-
subscale of the HOOS was calculated and P5, P6, P7, and P8 was taken for the vasive surgery (8 patients [15.7%]), sur-
transformed by using a formula provided WOMAC pain subscale; items S4 and gical hip luxation (6 patients [11.8%]),
in the user’s guide.15 S5 were added for WOMAC’s stiffness replacement of the artificial hip joint
The Western Ontario and McMaster subscale; and the sum of items A1 to A17 (2 patients [3.9%]), hip arthroscopy (2
Universities Osteoarthritis Index The was calculated for the WOMAC’s func- patients [3.9%]), and refixation of the
Western Ontario and McMaster Univer- tion subscale.4 To get original WOMAC greater trochanter (1 patient [2%]).
990 | december 2014 | volume 44 | number 12 | journal of orthopaedic & sports physical therapy

and the SF-36 were analyzed. The score
Characteristics and Internal of the SF-36 PF subscale was correlated
Consistency Values of Subscales of to the sum of all HOOS subscales, and a
TABLE 1
the German HOOS (n = 51) at First Mailing value of 0.71 was obtained. TABLE 3 shows
and Those of the Original HOOS 12* Spearman correlation coefficients for the
SF-36 PF and PCS subscale scores versus
HOOS Subscale German Original each HOOS subscale and the summary
Other symptoms (5 items) .87 .82 score.

Pain (10 items) .93 .93 For concurrent construct validity, the
Function in daily living (17 items) .89 .96 total of all HOOS subscales was correlat-
Function in sport and recreation (4 items) .88 .88 ed to the total WOMAC scores. The value
Hip-related quality of life (4 items) .88 .77 for Spearman rho was 0.92 (TABLE 3). No
Overall (40 items) .92 NS floor or ceiling effects were found.
Abbreviations: HOOS, Hip disability and Osteoarthritis Outcome Score; NS, not stated.
*Values are Cronbach alpha.
DISCUSSION
T
he objective of this study was
Test-Retest Reliability Overall to translate the original English
TABLE 2 and for Subscales of the German HOOS into conceptually equivalent
(n = 39) and the Original HOOS 12 German, following international guide-

lines. Furthermore, its validity and reli-
HOOS Subscale German* Original† ability were evaluated in people with hip
Other symptoms (5 items) 0.77 (0.63, 0.86) 0.86 problems after surgery.
Pain (10 items) 0.87 (0.79, 0.93) 0.89
Function in daily living (17 items) 0.80 (0.67, 0.88) 0.89 Psychometric Performance
Function in sport and recreation (4 items) 0.79 (0.66, 0.87) 0.91 of the Translated Measure
Hip-related quality of life (4 items) 0.79 (0.65, 0.87) 0.78 Reliability estimates from this study sug-
Overall (40 items) 0.85 (0.75, 0.91) NS gest that the German HOOS is internally
Abbreviations: HOOS, Hip disability and Osteoarthritis Outcome Score; NS, not stated. consistent and stable over time. All sub-
*Values are intraclass correlation coefficient for agreement (95% confidence interval).
scales met published criteria for internal
†
Values are intraclass correlation coefficient for agreement.
reliability.14
Test-retest reliability and internal
Unfortunately, 1 subject had to go to data from 39 of 51 patients who com- consistency were very good for both the
the hospital a second time because of pleted a second German HOOS were total score and the subscores. The quite
declining health. Her scores on the ques- analyzed. Correlations between the sums large mean difference could be a retest
tionnaires were included in the analysis of all subscales from the first and second bias due to (1) the selected patients being
where possible. mailed HOOS were calculated using a in an early stage of the rehabilitation pro-
2-way random-effects model.18 TABLE 2 cess and therefore not in a stable health
Translation Process shows the values for the intraclass corre- status; and/or (2) the patients having ex-
In the harmonization meeting, the most lation coefficients for agreement overall pected themselves to have a better health
discussed items were S1, P2, P3, P4, P10, and for the subscales of the German ver- status when completing the question-
A6, SP1, and SP4. sion of the HOOS. naires for the second time, which could
The Bland-Altman plot (FIGURE) be part of the bias related to the measure-
Reliability showed a mean  SD difference of 6.95 ment of change.16
To estimate internal consistency, the first  28.76, with a 95% confidence interval The German version showed simi-
mailed and returned German HOOS was of 35.71, –21.81 as the limits of agreement lar values of Cronbach alpha for all
used, and the Cronbach alpha was calcu- between the HOOS scores obtained at the subscales except for the hip-related
lated for each subscale. TABLE 1 presents 2 points in time. QoL subscale, which was higher in the
information about each HOOS subscale present study compared to that of the
and the values for the Cronbach alpha Validity original HOOS.12 A low Cronbach alpha
of the German and the original HOOS.12 To examine convergent construct valid- indicates a low correlation between the
To estimate test-retest reliability, the ity, the first distributed HOOS in German items in the subscale, which makes sum-

[ research report ]
cient of the sport and recreation subscale,
60 were similar.
Concurrent construct validity was
studied by calculating WOMAC scores
from the HOOS questionnaire. The
40
95% CI, upper boundary, 35.71 WOMAC, as part of the HOOS, is not an
Difference Between HOOS (T1 – T2)
independent comparator and therefore

led to high correlation values. Further-

20
more, the SF-36 as a generic health status
mean, 6.95
measure is probably not the best instru-
ment to be compared with a disease-spe-
0 cific questionnaire.
Another limitation could be that the
WOMAC was translated again, instead
–20 95% CI, lower boundary, –21.81 of using the validated German version by
Stucki et al.17
The lack of any floor or ceiling effects
–40
indicates that the original HOOS con-
tains items capturing patients at both
0 20 40 60 80 100 the lower and upper ends of the scale,

and that the study sample was balanced
Mean HOOS Scores (T1 and T2) regarding the severity of symptoms and/
or impairments.
FIGURE. Bland-Altman plot for the summed scores of the German HOOS at 8 weeks postsurgery (T1) and 1 week
later (T2) (n = 51). The dotted line represents the mean and the 2 continuous lines equal limits of agreement.
Abbreviations: CI, confidence interval; HOOS, Hip disability and Osteoarthritis Outcome Score; T1, time point 1
Study Strengths
(baseline: 8 weeks postsurgery); T2, time point 2 (retest: 9 weeks postsurgery). The strengths of the present study in-
clude the standardized methods used for
all procedures.2 Assessment of this pro-
Correlations Between the German HOOS, cess by examining reliability and validity
TABLE 3
Original HOOS, 13 the SF-36, and the WOMAC* does not depend on specific diagnoses of
hip problems (and by that it does not de-
German HOOS Original HOOS
pend on a specific patient selection), be-
cause all comparisons are made within
HOOS Subscale SF-36 PF SF-36 PCS WOMAC SF-36 PF
the same patient(s).1
Other symptoms (5 items) 0.46 0.46 NS 0.35
Pain (10 items) 0.56 0.52 NS 0.52
Limitations
Function in daily living (17 items) 0.64 0.62 NS 0.66
The rate of missing answers was gener-
Function in sport and recreation (4 items) 0.70 0.65 NS 0.49
ally low. Nonetheless, some patients re-
Hip-related quality of life (4 items) 0.52 0.55 NS 0.46
ported difficulties answering items S1,
Overall (40 items) 0.71 0.97 0.92 NS
A9, A13, A16, and P10. The most prob-
Abbreviations: HOOS, Hip disability and Osteoarthritis Outcome Score; NS, not stated; PCS, physical
component summary; PF, physical functioning subscale; SF-36, Medical Outcomes Study 36-Item lematic part of the questionnaire was
Short-Form Health Survey; WOMAC, Western Ontario and McMaster Universities Osteoarthritis the function in sport and recreation sub-
Index. scale. This subscale did not seem to be
*Validity statistics are Spearman rho (n = 51). All Spearman rho correlations were significant at the
P = .01 level (1-tailed). particularly applicable for seniors in the
present study.
Limitations regarding the generaliz-
ming the items unjustified.18 Conversely, The German HOOS demonstrated ability of the study results include the
Cronbach alpha values greater than .90, strong convergent construct validity. study having been conducted only in the
which was the case for 1 subscale, are in- Compared to the findings of Nilsdotter et German-speaking part of Switzerland
terpreted to mean that some items are al,13 the present study’s values for the PF and not in other European German-
redundant, which may limit the content subscale of the SF-36 versus the HOOS speaking countries. But, as language
validity.16,18 subscales, except for the correlation coeffi- differences relate only to the spoken dia-

lects, the German HOOS should be ap- tion of the minimal important differ- 10. Guillemin F, Bombardier C, Beaton D. Cross-
plicable outside of Switzerland. ence, be undertaken to complete its cultural adaptation of health-related quality of
Furthermore, the sample size was psychometric properties. Moreover, the life measures: literature review and proposed
quite small, despite the fact that, in total, applicability of the German HOOS in guidelines. J Clin Epidemiol. 1993;46:1417-1432.
11. Jüni P, Müller U. Erkrankungen des Bewegung-
267 patients underwent a THR at that German-speaking countries other than
sapparates. Available at: http://www.ispm.ch/
hospital in 2007. Moreover, the comorbid Switzerland should be examined in fur- fileadmin/SPM-Skript/Skript_Kapitel8.3.pdf.
conditions of the patients were not evalu- ther studies.
Accessed November 3, 2014.

ated, and the age range was large and not 12. Klässbo M, Larsson E, Mannevik E. Hip dis-
very common for hip OA. However, be- ability and Osteoarthritis Outcome Score. An
cause the aim of the study was to evalu- REFERENCES extension of the Western Ontario and McMaster
Universities Osteoarthritis Index. Scand J
ate the psychometric properties of the Rheumatol. 2003;32:46-51. http://dx.doi.
1. Angst F, Goldhahn J, Pap G, et al. Cross-
German HOOS and not to obtain norms cultural adaptation, reliability and validity of org/10.1080/03009740310000409)
for groups, the patient selection was ac- the German Shoulder Pain and Disability Index 13. Nilsdotter AK, Lohmander LS, Klässbo M,
ceptable. Another important limitation (SPADI). Rheumatology (Oxford). 2007;46:87- Roos EM. Hip disability and Osteoarthritis
92. http://dx.doi.org/10.1093/rheumatology/ Outcome Score (HOOS) – validity and re-
is the lack of longitudinal data to study
kel040 sponsiveness in total hip replacement. BMC
responsiveness. 2. Beaton DE, Bombardier C, Guillemin F, Ferraz Musculoskelet Disord. 2003;4:10. http://dx.doi.
MB. Guidelines for the process of cross-cultural org/10.1186/1471-2474-4-10
CONCLUSION adaptation of self-report measures. Spine 14. Nunnally JC. Psychometric Theory. New York,
(Phila Pa 1976). 2000;25:3186-3191. NY: McGraw-Hill; 1978.
T
3. Bellamy N. WOMAC Osteoarthritis Index. A 15. Roos E. Knee injury and Osteoarthritis Out-
he strong levels of reliability
User’s Guide. London, Canada: University of come Score. Available at: http://www.koos.nu/.
and validity support the potential Western Ontario; 1995. Accessed November 3, 2014.
of the German HOOS for use for 4. Bellamy N, Kean WF, Buchanan WW, Gerecz- 16. Streiner DL, Norman GR. Health Measurement
postoperative hip patients. Additional Simon E, Campbell J. Double blind randomized Scales: A Practical Guide to Their Development
controlled trial of sodium meclofenamate (Me- and Use. 3rd ed. Oxford, UK: Oxford University
psychometric testing, including respon-
clomen) and diclofenac sodium (Voltaren): post Press; 2003.
siveness to change and calculation of the validation reapplication of the WOMAC Osteo- 17. Stucki G, Meier D, Stucki S, et al. [Evaluation of
minimal important difference, should be arthritis Index. J Rheumatol. 1992;19:153-159.
undertaken. t
a German version of WOMAC (Western Ontario
5. Bland JM, Altman DG. Statistical meth-
and McMaster Universities) Arthrosis Index]. Z
ods for assessing agreement between two
Rheumatol. 1996;55:40-49.
methods of clinical measurement. Lancet.
KEY POINTS 1986;327:307-310. http://dx.doi.org/10.1016/

18. Terwee CB, Bot SD, de Boer MR, et al. Qual-
FINDINGS: The German version of the ity criteria were proposed for measurement
S0140-6736(86)90837-8
properties of health status questionnaires. J
HOOS is reliable and valid in this small 6. Bullinger M, Kirchberger I. SF-36 Fragebogen
zum Gesundheitszustand: Handanweisung. Clin Epidemiol. 2007;60:34-42. http://dx.doi.
group of Swiss patients with different org/10.1016/j.jclinepi.2006.03.012
Göttingen, Germany: Hogrefe; 1998.
hip problems. 7. Dauphinee SW, Gauthier L, Gandek B, Magnan 19. Ware JE, Jr., Sherbourne CD. The MOS 36-Item
IMPLICATIONS: This specific hip question- L, Pierre U. Readying a US measure of health Short-Form Health Survey (SF-36). I. Concep-
naire in German not only evaluates status, the SF-36, for use in Canada. Clin Invest tual framework and item selection. Med Care.
Med. 1997;20:224-238. 1992;30:473-483.
symptoms and function but also con-
8. Felson DT, Lawrence RC, Dieppe PA, et al. 20. Zacher J, Feldman D, Gerli R, et al. A com-
tains items about hip-related QoL. It Osteoarthritis: new insights. Part 1: the parison of the therapeutic efficacy and
can be easily accessed at http://www. disease and its risk factors. Ann Intern tolerability of etoricoxib and diclofenac
koos.nu and downloaded free of charge Med. 2000;133:635-646. http://dx.doi. in patients with osteoarthritis. Curr Med
in different languages. org/10.7326/0003-4819-133-8-200010170- Res Opin. 2003;19:725-736. http://dx.doi.
00016 org/10.1185/030079903125002469
CAUTION: Use of the German HOOS for
9. Finch E, Brooks D, Stratford P, Mayo N. Physical
hip patients after surgery is recom- Rehabilitation Outcome Measures: A Guide to
@
mended, with the proviso that addi- Enhanced Clinical Decision Making. 2nd ed.
Baltimore, MD: Lippincott Williams & Wilkins;
MORE INFORMATION
tional psychometric testing, including
2002. WWW.JOSPT.ORG
responsiveness to change and calcula-

44-12 Blasimann.indd 994

APPENDIX
11/17/2014 5:49:41 PM

APPENDIX
journal of orthopaedic & sports physical therapy | volume 44 | number 12 | december 2014 |
995
11/17/2014 5:49:41 PM

APPENDIX
11/17/2014 5:49:42 PM

APPENDIX
journal of orthopaedic & sports physical therapy | volume 44 | number 12 | december 2014 |
997
11/17/2014 5:49:42 PM

Jospt 2014 4994

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Jospt 2014 4994

Uploaded by

Copyright:

Available Formats

[ research report ]

Translation, Cross-cultural Adaptation,

German Version of the Hip Disability

to assess patients’ perceptions about their

44-12 Blasimann.indd 989 11/17/2014 4:46:43 PM

44-12 Blasimann.indd 990 11/17/2014 4:46:43 PM

Other symptoms (5 items) .87 .82 score.

(n = 39) and the Original HOOS 12 German, following international guide-

44-12 Blasimann.indd 991 11/17/2014 4:46:44 PM

independent comparator and therefore

led to high correlation values. Further-

0 20 40 60 80 100 the lower and upper ends of the scale,

44-12 Blasimann.indd 992 11/17/2014 4:46:45 PM

Accessed November 3, 2014.

KEY POINTS 1986;327:307-310. http://dx.doi.org/10.1016/

44-12 Blasimann.indd 993 11/17/2014 4:46:45 PM

44-12 Blasimann.indd 994

44-12 Blasimann.indd 995

44-12 Blasimann.indd 996

44-12 Blasimann.indd 997

You might also like