Sensibility of five at-work productivity measures was endorsed by patients with osteoarthritis or rheumatoid arthritis

Article (PDF Available)inJournal of clinical epidemiology 66(5) · March 2013with39 Reads
DOI: 10.1016/j.jclinepi.2012.12.009 · Source: PubMed
Abstract
Objective: To examine and compare the sensibility attributes (face/content validity and feasibility) of five at-work productivity measures from the perspective of patients with osteoarthritis (OA) or rheumatoid arthritis (RA). Study design and setting: Workers with OA or RA (n = 250) completed a survey that includes five at-work productivity (presenteeism) measures and questions asking about their comprehensiveness, understandability, length, and suitability of response options. A final question asked respondents which single measure was considered "best" overall. Measures compared included the Workplace Activity Limitations Scale (WALS), Stanford Presenteeism Scale, Endicott Work Productivity Scale, Work Instability Scale for Rheumatoid Arthritis (RA-WIS), and Work Limitations Questionnaire (WLQ-25). Sensibility performance was assessed quantitatively (% respondent endorsement) and qualitatively via written feedback. Results: The WLQ-25 was considered most comprehensive (endorsed by 92.8%), the WALS performed best in terms of understandability (97.6%) and suitability of response options (97.9%), and the RA-WIS was favored in terms of length (91.6%). Consistent sensibility performance between OA and RA was found. The WALS (32.6%) and WLQ-25 (30.0%) were moderately preferred in the final overall appraisal. Conclusion: Sensibility criteria were generally met by all five at-work productivity measures. Variable endorsement levels across specific sensibility attributes were also revealed across the measures compared.
Sensibility of five at-work productivity measures was endorsed
by patients with osteoarthritis or rheumatoid arthritis
Kenneth Tang
a,b,c,
*
, Dorcas E. Beaton
a,b,c
, Diane Lacaille
d,e
, Monique A.M. Gignac
f,g
,
Claire Bombardier
b,c,h,i,j
, and Canadian Arthritis Network (CAN) Work Productivity Group
a
Mobility Program Clinical Research Unit, Li Ka Shing Knowledge Institute of St. Michael’s Hospital, 30 Bond Street, Toronto, Ontario, Canada M5B1W8
b
Institute of Health Policy, Management and Evaluation, University of Toronto, 155 College Street, Suite 425, Toronto, Ontario, Canada M5T 3M6
c
Institute for Work & Health, 481 University Avenue, Suite 800, Toronto, Ontario, Canada M5G 2E9
d
Arthritis Research Centre of Canada, Milan Ilich Arthritis Research Centre, 5591 No. 3 Road, Richmond, British Columbia, Canada V6X 2C7
e
Division of Rheumatology, Department of Medicine, University of British Columbia, 2775 Laurel Street, 10th Floor, Vancouver,
British Columbia, Canada V5Z 1M9
f
Arthritis Community Research and Evaluation Unit, Toronto Western Research Institute, University Health Network, 399 Bathurst Street, MP-10th Floor,
Suite 316, Toronto, Ontario, Canada M5T 2S8
g
Dalla Lana School of Public Health, University of Toronto, 155 College Street, 6th floor, Toronto, Ontario, Canada M5T 3M7
h
Division of Clinical Decision Making and Health Care, Toronto General Research Institute, University Health Network, 7-504, 610 University Avenue,
Toronto, Ontario, Canada M5G 2M9
i
Division of Rheumatology, Department of Medicine, University of Toronto, Suite RFE 3-805, 200 Elizabeth Street, Toronto, Ontario, Canada M5G 2C4
j
Mount Sinai Hospital, 600 University Avenue, Toronto, Ontario, Canada M5G 1X5
Accepted 23 December 2012; Published online 1 March 2013
Abstract
Objective: To examine and compare the sensibility attributes (face/content validity and feasibility) of five at-work productivity
measures from the perspective of patients with osteoarthritis (OA) or rheumatoid arthritis (RA).
Study Design and Setting: Workers with OA or RA (n 5 250) completed a survey that includes five at-work productivity (presentee-
ism) measures and questions asking about their comprehensiveness, understandability, length, and suitability of response options. A final
question asked respondents which single measure was considered ‘best’ overall. Measures compared included the Workplace Activity Lim-
itations Scale (WALS), Stanford Presenteeism Scale, Endicott Work Productivity Scale, Work Instability Scale for Rheumatoid Arthritis
(RA-WIS), and Work Limitations Questionnaire (WLQ-25). Sensibility performance was assessed quantitatively (% respondent endorse-
ment) and qualitatively via written feedback.
Results: The WLQ-25 was considered most comprehensive (endorsed by 92.8%), the WALS performed best in terms of understand-
ability (97.6%) and suitability of response options (97.9%), and the RA-WIS was favored in terms of length (91.6%). Consistent sensibility
performance between OA and RA was found. The WALS (32.6%) and WLQ-25 (30.0%) were moderately preferred in the final overall
appraisal.
Conclusion: Sensibility criteria were generally met by all five at-work productivity measures. Variable endorsement levels across spe-
cific sensibility attributes were also revealed across the measures compared. Ó 2013 Elsevier Inc. All rights reserved.
Keywords: Outcome measure; Face and content validity; Feasibility; Work productivity; Presenteeism; Arthritis
Funding: Funding for this study was provided by a research grant from
the Canadian Arthritis Network (part of the Networks of Centres of Excel-
lence) in partnership with The Arthritis Society of Canada and also by an
unrestricted grant from Abbott. Mr. K.T. is a recipient of a Canadian Insti-
tutes of Health Research (CIHR) PhD Fellowship, Canadian Arthritis Net-
work Graduate Award, and Syme Fellowship from the Institute for Work &
Health. Dr. D.E.B. was supported by a CIHR New Investigators award dur-
ing the conduct of this study. Dr. D.L. holds the Mary Pack Chair in Arthri-
tis Research from the University of British Columbia and The Arthritis
Society of Canada. Dr. C.B. is a recipient of a Canada Research Chair in
Knowledge Transfer for Musculoskeletal Care and a Pfizer Chair in
Rheumatology (Division of Rheumatology, Faculty of Medicine, Univer-
sity of Toronto).
Conflict of interest statement: All authors declare no financial or intel-
lectual conflicts of interest. We declare that all sources of funding support
had no direct role in the study design, data collection, analysis and inter-
pretation of the data; writing of the article; approval of article content;
or in the decision to publish this work. Neither the submission nor publi-
cation of this article was contingent on the approval of Abbott.
* Corresponding author. Tel.: þ1-416-864-6060x77031; fax: þ1-416-
864-5003.
E-mail address: ken.tang@mail.utoronto.ca (K. Tang).
0895-4356/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jclinepi.2012.12.009
Journal of Clinical Epidemiology 66 (2013) 546e556
Downloaded from ClinicalKey.com at University of British Columbia April 19, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.
1. Introduction
The impact of arthritis on a person’s ability to meet work
demands is an important concern [1e4], although the ideal
measure(s) to capture the extent of this impact remains un-
clear. In addition to the traditional indicators of work ab-
senteeism (e.g., days off work), recent studies have
emphasized the importance of examining ‘on-the-job’
problems (at-work productivity loss or presenteeism) expe-
rienced by workers with arthritis [5e7]. As more and more
individuals with arthritis are able to continue to work given
recent advancements in therapies, the need for accurate and
precise evaluations of presenteeism has gained importance
and research attention. In fact, compare d with absenteeism,
presenteeism has shown to contribute to an even greater
proportion of the indirect economic costs of arthritis
[8,9], giving this concept clear economic relevance. The
measures of presenteeism measures are also increasingly
used as study outcomes in rheumatology clinical trials
[10e12] as there is an increasing recognition that work
issues and potential cost benefits of therapeutic interven-
tions are important to different stakeholder s, including
patients/workers, employers, industry, and policy makers.
The number of self-report presenteeism measures is on
the rise [5,13e15], many of which have potential applica-
bility in clinical trials or employment-related research.
Some could also be applied to estimate costs associated
with health-related productivity loss. For example, the
Work Limitations Questionnaire (WLQ-25) [16] assesses
the proportion of time workers have difficulty over various
work domains. The Workplace Activity Limitatio ns Scale
(WALS) [17] takes a different approach as it is aimed at
quantifying the degree of difficulty a worker experiences
while performing various job-related tasks. Yet, another ex-
ample is the Work Instability Scale for Rheumatoid Arthri-
tis (RA-WIS) [18], which is designed as a prognostic
indicator of future work loss and has potential applicability
to help inform vocational decisi on making (e.g., the need
for workplace interventions).
To quantify patient experiences (e.g., symptoms, work,
and health-related quality of life), choosing the ideal out-
come measure(s) in a given situation may involve not only
psychometric considerations but also ‘sensibility’ consid-
erations. Sensibility is a term originally coined by Feinstein
[19] to describe the importance for instruments to demon-
strate fundamental attributes such as face/content validity
and feasibility. Recently, others have also emphasized the
need to directly appraise these qualitative attributes (also
referred to as clinical utility, practicality,orapplicability)
from the perspectives of both end users (e.g., researchers/
clinicians) and respondents (e.g., patients) [20e22]. In to-
day’s patient-oriented approach to health care, engaging pa-
tients in the development/testing of outcome measures are
increasingly relevant in rheumatology [23,24] and also
mandated by regulatory agencies, such as the US Food
and Drug Administration [25]. When assessing the value
of a health intervention, it is important to be able to dem-
onstrate efficacy on outcomes that capture concepts deemed
meaningful (i.e., what matters) to the target patient popula-
tion. Irrespective of its psychometric robustness, if an out-
come measure fails to meet conceptual needs, or if it is
impractical to apply, it may not be the o ptimal choice for
a given circumstance.
Sensibility appraisals of work outcome measures are rel-
evant for several additional reasons: (1) there is substantial
diversity in available perspectives and approaches to quan-
tify the impact of health on work, but specific work con-
cepts (e.g., ability vs. productivity) that resonate most
with patients/workers remain unclear [5,13]; (2) job context
can vary considerably among workers; thus, there is a need
to examine whether specific work measures are similarly
relevant across different occupational sectors; and (3) the
evolving nature of the employment and labor market
(e.g., change in job demands over time because of techno-
logical advancements) entails a need for periodic (re)ap-
praisals of available outcome measur es to ensure that
they remain optimal for capturing what matters to the
present-day worker.
Research that examines the direct comp arability of mea-
sures in a controlled sample is useful for gaining insights on
the measures’ relative strengths and limitations
[5,7,13,26,27]. To date, however, most head-to-head studies
on work measures have mainly focused on psychometrics
[27e34] as comparisons of sensibility attributes have been
rarely evaluated. This study examined and compared the
sensibility attributes (comprehensiveness, understandabil-
ity, length, and suitability of response options) of five
at-work productivity measures from the perspective of pa-
tients with osteoarthritis (OA) or rheumatoid arthritis (RA).
2. Methods
2.1. Participants
Study participants were workers with arthritis (n 5
250)
recruited
by convenience sampling from three sites: two
tertiary-level rheumatology clinics in urban teaching
hospitals (n 5 142) in Toronto, Ontario, Canada, and an
outpatient arthritis treatment program providing multi-
disciplinary services (n 5 108) in Vancouver, British Co-
lumbia, Canada. Inclusion criteria were (1) attendance at
an outpatient rheumatology clinic with a rheumatologist
diagnosis of either OA or RA (Toronto) or attendance at
an arthritis treatment program within the past 2 years, with
OA or RA recorded as the reason for referral by the refer-
ring physician (Vancouver); (2) participating in paid or un-
paid work (e.g., homemaking) within 1 month before
recruitment; and (3) providing informed written consent.
Respondents were excluded if they did not speak English,
547K. Tang et al. / Journal of Clinical Epidemiology 66 (2013) 546e556
Downloaded from ClinicalKey.com at University of British Columbia April 19, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.
What is new?
Patients’ viewpoints on the sensibility (face/con-
tent validity and feasibility) of self-reported out-
come measures and their ability to capture what
matters are rarely investigated. This study pro-
vided a head-to-head comparison of the sensibility
of five at-work productivity measuresdproviding
a perspective beyond psychometrics.
Methods to appraise the qualitative attributes of
patient-reported outcome measur es are scarce in
the literature. A novel, quantitative, group-level ap-
proach to evaluate the sensibility of work outcome
measures from the patients’ perspective was
applied.
Variable performance across four key sensibility
criteria was found among the five at-work produc-
tivity measures compared. Overall, the Workplace
Activity Limitations Scale and Work Limitations
Questionnaire were most preferred by patients with
osteoarthritis or rheumatoid arthritis.
Users engaged in an evidence-based process to se-
lect work outcome measures for application in ar-
thritis should consider sensibility attributes in
conjunction with available psychometric evidence
to inform their choice(s).
which was the only available language for the survey. Re-
search ethics approval was obtained from all participating
institutions.
2.2. Data collection
This study included only baseline data (collected between
2005 and 2006) from a 12-month survey as sensibility of pre-
senteeism measures was assessed at baseline. A detailed de-
scription of the study participants has been published
elsewhere [28]. At each recruitment site, participants were ran-
domly assigned to complete the study questionnaires (self-
administration) in either paper (n 5 130) or computer touch
screen (n 5 120) formats. The baseline survey comprised
questions on sociodemographic, health-, and work-related
variables, in addition to the five presenteeism measures. The
measures chosen were (1) considered to have potential appli-
cability in arthritis, (2) specifically focused on at-work diffi-
culties, and (3) of interest for comparison because they
offered different conceptualizations of presenteeism (Table 1).
2.3. Measures (fielded in the following order)
2.3.1. Workplace Activity Limitations Scale
The 12-item WALS is designed to capture arthritis-
related workplace activity limitations [17,35,36]. Scale
items assess the degree of difficulty with various job-
related tasks that tax upper and/or lower limb function
(e.g., gripping and crouching) and the difficulties with com-
muting, scheduling, concentration, and pace of work. Re-
spondents are asked about difficulties ‘in general’ or
‘typically’ but without a defined recall period. Response
options are based on a four-point Likert scale from ‘no dif-
ficulty (score 0)’ to ‘not able to do (score 3)’ plus two
other options: ‘difficulty unrelated to arthritis’ and ‘not
applicable’ (both assigned a score of 0). The WALS has
a score range between 0 and 36.
2.3.2. Stanford Presenteeism Scale
The Stanford Presenteeism Scale (SPS) 6 [37] is a short
derivative of the original SPS-32 [38]. Intended for nonspe-
cific clinical and employee populations, the SPS-6 mea-
sures a worker’s ability to focus on and complete work
tasks despite health-related distractions, based on a 1-
month recall period [37,38]. Scale content of the SPS-32
was originally formulated based on a literature review
and developers’ experience [37]. The SPS consists of six
questions on a five-point Likert scale, with response options
from ‘strongly agree (score 1)’ to ‘strongly disagree
(score 5)’ to provide a scale score from 6 to 30 (30, most
presenteeism).
2.3.3. Endicott Work Productivity Scale
The Endicott Work Productivity Scale (EWPS) was de-
veloped to quantify the frequency of unproductive attitudes
and behaviors over a 1-week period, for a broad range of
diseases and occupations [38,39]. It covers four domains:
attendance, quality of work, performance capac ity, and per-
sonal factors (social, mental, physical, and emotional) in 25
items. Five response options are offered for each item
(never, score 0, to almost always, score 4), and the scale
is scored of 100 (100, lowest productivity).
2.3.4. Work Instability Scale for Rheumatoid Arthritis
The RA-WIS was developed for RA to assess the degree
of mismatch between workers’ functional abilities and job
demands and identify individuals in need of workplace
modifications to sustain employment [18]. Scale items were
initially generated from qualitative interviews with workers
with early RA. Respondents are asked to respond based on
experiences ‘at the moment. The scale consists of 23
items (yes, score 1, and no, score 0), in which a total score
of !10 represents low work instability (WI), 10e 17 repre-
sents moderate WI, and O17 represents high WI [18]
.
Hi
gher WI has shown to be predictive of future arthritis-
related work transitions (e.g., disability leaves and job
changes) among workers with arthritis [40].
2.3.5. Work Limitations Questionnaire
The WLQ-25 was developed to measure the impact of
chronic diseases and treatment on work performance
[16,41] and has been previously validated in both OA
548 K. Tang et al. / Journal of Clinical Epidemiology 66 (2013) 546e556
Downloaded from ClinicalKey.com at University of British Columbia April 19, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.
Table 1. Summary of the conceptual and scaling properties of the five at-work productivity measures compared
Measures Concept
Intended
population
Number
of items
Response options (number
of categories) Items with 5% ‘‘not applicable’’ response
a
Item nonresponse
b
Mean ± SD
%of
scale
WALS Degree of difficulty with specific job-related
tasks
Arthritis (IA
and OA)
12 No difficulty to not able to do (4)
c
OA: one itemdprolonged standing (9.4%)
RA: two itemsdprolonged sitting (5.0%);
prolonged standing (5.9%)
OA: 0.1 6 0.5
RA: 0.1 6 0.4
1.1
0.7
SPS-6 Degree of agreement on level of
concentration at work and ability to
complete work
Nonspecific 6 Strongly disagree to strongly agree (5) Not applicable response option not offered OA: 0.0 6 0.2
RA: 0.0 6 0.1
0.7
0.3
EWPS Frequency of unproductive work behaviors Nonspecific 25 Never to almost always (5) Not applicable response option not offered OA: 0.3 6 1.0
RA: 0.5 6 1.9
1.3
2.0
RA-WIS Degree of mismatch between functional
abilities and work demands
RA 23 Yes or no (2) Not applicable response option not offered OA: 0.2 6 0.9
RA: 0.6 6 2.5
1.0
2.4
WLQ-25 Proportion of time with difficulty with
specific work experiences
Chronic health
conditions
25 None of the time to all of the time (5)
d
OA: three itemsdlifting, carrying, moving
object O10 lbs (35.4%); repeating the
same motion (20.2%); helping others get
work done (10.0%).
RA: four itemsdsticking to routine/
schedule (10.1%); lifting, carrying,
moving objects O10 lbs (30.3%);
repeating the same motion (19.5%);
bending, twisting, reaching (14.4%).
OA: 1.5 6 2.6
RA: 1.5 6 2.2
5.8
6.0
Abbreviations: SD, standard deviation; WALS, Workplace Activity Limitations Scale; SPS, Stanford Presenteeism Scale; EWPS, Endicott Work Productivity Scale; RA-WIS, Work Instability
Scale for Rheumatoid Arthritis; WLQ-25, Work Limitations Questionnaire; RA, rheumatoid arthritis; OA, osteoarthritis; IA, inflammatory arthritis.
a
Not applicable response options: ‘‘not applicable’’ for the WALS and ‘‘does not apply to my job’’ for the WLQ-25.
b
Items not completed by respondent (i.e., incompletes).
c
Two additional response options are available for WALS items: ‘‘difficulty unrelated to arthritis’’ and ‘‘not applicable’’ (both scored as 0).
d
Does not apply to my job response option is also available for WLQ-25 items (treated as missing value, if selected).
549K. Tang et al. / Journal of Clinical Epidemiology 66 (2013) 546e556
Downloaded from ClinicalKey.com at University of British Columbia April 19, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.
[42] and RA [43]. Que stionnaire items ask about the propor-
tion of time (none of the time to all the time) over a 2-week
period, in which difficulty is experienced in four domains/
subscales (each scored from 0 to 100): time management
(five item s), physical demands (PD, six items), mental inter-
personal (nine items), and output demands (five items). Con-
tent and format of the WLQ-25 were originally derived from
focus groups and cognitive interviewing of workers with
gastrointestinal disorders, psychiatric disorders, respiratory
disorders, or epilepsy [16]. In relation to other subscales, in-
struction for items in the PD subscale is reversed by design
(i.e., asks about proportion of time without difficulty). Lern-
er et al. [44] also developed a formula (WLQ Index) that
converts WLQ -25 subscale scores into an estimate of per-
centage productivity loss compared with healthy controls,
which provides a metric of economic impact.
2.4. Sensibility assessment
Based on a review of the existing theoretical frameworks
[19e22], four broad criteria were designated for direct eval-
uation from patients, (1) comprehensiveness, (2) understand-
ability, (3) appropriateness of length, and (4) suitability of
response options, as these were considered key attributes that
an outcome measure with high sensibility should demon-
strate. If a measure is not sufficiently comprehensive, it sug-
gests that relevant constructs had been omitted. If a measure
is not understandable by the respondents, it suggests that
items were not properly phrased even if the instrument was
conceptually on target. If the measure is deemed too short
or too long by respondents, it provides another indication
of the lack of comprehensiveness and/or the possibility of re-
dundant content, in addition to feasibility issues (e.g., respon-
dent burden). And finally, if a measure does not have suitable
response options, then item responses would unlikely reflect
respondents’ experiences in a faithful manner. Thus, en-
dorsement of these collective sensibility criteria would affirm
that a measure is capturing the patients’ work experience well
and also feasible for use. To evaluate these criteria, we had
developed four multiple-choice primary questions (three
yes/no items and one item with Likert scale response op-
tions), plus two additional openeended questions to invite
qualitative (written) feedback from respondents. These six
questions were placed immediately after each of the five
presenteeism measures in the survey (Table 2). At the very
end of the survey, we asked ‘which of the five measures do
you think is best for asking people with arthritis about their
difficulties at work?’ to allow respondents to offer a final
overall appraisal. This set of sensibility questions has been
successfully applied in a previous study [27] and was con-
sidered appropriate and relevant for the present study. Fi-
nally, for each measure, we also evaluated the frequency of
‘not applicable’ responses (available only for WALS and
WLQ-25) and nonresponse (i.e., incompletes) at the item
level to gain further insights on content relevance and
comprehensiveness.
2.5. Analytic approach and performance criteria for
sensibility
A quantitative approach was considered most ideal for
providing an objective evaluation of the sensibility of
Table 2. Questions fielded to assess the sensibility of at-work productivity measures
Sensibility assessment Associated question(s) (response options)
Performance criteria (O70% hypothesized
for all measures)
Criterion #1: Comprehensiveness Does this questionnaire ask enough about how your
arthritis has affected your ability to work? ( yes
a
or no)
Additional openeended question: if no, why?
!70% ‘‘yes’’ 5 poor
70e79.9% ‘‘yes’’ 5 acceptable
80e89.9% ‘‘yes’’ 5 good
90e94.9% ‘‘yes’’ 5 very good
95e100% ‘‘yes’’ 5 excellent
Criterion #2: Understandability Did you have any difficulty understanding this
questionnaire? ( yes or no
a
)
Additional openeended question: if yes, why?
!70% ‘‘no’’ 5 poor
70e79.9% ‘‘no’’ 5 acceptable
80e89.9% ‘‘no’’ 5 good
90e94.9% ‘‘no’’ 5 very good
95e100% ‘‘no’’ 5 excellent
Criterion #3: Appropriateness of length Did you find this questionnaire to be (too long, a little too
long, a good length
a
, a little short, or too short)?
!70% ‘‘a good length’’ 5 poor
70e79.9% ‘‘a good length’’ 5 acceptable
80e89.9% ‘‘a good length’’ 5 good
90e94.9% ‘‘a good length’’ 5 very good
95e100% ‘‘a good length’’ 5 excellent
Criterion #4: Suitability of response
options
Did the wording of the response options make sense to
you? ( yes
a
or no)
!70% ‘‘yes’’ 5 poor
70e79.9% ‘‘yes’’ 5 acceptable
80e89.9% ‘‘yes’’ 5 good
90e94.9% ‘‘yes’’ 5 very good
95e100% ‘‘yes’’ 5 excellent
Final/overall appraisal question Which questionnaire do you think is best for asking
people with arthritis about their work? ( forced choice
among five measures)
Not applicable
a
Response choice indicates ‘‘endorsement’’ of sensibility criterion by respondent.
550 K. Tang et al. / Journal of Clinical Epidemiology 66 (2013) 546e556
Downloaded from ClinicalKey.com at University of British Columbia April 19, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.
presenteeism measures at the group level; yet, few perfor-
mance guidelines directly applicable for such evaluations
currently exist. In the study by Rowe and Oxman [45],an
early application of Feinstein’s framework, a 12-item ques-
tionnaire (each item on a seven-point Likert scaling, 0e7),
was applied to examine the sensibility of a quality-of-life in-
strument among patients with asthma. Their expectation was
that an acceptable instrument would achieve a sample mean
of 5 (of 7) for at least 10 of 12 items, with no items having
a mean of 3 on the sensibility questionnaire. Approximat-
ing them, we have adopted a 70% patient endorsement rate
as the minimum standard for acceptable sensibility in our
study, applicable for each of the four primary sensibility
questions (Table 2). For each question, chi-square tests of
proportion were also performed to evaluate whether sensibil-
ity performance differed between arthritis type (OA vs. RA)
or mode of survey administration (paper vs. touch screen)
subgroups. Feedback comments provided by patients on
openeended questions were reviewed, and the two most fre-
quently expressed comments pertaining to each of the five
measures were summarized. For the final appraisal question,
we examined the response distribution for the full cohort and
also after stratification by arthritis type, method of survey ad-
ministration, and broad employment groups based on the Na-
tional Occupational Classification developed by Human
Resources and Skills Development Canada [46]. All quanti-
tative data analysis was conducted using SAS Version 9.1
(Statistical Analysis System Institute, Cary, NC).
3. Results
The study sample consisted of 250 workers with either
OA (n 5 130) or RA (n 5 120); the major ity of whom were
female (82.7%). Mean age was 50.6 years [standard devia-
tion (SD), 9.2; range, 19e65]. Workers were employed in
various occupational sectors: ‘business, finance, and ad-
ministration’ (44.1%), ‘health, science, art, and sports’
(31.1%), ‘sales and services’ (17.6%), and ‘trades, trans-
port, and equipment operators’ (7.2%). At the time of sur-
vey, most respondents were engaged in either full-time
(68.8%; mean hours worked/week, 41.5; SD, 11.8) or
part-time employment (27.2%; mean hours worked/week,
23.5; SD, 11.2). Other participants (4.0%) were either
homemakers, doing volunteer work, or currently on sick
leave because of arthritis (but working within the past
month).
3.1. Comprehensiveness
A summary of patient endorsem ent levels for each sen-
sibility criterion is provided in Figures 1 and 2. In terms of
comprehensiveness, 92.8% of the respondents endorsed the
WLQ-25, which was highest among the five measures. In
contrast, the WALS was bo rderline-acceptable (70% en-
dorsement), suggesting that there may be room for some
improvements in this area. Endorsement levels for the other
three measures ranged between 73.4 and 85.6%. Frequency
of ‘not applicable’ type responses for either the WLQ-25
or WALS was generally low, which was indicative of good
content relevance (Table 1). However, a few notable excep-
tions were found in the WLQ-25 PD subscale. The ‘does
not apply to my job’ option was selected by O20% of re-
spondents for lifting/carrying/moving objects O10 pounds
(OA, 35.4%; RA, 30.3%) and repeating the same motion
(OA, 20.2%; RA, 19.5%). In terms of item nonresponse
(i.e., incompletes), low rates were found across all five
measures (0.3e6.0%).
Fig. 1. Level of patient endorsement (%) for each of five at-work productivity measures across four sensibility criteria (full sample). WALS, Work-
place Activity Limitations Scale; SPS, Stanford Presenteeism Scale; EWPS, Endicott Work Productivity Scale; RA-WIS, Work Instability Scale for
Rheumatoid Arthritis; WLQ-25, Work Limitations Questionnaire.
551K. Tang et al. / Journal of Clinical Epidemiology 66 (2013) 546e556
Downloaded from ClinicalKey.com at University of British Columbia April 19, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.
3.2. Understandability
Understandability was endorsed by O90% of respon-
dents for four of five presenteeism measures. The WALS
and EWPS were ranked highest with 97.6% support each.
The WLQ-25 was endorsed by 78.6% of the respondents.
3.3. Appropriateness of length
Four of five measures were considered to have ‘a good
length’ by O70% of patients, but only two had O80%
support (EWPS and RA-WIS). The SPS-6 fell just short
of the endorsement level required (68.8% felt that it was
of a good length), with approximately a quarter of the
sample indicating that it was ‘a little too short’ (24.7%).
Respondents who did not consider the WALS to be of good
length mostly felt that it was a little too short (18.0%). For
the WLQ-25, ‘a little too long’ was endorsed second most
(17.0%).
3.4. Suitability of response options
All five measures rated high in terms of suitability of re-
sponse options as endorsement levels ranged from 86.2%
(WLQ-25) to 97.9% (WALS) . At the arthritis-type sub-
group level, the range of support was between 90.8% and
96.1% for OA and between 81.0% and 100.0% for RA.
Fig. 2. Level of patient endorsement (%) for each of five at-work productivity measures across four sensibility criteria [stratified by arthritis type: OA
(top graph) vs. RA (bottom graph)]. WALS, Workplace Activity Limitations Scale; SPS, Stanford Presenteeism Scale; EWPS, Endicott Work Pro-
ductivity Scale; RA-WIS, Work Instability Scale for Rheumatoid Arthritis; WLQ-25, Work Limitations Questionnaire; OA, osteoarthritis; RA, rheu-
matoid arthritis.
552 K. Tang et al. / Journal of Clinical Epidemiology 66 (2013) 546e556
Downloaded from ClinicalKey.com at University of British Columbia April 19, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.
For all five measures, no statistical differences in en-
dorsement levels were found between either arthritis type
or method of survey administration subgroups over any of
the sensibility criterion assessed (all chi-square tests,
P O 0.05).
3.5. Feedback from respondents
The two most frequent comments by respondents with
regard to either the comprehensiveness or understandability
of the measures are summarized in Table 3. Of note, ‘con-
tent not specific enough to my job’ was expressed by at
least some participants for all five measures.
3.6. Overall preference
The final apprais al revealed a moderate preference for
two measures, the WALS and WLQ-25, which were consid-
ered ‘best’ by 32.6% and 30.0% of our sample, respec-
tively ( Table 4). The SPS-6 was least preferred overall
(2.6% consi dered as best). No significant differences in
overall preference of measures were found for arthritis
type, method of survey administration, or occupational sec-
tor subgroups (all chi-square tests, P O 0.05).
4. Discussion
This study examined the comparability of five at-work
productivity measures from a fundamentally important yet
often overlooked perspective of sensibility. According to
patients with arthritis, key sensibility criteria were gener-
ally met by all measuresdfour of which were developed
fairly recently (between 2001 and 2004) with one dating
back to 1997 (EWPS). Positive ratings for
comprehensiveness (70.0e92.8% endorsement) provided
perhaps the clearest indication that these measures (con-
tinue to) have high relevance for present-day workers em-
ployed over diverse occupa tional sectors. This was further
reinforced by our synthesis of patient feedback, which indi-
cated that relatively few felt that certain aspects of their
jobs were not adequately captured by specific measures.
Several interesting patterns of endorsement emerged
across the four sensibility criteria assessed. Some coupling
between ratings for understandability and suitability of re-
sponse options was observed as measures scoring high on
one criterion also tended to be high on the other. The
WLQ-25 notably received slightly lower ratings on these
criteria compared with other measures, although it comfort-
ably exceeded our 70% cut-off on both counts. We suspect
that some of the difficulties with this tool could be linked to
reversals in the orientation of instructions across the mea-
sure’s four subscales (varies between asking about propor-
tion of time with or without difficulty). In fact, from
openeended feedback, some respondents remarked about
the need to reread instructions as they responded to
WLQ-25 items; thus, this ‘flip’ may have been a source
of some confusion [47]. Ratings for comprehensiveness
and appropriateness of length varied the most across the
five measures. These criteria may be linked, in that more
comprehensive measures also tend to be longer. Both the
WALS and SPS-6 received !80% support for these two
criteria, whereas the EWPS, RA-WIS, and WLQ-25 all re-
ceived 80% support, sugges ting that respondents were
generally not deterred by the increased length of what they
also felt were more comprehensive measures.
A main strength of this study is the highly patient-
centered perspective of our evaluation, which is in concert
with the expectations of today’s patient-reported outcome
Table 3. Summary of written feedback from respondents
Measures
n (%) Indicating measure
lacked comprehensiveness
n (%) Indicating measure
lacked understandability Top two most frequent feedback from respondents
a
WALS 73 (30.0) 6 (2.4) Content not specific enough to my job (n 5 15)
b
Does not accommodate for fluctuations over time (n 5 5)
b
SPS-6 65 (26.6) 21 (8.5) Content not specific enough to my job (n 5 22)
b
Focused too much on emotional demands, not enough on physical de-
mands (n 5 2)
b
EWPS 45 (18.1) 6 (2.4) Questions on work problems lacked connection with health (arthritis)
(n 5 20)
b
Content not specific enough to my job (n 5 6)
b
RA-WIS 36 (14.4) 14 (5.7) More response options is needed in addition to ‘‘yes’’ and ‘‘no’’ (n 5 14)
b
Content not specific enough to my job (n 5 2)
b
WLQ-25 18 (7.2) 53 (21.4) Difficulty with flipping of orientation of instruction (proportion of time with
difficulty vs. without difficulty)dneeded to refer back to instruction often
(n 5 30)
c
Content not specific enough to my job (n 5 3)
b
Abbreviations: WALS, Workplace Activity Limitations Scale; SPS, Stanford Presenteeism Scale; EWPS, Endicott Work Productivity Scale; RA-
WIS, Work Instability Scale for Rheumatoid Arthritis; WLQ-25, Work Limitations Questionnaire.
a
Opportunity for feedback was provided for respondents who did not endorse either the comprehensiveness or understandability of specific at-
work productivity measures.
b
Feedback pertains to the comprehensiveness of the measures.
c
Feedback pertains to the understandability of the measures.
553K. Tang et al. / Journal of Clinical Epidemiology 66 (2013) 546e556
Downloaded from ClinicalKey.com at University of British Columbia April 19, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.
measures [25]. Although the concept of sensibility is not
new, most current approaches continue to remain heavily
reliant on subjective appraisals (e.g., use of checklists com-
pleted by individual researchers/clinicians). A group-level
quantitative evaluation offers an alternative approach that
may be less susceptible to individual rater’s biases. Our de-
sign has some limitations: subject recruitment was conduct-
ed by convenient sampling; the order of the measures
introduced in our survey was not randomized, although
no systematic ‘order’ effect on sensibility performance
was evident in our data except slightly higher rates of item
nonresponse (5.8e6.0%) for the WLQ-25
(last measure fielded) compared with other measures
(0.3e2.4%); also, it should be considered that our final ap-
praisal question enabled only one best measure to be se-
lected, although it is quite possible that respondents may
have seen little difference among the measures being com-
pared. Opportunities to expand sensibility evaluations could
be considered in the future. Although we asked about com-
prehensiveness, our approach did not ‘penalize’ measures
for having redundant or irrelevant items. Also, our opene
ended questions were phrased in a way that asked res-
pondents to offer only criticisms; thus, future studies may
consider offering respondents the opportunity to remark
on the perceived strengths of the measures as well. Further-
more, we had not asked whether the recommended scoring
procedure for the respective scales appear sensible to the
respondents (e.g., equal weighting of all items).
In conclusion, although sensibility crite ria were general
met by all the five at-work productivity measures, it was also
evident that among the measures compared, no single tool
had emerged as being clear ‘superior. In fact, unique
strengths and limitations of the competing measures across
the different sensibility criteria were revealed. A moderate
overall preference for the WALS and WLQ-25 was
indicated by our patient sample, whereas the SPS-6 was
clearly least preferred. To identify appropriate patient-
reported outcome measure(s) for a given need, sensibility
attributes of candidate instruments are worthy of consider-
ation alongside psychometric evidence. Currently, evidence
supporting the reliability, validity, and responsiveness of
the WALS [17,28,35,36], RA-WIS [28,48,49], and WLQ-
25 [28,42,43] in arthritis is more readily available than that
of the EWPS [28] and SPS-6 [28]. Moreover, among these
studies, a recent head-to-head comparison of these same five
measures in arthritis found that the WALS, RA-WIS, and to
a lesser extent the WLQ-25 were superior to the EWPS and
SPS-6 in terms of overall psychometric performance [28].
As such, from an evidence-based perspective, the former
three presenteeism measures may be regarded as more justifi-
able options for use in this population than the latter two
scales at the present time . That said, one should also be mind-
ful of the fact that not all currently available presenteeism
measures have been included in our comparison. Finally,
we propose several areas of further research to advance our
understanding on the sensibility of work outcome measures:
(1) qualitative research (e.g., cognitive testing) to explore the
rationale behind sensibility ratings provided by individual re-
spondents and preferences for specific measures over others,
(2) additional head-to-head comparisons of measures in other
health populations to elucidate the generaliza bility of our
current findings, and finally (3) periodic sensibility reap-
praisals as labor markets evolve and new instruments emerge
over time.
Acknowledgments
The authors acknowledge the participa ting institutions for
this study: the Mount Sinai Hospital (Toronto, Ontario, Can-
ada), Martin Family Centre for Arthritis Care and Research at
St. Michael’s Hospital (Toronto, Ontar io, Canada), and Mary
Pack Arthritis Program (Vancouver, British Columbia,
Canada). The authors also thank the Institute for Work &
Health and Arthritis Community Research & Evaluation Unit
for providing in-kind support.
Table 4. Patient preference (% endorsement) on the ‘‘best’’ measure for capturing the impact of arthritis on work
Measures WALS SPS-6 EWPS RA-WIS WLQ-25
Full sample (N 5 233) 32.6 (1) 2.6 (5) 16.3 (4) 18.5 (3) 30.0 (2)
Stratified by arthritis type
RA (n 5 112) 30.4 (1) 1.8 (5) 19.6 (4) 22.3 (3) 25.9 (2)
OA (n 5 121) 34.7 (1) 3.3 (5) 13.2 (4) 14.9 (3) 33.9 (2)
Stratified by mode of administration
Paper (n 5 120) 28.3 (2) 1.7 (5) 16.7 (4) 17.5 (3) 35.8 (1)
Touch screen (n 5 113) 37.2 (1) 3.5 (5) 15.9 (4) 19.5 (3) 23.9 (2)
Stratified by broad occupational categories
Business, finance, and administration (n 5 92) 38.0 (1) 1.1 (5) 16.3 (3) 14.1 (4) 30.4 (2)
Health, science, art, and sports (n 5 78) 29.5 (2) 5.1 (5) 16.7 (4) 18.0 (3) 30.8 (1)
Sales and services (n 5 37) 27.0 (2) 0.0 (5) 18.9 (4) 29.7 (1) 24.3 (3)
Trades, transport, and equipment operators (n 5 16) 43.8 (1) 6.3 (4) 12.5 (3) 0.0 (5) 37.5 (2)
Abbreviations: WALS, Workplace Activity Limitations Scale; SPS, Stanford Presenteeism Scale; EWPS, Endicott Work Productivity Scale; RA-
WIS, Work Instability Scale for Rheumatoid Arthritis; WLQ-25, Work Limitations Questionnaire; RA, rheumatoid arthritis; OA, osteoarthritis.
Comparative ranking of the five measures in parentheses.
No statistical differences were found in levels of patient endorsement for arthritis type, method of survey administration, or occupational sub-
groups (all chi-square tests of proportion, P O 0.05).
554 K. Tang et al. / Journal of Clinical Epidemiology 66 (2013) 546e556
Downloaded from ClinicalKey.com at University of British Columbia April 19, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.
In addition to the list of authors, investigators of the Ca-
nadian Arthritis Network Work Productivity project also in-
clude Dr. Aslam H. Anis, PhD, and Dr. Elizabeth M.
Badley, PhD. We thank them for their contributions and
support to this work.
Finally, the authors acknowledge individuals who have
made contributions to the overall Canadian Arthritis Net-
work Work Productivity project beyond the present study:
Xingshan Cao (Data Analyst), Paul Clarke (Research Coor-
dinator), Timea Donka (Research Assistant), Rebecca Dub
e
(Research Assistant), Katherine Edwards (Research Assis-
tant), Taucha Inrig (Research Assistant), Carol Kennedy
(Research Assistant), Jessica Lee (Research Coordinator),
Xin Li (Postdoctoral Fellow), Samra Mian (Research Assis-
tant), Ludmila Mironyuk (Research Coordinator), Anusha
Raj (Research Associate), Pam Rogers (Research Coordi-
nator), Rebeka Sujic (Research Coordinator), Debbie Sut-
ton (Data Analyst), Ada Todd (Research Coordinator),
Dwayne Van Eerd (Research Coordinator), Rebecca Wick-
ett (Research Coordinator), and Jessica Widdifield (Re-
search Coordinator), and Wei Zhang (Graduate Student).
References
[1] Sokka T, Kautiainen H, Pincus T, Verstappen SM, Aggarwal A,
Alten R, et al. Work disability remains a major problem in rheuma-
toid arthritis in the 2000s: data from 32 countries in the QUEST-
RA study. Arthritis Res Ther 2010;12(2):R42.
[2] Allaire S, Wolfe F, Niu J, LaValley MP. Contemporary prevalence
and incidence of work disability associated with rheumatoid arthritis
in the US. Arthritis Rheum 2008;59:474e80.
[3] Backman CL. Employment and work disability in rheumatoid arthri-
tis. Curr Opin Rheumatol 2004;16(2):148e52.
[4] Pincus T, Mitchell JM, Burkhauser RV. Substantial work disability
and earnings losses in individuals less than age 65 with osteoarthritis:
comparisons with rheumatoid arthritis. J Clin Epidemiol 1989;42:
449e57.
[5] Escorpizo R, Bombardier C, Boonen A, Hazes JM, Lacaille D,
Strand V, et al. Worker productivity outcome measures in arthritis.
J Rheumatol 2007;34:1372e80.
[6] Loeppke R, Hymel PA, Lofland JH, Pizzi LT, Konicki DL,
Anstadt GW, et al. Health-related workplace productivity measure-
ment: general and migraine-specific recommendations from the
ACOEM Expert Panel. J Occup Environ Med 2003;45(4):349e59.
[7] Lofland JH, Pizzi L, Frick KD. A review of health-related workplace
productivity loss instruments. Pharmacoeconomics 2004;22(3):
165e84.
[8] Collins JJ, Baase CM, Sharda CE, Ozminkowski RJ, Nicholson S,
Billotti GM, et al. The assessment of chronic health conditions on
work performance, absence, and total economic impact for em-
ployers. J Occup Environ Med 2005;47(6):547e57.
[9] Li X, Gignac MA, Anis AH. The indirect costs of arthritis resulting
from unemployment, reduced performance, and occupational
changes while at work. Med Care 2006;44:304e10.
[10] Anis A, Zhang W, Emery P, Sun H, Singh A, Freundlich B, et al. The
effect of etanercept on work productivity in patients with early active
rheumatoid arthritis: results from the COMET study. Rheumatology
2009;48(10):1283e9.
[11] Bejarano V, Quinn M, Conaghan PG, Reece R, Keenan AM,
Walker D, et al. Effect of the early use of the anti-tumor necrosis fac-
tor adalimumab on the prevention of job loss in patients with early
rheumatoid arthritis. Arthritis Rheum 2008;59:1467e74.
[12] Kavanaugh A, Smolen JS, Emery P, Purcaru O, Keystone E,
Richard L, et al. Effect of certolizumab pegol with methotrexate on
home and work place productivity and social activities in patients
with active rheumatoid arthritis. [Erratum appears in Arthritis
Rheum. 2010 Oct;62(10):1514]. Arthritis Rheum 2009;61:
1592e600.
[13] Beaton D, Bombardier C, Escorpizo R, Zhang W, Lacaille D,
Boonen A, et al. Measuring worker productivity: frameworks and
measures. J Rheumatol 2009;36:2100e9.
[14] Tang K, Escorpizo R, Beaton DE, Bombardier C, Lacaille D,
Zhang W, et al. Measuring the impact of arthritis on worker produc-
tivity: perspectives, methodologic issues, and contextual factors.
J Rheumatol 2011;38:1776e90.
[15] Tang K, Beaton DE, Boonen A, Gignac MAM, Bombardier C. Mea-
sures of work disability and productivity: Rheumatoid Arthritis Spe-
cific Work Productivity Survey (WPS-RA), Workplace Activity
Limitations Scale (WALS), Work Instability Scale for Rheumatoid
Arthritis (RA-WIS), Work Limitations Questionnaire (WLQ), and
Work Productivity and Activity Impairment Questionnaire (WPAI).
Arthritis Care Res (Hoboken) 2011;63(Suppl 11):S337e49.
[16] Lerner D, Amick BC 3rd, Rogers WH, Malspeis S, Bungay K,
Cynn D. The Work Limitations Questionnaire. Med Care 2001;39:
72e85.
[17] Gignac MA, Badley EM, Lacaille D, Cott CC, Adam P, Anis AH.
Managing arthritis and employment: making arthritis-related work
changes as a means of adaptation. Arthritis Rheum 2004;51:909e16.
[18] Gilworth G, Chamberlain MA, Harvey A, Woodhouse A, Smith J,
Smyth MG, et al. Development of a work instability scale for rheu-
matoid arthritis. Arthritis Rheum 2003;49:349e54.
[19] Feinstein AR. The theory and evaluation of sensibility. In:
Feinstein AR, editor. Clinimetrics. New Haven, MA: Y.U. Press;
1987:141e66.
[20] Auger C, Demers L, Swaine B. Making sense of pragmatic criteria
for the selection of geriatric rehabilitation measurement tools. Arch
Gerontol Geriatr 2006;43(1):65e83.
[21] Buchbinder R, Goel V, Bombardier C, Hogg-Johnson S. Classification
systems of soft tissue disorders of the neck and upper limb: do they sat-
isfy methodological guidelines? J Clin Epidemiol 1996;49:141e9.
[22]
Terwee CB, Mokkink LB, van Poppel MN, Chinapaw MJ, van MW,
de Vet HC. Qualitative attributes and measurement properties of
physical activity questionnaires: a checklist. Sports Med 2010;
40(7):525e37.
[23] Beaton DE, Boers M, Tugwell P. Assessment of health outcomes. In:
Firestein GS, editor. Kelley’s textbook of rheumatology. 8th ed. Phil-
adelphia, PA: Saunders; 2009:463e74.
[24] Boers M, Brooks P, Strand CV, Tugwell P. The OMERACT filter for
outcome measures in rheumatology. J Rheumatol 1998;25:198e9.
[25] U.S.Department o f Health an d Human Services FDA Center for
Drug Evaluation and Research, U.S.Departm ent of Health and Hu-
man Services FDA C enter for Biologics Evaluation and Research ,
U.S.Department o f Health and Human Services FDA Center for De-
vices and Radiological Health. Guidance for industry: patient-
reported outcome mea sures: use in medical product development
to support labeling claims: draft guidance. He alth Qual Life Out-
comes 20 06;4:79.
[26] Prasad M, Wahlqvist P, Shikiar R, Shih YC. A review of self-report
instruments measuring health-related work productivity: a patient-
reported outcomes perspective. Pharmacoeconomics 2004;22(4):
225e44.
[27] Tang K, Pitts S, Solway S, Beat on D. Comparison of the psycho-
metric properties of four at-work disability measures in workers
with shoulder or elbow disorders. J Occu p Rehab il 2009;19(2):
142e 54.
[28] Beaton DE, Tang K, Gignac MA, Lacaille D, Badley EM, Anis AH,
et al. Reliability, validity, and responsiveness of five at-work produc-
tivity measures in patients with rheumatoid arthritis or osteoarthritis.
Arthritis Care Res (Hoboken) 2010;62(1):28e37.
555K. Tang et al. / Journal of Clinical Epidemiology 66 (2013) 546e556
Downloaded from ClinicalKey.com at University of British Columbia April 19, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.
[29] Brouwer WB, Koopmanschap MA, Rutten FF. Productivity losses
without absence: measurement validation and empirical evidence.
Health Policy 1999;48(1):13e27.
[30] Lavigne JE, Phelps CE, Mushlin A, Lednar WM. Reductions in indi-
vidual work productivity associated with type 2 diabetes mellitus.
Pharmacoeconomics 2003;21(15):1123e34.
[31] Meerding WJ, W IJ, Koopmanschap MA, Severens JL, Burdorf A.
Health problems lead to considerable productivity loss at work
among workers with high physical load jobs. J Clin Epidemiol
2005;58:517e23.
[32] Ozminkowski RJ, Goetzel RZ, Chang S, Long S. The application of
two health and productivity instruments at a large employer. J Occup
Environ Med 2004;46(7):635e48.
[33] Sanderson K, Tilse E, Nicholson J, Oldenburg B, Graves N. Which
presenteeism measures are more sensitive to depression and anxiety?
J Affect Disord 2007;101(1e3):65e74.
[34] Turpin RS, Ozminkowski RJ, Sharda CE, Collins JJ, Berger ML,
Billotti GM, et al. Reliability and validity of the Stanford Presentee-
ism Scale. J Occup Environ Med 2004;46(11):1123e33.
[35] Gignac MA. Arthritis and employment: an examination of behavioral
coping efforts to manage workplace activity limitations. Arthritis
Rheum 2005;53:328e36.
[36] Gi gnac MA, Cao X, Tang K, Beaton DE. Examination of
arthrit is-related work place activity limitations and intermittent
disability over four-and-a-half years and its relationship to job
modifications an d outcomes. Arthritis Care Res (Hoboke n)
2011;63(7):953e62.
[37] Koopman C, Pelletier KR, Murray JF, Sharda CE, Berger ML,
Turpin RS, et al. Stanford presenteeism scale: health status
and em ployee productivity. J Occup Environ Me d 2002;44(1):
14e20.
[38] Lynch W, Riedel J. Measuring employee productivity: a guide to self-
assessment tools. Scottsdale, AZ: Institute for Health & Productivity
Management & William Mercer; 2001.
[39] Endicott J, Nee J. Endicott Work Productivity Scale (EWPS): a new
measure to assess treatment effects. Psychopharmacol Bull 1997;
33(1):13e6.
[40] Tang K, Beaton DE, Gignac MA, Lacaille D, Zhang W,
Bombardier C. The Work Instability Scale for rheumatoid arthritis
predicts arthritis-related work transitions within 12 months. Arthritis
Care Res (Hoboken) 2010;62(11):1578e87.
[41] Amick BC 3r d, Lerner D, Rogers WH, Rooney T, Katz JN. A re-
view of health-related work outcome measures and their uses,
and recommended measures. Spine (Phila Pa 1976) 2000;25( 24):
3152e60.
[42] Lerner D, Reed JI, Massarotti E, Wester LM, Burke TA. The Work
Limitations Questionnaire’s validity and reliability among patients
with osteoarthritis. J Clin Epidemiol 2002;55:197e208.
[43] Walker N, Michaud K, Wolfe F. Work limitations among working
persons with rheumatoid arthritis: results, reliability, and validity of
the work limitations questionnaire in 836 patients. J Rheumatol
2005;32:1006e12.
[44] Lerner D, Rogers WH, Chang H. Scoring the work limitations ques-
tionnaire (WLQ) and the WLQ index for estimating work productiv-
ity loss. Technical Report 2003.
[45] Rowe BH, Oxman AD. An assessment of the sensibility of a quality-
of-life instrument. Am J Emerg Med 1993;11:374e80.
[46] Human Resources and Skills Development Canada. National
occupational classification. Government of Canada; 2011. Available
at http://www5.hrsdc.gc.ca/NOC/English/NOC/2006/Welcome.aspx.
Accessed May 23, 2011.
[47] Tang K, Beaton DE, Amick BC III, Hogg-Johnson S, Cote P,
Loisel P. Confirmatory factor analysis of the Work Limitations Ques-
tionnaire (WLQ-25) in workers’ compensation claimants with
chronic upper-limb disorders. J Occup Rehabil 2012. [Epub ahead
of print].
[48] Tang K, Beaton DE, Lacaille D, Gignac MA, Zhang W, Anis AH,
et al. The Work Instability Scale for Rheumatoid Arthritis (RA-
WIS): does it work in osteoarthritis? Qual Life Res 2010;19:
1057e68.
[49] Tan g K. Disease-related d ifferen tial item functioning in the w ork
instability scale for rheumatoid arthritis: converging results from
three methods. Arthritis Care Res (Hoboken) 2011;63(8):
1159e69.
556 K. Tang et al. / Journal of Clinical Epidemiology 66 (2013) 546e556
Downloaded from ClinicalKey.com at University of British Columbia April 19, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.
    • "In contrast, predictors for sick leave and at-work productivity loss have not been well researched, although interest in atwork productivity loss has been increasing since 49% of patients with RA have this experience [4]. The importance of at-work productivity loss has been recognized by the OMERACT (Outcome Measures in Rheumatology) initia- tive [7] and recently, measures for at-work productivity loss have been identified and validated8910. Since measures for at-work productivity loss are now available, it is possible to investigate the work functioning of patients with RA. "
    [Show abstract] [Hide abstract] ABSTRACT: The aim of this study was to determine which combination of personal, disease-related and environmental factors is best associated with at-work productivity loss in patients with rheumatoid arthritis (RA), and to determine whether at-work productivity loss is associated with the quality of life for these patients. This study is based on cross-sectional data. Patients completed a questionnaire with personal, disease-related and environmental factors (related to the work environment), and clinical characteristics were obtained from patient medical records. At-work productivity loss was measured with the Work Limitations Questionnaire, and quality of life with the RAND 36. Using linear regression analyses, a multivariate model was built containing the combination of factors best associated with at-work productivity loss. This model was cross-validated internally. We furthermore determined whether at-work productivity loss was associated with quality of life using linear regression analyses. We found that at-work productivity loss was associated with workers who had poorer mental health, more physical role limitations, were ever treated with a biological therapeutic medication, were not satisfied with their work, and had more work instability (R(2) = 0.50 and R(2) following cross-validation was 0.32). We found that at-work productivity loss was negatively associated with health-related quality of life, especially with dimensions of mental health, physical role limitations, and pain. We found that at-work productivity loss was associated with personal, work-related, and clinical factors. Although our study results should be interpreted with caution, they provide insight into patients with RA who are at risk for at-work productivity loss.
    Full-text · Article · May 2015
    • "Four of the questionnaire items were reverse scored so that higher scores indicated greater sensibility. According to criteria proposed by Rowe and Oxman (1993), an instrument is considered sensible if mean scores of 5 are obtained for at least 80% of the questionnaire items and if no questionnaire items receive a mean rating of 3. Similar criteria have previously been used for establishing the sensibility of instruments in clinical domains (Rowe and Oxman, 1993; O'Brien et al., 2013; Tang et al., 2013). Interview transcripts underwent directed content analysis (Hsieh and Shannon, 2005) in which an existing sensibility framework (Rowe and Oxman 1987 ) served to inform the deductive , structured approach to the initial coding process. "
    [Show abstract] [Hide abstract] ABSTRACT: a b s t r a c t Sound application of clinical reasoning (CR) by the physical therapist is critical to achieving optimal patient outcomes. As such, it is important for institutions granting certification in orthopaedic manual physical therapy (OMPT) to ensure that the assessment of CR is sufficiently robust. At present, the dearth of validated instruments to assess CR in OMPT presents a serious challenge to certifying institutions. Moreover, the lack of documentation of the development process for instruments that measure CR pose additional challenges. The purpose of this study is to evaluate the sensibility of a newly developed instrument for assessing written responses to a test of CR in OMPT; a 'pilot' phase that examines instrument feasibility and acceptability. Using a sequential mixed-methods approach, Canadian OMPT examiners were recruited to first review and use the instrument. Participants completed a sensibility questionnaire followed by semi-structured interviews, the latter of which were used to elaborate on questionnaire responses regarding feasibility and acceptability. Eleven examiners completed the questionnaire and interviews. Questionnaire results met previously established sensibility criteria, while interview data revealed participants' (dis)comfort with exerting their own judgment and with the rating scale. Quantitative and qualitative data provided valuable insight regarding content validity and issues related to efficiency in assessing CR competence; all of which will ultimately inform further psychometric testing. While results suggest that the new instrument for assessing clinical reasoning in the Canadian cer-tification context is sensible, future research should explore how rater judgment can be utilized effec-tively and the mental workload associated with appraising clinical reasoning.
    Full-text · Article · Oct 2014
  • [Show abstract] [Hide abstract] ABSTRACT: The objective of the Outcome Measures in Rheumatology (OMERACT) Worker Productivity working group is to identify worker productivity outcome measures that meet the requirements of the OMERACT filter. At the OMERACT 11 Workshop, we focused on the at-work limitations/productivity component of worker productivity (i.e., presenteeism) - an area with diverse conceptualization and instrumentation approaches. Various approaches to quantify at-work limitations/productivity (e.g., single-item global and multi-item measures) were examined, and available evidence pertaining to OMERACT truth, discrimination, and feasibility were presented to conference participants. Four candidate global measures of presenteeism were put forth for a plenary vote to determine whether current evidence meets the OMERACT filter requirements. Presenteeism globals from the Work Productivity and Activity Impairment Questionnaire (72% support) and Rheumatoid Arthritis-specific Work Productivity Survey (71% support) were endorsed by conference participants; however, neither the presenteeism global item from the Health and Work Performance Questionnaire nor the Quantity and Quality method achieved the level of support required for endorsement at the present time. The plenary was also asked whether the central item from the Work Ability Index should also be considered as a candidate measure for potential endorsement in the future. Of participants at the plenary, 70% supported this presenteeism global measure. Progress was also made in other areas through discussions at individual breakout sessions. Topics examined include the merits of various multi-item measures of at-work limitations/productivity, methodological issues related to interpretability of outcome scores, and approaches to appraise and classify contextual factors of worker productivity. Feedback gathered from conference participants will inform the future research agenda of the working group.
    Full-text · Article · Oct 2013
Show more