VALUE IN HEALTH 19 (2016) 544–551

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/jval

Patient-Reported Outcome
Performance and Validation of Chronic Liver Disease
Questionnaire-Hepatitis C Version (CLDQ-HCV) in Clinical Trials
of Patients with Chronic Hepatitis C
Zobair M. Younossi, MD, MPH1,2,*, Maria Stepanova, PhD3, Linda Henry, PhD2,3
1
Department of Medicine, Center for Liver Diseases, Inova Fairfax Hospital, Falls Church, VA, USA; 2Betty and Guy Beatty Center for
Integrated Research, Inova Health System, Falls Church, VA, USA; 3Center for Outcomes Research in Liver Diseases, Washington,
DC, USA

AB STR A CT

Background: The hepatitis C virus (HCV) infection has tremendous
clinical, health-related quality-of-life (HRQOL), and economic burden on
patients and the society. To assess the comprehensive impact of HCV
infection, systematic tracking of HRQOL in patients with HCV infection is
important. Objective: The aim of this study was to systematically validate
an HCV-specific HRQOL instrument, the Chronic Liver Disease
Questionnaire-Hepatitis C Version (CLDQ-HCV), in patients with chronic
HCV infection. Methods: The CLDQ-HCV has 29 items in four domains,
each scored on a Likert scale of 1 –to 7. We used a large cohort of patients
with HCV infection enrolled in clinical trials (N ¼ 4142) to test internal
consistency, validity, and responsiveness, and we used another cohort of
untreated patients with HCV infection (N ¼ 36) to assess test-retest
reliability. Results: The CLDQ-HCV performed well in all the psychometric
assessments. In particular, the Cronbach alphas ranged from 0.84 to 0.94
for the four domains. The item-to-own-dimension correlations exceeded
0.6 for 27 of the 29 items. Of the clinical and demographic parameters, the
presence of cirrhosis and history of psychiatric conditions were discrimi-
nated best by the CLDQ-HCV (all P o 0.0001). The domains’ correlations
with similar domains of the 36-item short form health survey exceeded
0.8. The responsiveness to significant clinical outcomes such as develop-
ing treatment-induced anemia and clearance of HCV infection was
notable (up to 0.70 for anemia and up to þ0.85 for achieving sustained
virologic response; all P o 0.0001). Test-retest reliability showed intraclass
correlations of 0.84 to 0.93 between multiple administrations. Conclu-
sions: The CLDQ-HCV is a fully validated, simple-to-administer HCV-
specific instrument for patients with HCV infection that could be
considered in studies of HCV-infected patients.
Keywords: hepatitis C, HRQOL, measurement tool, patient-reported
outcomes, quality of life.
Copyright & 2016, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.

[9,10]. Until recently, a triple therapy combination of pegylated

Introduction
Hepatitis C virus (HCV) is a systemic virus with both hepatic and
extrahepatic manifestations [1–7]. In terms of liver disease associated
with HCV infection, chronic hepatitis C (CH-C) is considered one of
the most common and potentially devastating causes of liver disease
worldwide. In fact, the clinical, economic, and patient experience
impact of HCV infection has been estimated to cause tremendous
burden on patients, their families, and the society [5,7]. However,
HCV cure, as indicated by achieving sustained virologic response
(SVR) after treatment, is associated with improvement of clinical,
economic, and patient-reported outcomes (PROs) [5,8].
Historically, HCV treatment with interferon-based regimens
was associated with low SVR rates and substantial side effects
interferon-alpha, ribavirin, and first-generation direct-acting anti-
viral (DAA) agents was the standard of care for treating CH-C [11].
Postapproval data for these regimens suggested relatively low
efficacy and unfavorable safety profile. Furthermore, interferon-
containing regimens had a tremendously negative impact on PROs
and patients’ health-related quality of life (HRQOL) during treat-
ment [12–15]. In late 2013, regimens that included second-
generation DAA agents were approved, and were soon followed
by the approval of all-oral regimens that had substantially higher
SVR rates as well as better tolerability and cost-effectiveness
profiles [16–27]. Since then, the improvement in HCV infection
treatment has been clearly shown not only in terms of the
superior efficacy and simplicity of the new regimens but also as

Conflicts of interest: Z. M. Younossi is a consultant to the advisory boards of Abbvie, Intercept, Gilead Sciences, Salix, GSK, BMS, and
Janssen. M. Stepanova and L. Henry have no conflicts of interest with regard to the content of this article.
* Address correspondence to: Zobair M. Younossi, Betty and Guy Beatty Center for Integrated Research, Claude Moore Health Education
and Research Building, 3300 Gallows Road, Falls Church, VA 22042.
E-mail: zobair.younossi@inova.org.
1098-3015$36.00 – see front matter Copyright & 2016, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.jval.2016.02.005
 VALUE IN HEALTH 19 (2016) 544–551
a substantially improved side-effect profile, which has led to
enhanced HRQOL and patient experiences [28–33].
In the context of these new treatment regimens for HCV,
documentation of PROs during treatment and after achieving SVR
has become increasingly important to provide the most complete
assessment of the effect of these treatment regimens on patients and
their well-being. Although both generic and disease-specific HRQOL
instruments have been shown to be useful in clinical research, it is
the disease-specific instruments that are supposed to be more
sensitive with regard to both the aspects of PROs that are most likely
impaired in affected patients and the changes in patients’ well-being
associated with different stages of disease [34,35]. Thus, these
disease-specific instruments are expected to be especially useful for
measuring PROs during clinical trials. Nevertheless, it is important
that both disease-specific and generic instruments undergo psycho-
metric testing for reliability, responsiveness, and validity.
The Chronic Liver Disease Questionnaire-Hepatitis C Version
(CLDQ-HCV) is an instrument designed specifically to assess
HRQOL in patients with HCV infection [36,37]. The aim of this
study was to test the CLDQ-HCV for reliability, responsiveness,
and validity in patients with CH-C.
Methods
The CLDQ-HCV
The CLDQ-HCV is a self-administered questionnaire with 29
items. It was originally developed as a modified version of the
CLDQ to apply specifically to patients with HCV. The details of its
development and preliminary validation have been published
[36–38]. In short, for its design, the original items from the CLDQ
were augmented by adding the items that were expected to be
pertinent to patients with HCV infection. These items were
selected from various sources (generic and liver-specific instru-
ments, interviews, and focus groups with patients with HCV
infection). The initial questionnaire (item selection question-
naire) contained 77 items. An impact scores analysis and a factor
analysis were performed to reduce the number of items and place
the items into different domains. After the item reduction step,
40 items remained, and then, after a principle-component anal-
ysis with the selection of factors with eigenvalues greater than 1,
four factors were selected. Using varimax rotation, redundancies
were eliminated, and a final version of the CLDQ-HCV was
developed with 29 items divided into four nonoverlapping
domains: activity/energy, emotional, worry, and systemic [36].
For the CLDQ-HCV, each item was structured in an open-
ended fashion and scored on a Likert scale ranging from 1 to 7:
the lowest score of 1 (a problem is experienced “All of the time” or
not experienced “None of the time”) to the highest score of 7 (a
problem is experienced “None of the time” or not experienced
“All of the time”). The intermediate scores included “Most of the
time,” “A good bit of the time,” “Some of the time,” “A little of the
time,” and “Hardly any of the time.” The four domains were
calculated as an average of their constituent items, and the total
score was the average of four domains. A 2-week recall period
was suggested for all items [36].
Psychometric Assessment of the CLDQ-HCV
We performed an assessment of the CLDQ-HCV for validation,
responsiveness, internal consistency, and test-retest reliability.
Study population: Validation cohort
For the purpose of validation of the CLDQ-HCV, we used a sample
of participants from a number of phase 3 trials of new DAA drugs
(sofosbuvir or a fixed-dose combination of sofosbuvir and
545
ledipasvir) who completed the CLDQ-HCV at multiple time points
before, during, and after treatment. Patients enrolled were adults
(18 years or older) of all age groups, both sex, all HCV genotypes,
treatment-naive or experienced, with or without compensated
cirrhosis, with or without coinfection with HIV, and enrolled in
North America, Europe, Australia, or New Zealand. Excluded from
all trials were patients with coinfection with hepatitis B virus,
patients with a history of any indication of hepatic decompensa-
tion, and patients with current pregnancy or living with a
pregnant partner. Extensive medical history was collected at
the time of screening for all trials’ participants.
For internal consistency assessment, all time points (before,
during, and after treatment) were used. For assessment of
validity and discriminatory power, only one pretreatment base-
line time point was used. For assessment of responsiveness, on-
treatment and post-treatment time points were selected in
addition to the baseline reference time point.
Study population: Test-retest reliability cohort
For the purpose of retest reliability assessment of the CLDQ-HCV,
we administered the instrument 2 to 4 times to patients receiving
care for their HCV infection in an outpatient clinic. Inclusion
criteria were as follows: patients aged 18 years or older, chronic
HCV infection for at least a year, not receiving any anti-HCV
treatment at the moment or between consecutive CLDQ-HCV
administrations, ability to read and understand English, and
willingness to give an informed consent.
The CLDQ-HCV was self-administered by patients in an electronic
form (a tablet was provided) in a clinic room before they received any
information related to their present health status during two
separate off-treatment visits a few weeks apart. Most of the patients
also completed a paper version of the same questionnaire immedi-
ately after completion of the electronic version.
Statistical Analysis
The standard tests were used to assess test-retest reliability,
internal consistency, validity, and responsiveness of the CLDQ-
HCV. Only records without missing items were used.
Test-retest reliability was assessed by several methods. First,
correlations between two paper-based administrations were
calculated for all individual items and all the summary domains.
Similarly, correlations between two electronic administrations
were also calculated. Next, correlations between paper and
electronic administrations were calculated. In addition, for the
same pair of administrations, the distributions of differences in
scores between multiple administrations were also calculated.
These included the mean differences in the item values, the SDs,
and the maximum absolute values of the differences; the median
differences were also statistically compared with 0 by a non-
parametric test for matched pairs. Finally, intraclass correlation
coefficients (ICCs), which are another standard indicator of
reliability, were calculated for the summary domains. A general
linear model was used for each domain separately. This model
used the subject identification (ID) and the administration ID
(both parameters categorical; up to four administrations per
subject) as two predictors of an outcome; it was expected that P
values for the administration ID parameter would be substan-
tially insignificant so that the outcome (a CLDQ-HCV domain)
would be driven solely by the subject. The ICC was the ratio of
between-subjects variance to the total sample variance.
Internal consistency was assessed by calculating Cronbach
alphas for the CLDQ-HCV domains, and by calculating the item-
to-own-domain correlations for all the CLDQ-HCV items after
adjustment for overlap [39].
Validity was assessed by evaluation of the association of
CLDQ-HCV scores with demographic and clinical parameters.
546 VALUE IN HEALTH 19 (2016) 544–551
Patients of older age, female patients, and those having comor-
bidities or other clinical risk factors were expected to score lower.
The validation parameters used in this study were age, sex, the
presence of compensated cirrhosis, pretreatment anemia (base-
line hemoglobin level of o12 g/dl in females, o13 g/dl in males
[40]), high HCV viral load (46 log 10 IU/ml), elevated alanine
aminotransferase level (above 1.5 sex-based upper limits of the
norm), coinfection with HIV, having a history of unsuccessful
anti-HCV treatment, and having a history of pretreatment
depression or clinically overt fatigue. A series of Wilcoxon non-
parametric tests was used to compare the CLDQ-HV scores
between the patients from the aforementioned categories.
The 36-item short form health survey (SF-36) instrument was
used as a reference “criterion standard” for assessment of validity
of the CLDQ-HCV. Correlations of the CLDQ-HCV domains with
the SF-36 domains were calculated, and correlations with the
most related domains of the SF-36 were expected to be the
highest. In addition, patients with the lowest possible and the
highest possible SF-36 scores were selected, and the proportions
of patients who would also score highest (lowest) in the CLDQ-
HCV were calculated.
Responsiveness was assessed using multiple time points for
the same patients. Two major outcomes, development of severe
anemia during treatment (o10 g/dL) and sustained clearance of
HCV infection, which the CLDQ-HCV was designed to be partic-
ularly responsive to, were studied. These outcomes were cap-
tured from the baseline survey results contrasted to the last day
of anti-HCV treatment survey results in patients who developed
severe treatment-emergent anemia and in patients who achieved
SVR at post-treatment weeks 12 and 24 and compared with 0 by a
Wilcoxon signed rank test.
P values less than 0.05 were considered statistically signifi-
cant. All analyses were performed using SAS 9.3 (SAS Institute,
Cary, NC). The Inova Institutional Review Board approved the
study. All subjects had given informed consent for participation
in the study.
Results
Validity, Discriminatory Power, and Responsiveness
In this part of the study, we included data from 4142 patients who
participated in phase 3 clinical trials of sofosbuvir-based regi-
mens for treatment of chronic HCV infection and had completed
the CLDQ-HCV at baseline. The patients were, on average, 52.3 ⫾
9.8 years old; 67% were men, 83% white, 67% enrolled in the
United States, 19% cirrhotic, and 20% with HIV coinfection. The
CLDQ-HCV was administered at multiple time points before,
during, and after treatment; thus, a total of 22,142 entries were
used in this study (5.3 per patient); nevertheless, 1,514 entries
(o7%) were excluded from the internal consistency analysis
because of missing items.
Table 1 – Internal consistency of the CLDQ-HCV (N ¼ 20,628 entries).
Domain No. of items Cronbach α
Raw Standardized
Activity/energy 6 0.926 0.926
Emotional 9 0.939 0.939
Worry 8 0.906 0.907
Systemic 6 0.838 0.844
CLDQ-HCV, Chronic Liver Disease Questionnaire-Hepatitis C Version.
Sufficiently high internal consistency was found (Table 1). In
particular, the Cronbach alpha coefficient, which is supposed to
reflect the measure of association between the items that
presumably measure the same construct, ranged from 0.84 for
the systemic domain to 0.94 for the emotional domain. Further-
more, in a series of one-item exclusions (adjustment for overlap),
the resulting Cronbach alphas did not change substantially to
any direction, indicating that the items were neither too corre-
lated nor too different from other items belonging to the same
domain. In fact, the greatest change in Cronbach alpha was for
the Q2 item (“o…4 suffered from bodily pain o…4?”) from the
systemic domain (0.038). The item-to-own-dimension correla-
tions were above 0.6 for 27 items; the remaining 2 items were Q29
of the worry domain (“o…4 costs of medications?”) and Q25 of
the systemic domain (“o…4 decreased interest in sex?”).
The histograms of the CLDQ-HCV domain score in Figure 1
show substantial skewness toward the highest possible values. In
fact, except for the systemic domain, the proportions of CLDQ-
HCV entries with domain values of 7 ranged from 43% to 59%. At
the same time, the proportions of entries with values of 1 did not
exceed 0.1%; also, at most, 2% of entries had a value of 2 for a
domain score.
Validity of the CLDQ-HCV and its discriminant function in
relation to demographic and clinical parameters is presented in
Table 2 (see Table 1 and Table 2 in Supplemental Materials found
at http://dx.doi.org/doi:10.1016/j.jval.2016.02.005 for age-based
distribution and for the distribution of discriminating parameters
by CLDQ-HCV quartiles, respectively). The average baseline
CLDQ-HCV scores were as follows: activity/energy ¼ 5.33 ⫾ 1.36,
emotional ¼ 5.42 ⫾ 1.21, worry ¼ 5.49 ⫾ 1.26, systemic ¼ 5.07 ⫾
1.29, total ¼ 5.33 ⫾ 1.13. As expected, men had significantly
higher scores than did women (by þ0.11 to þ0.27; all P o 0.02),
and the youngest group (o45) had the highest scores for the
activity/energy and systemic domains but not for the emotional
and worry domains. The association of lower CLDQ-HCV scores
with the presence of anemia, high HCV viral load, high alanine
aminotransferase level, or coinfection with HIV was moderate at
best (Table 2). However, the association of lower CLDQ-HCV
scores with a history of unsuccessful anti-HCV treatment, the
presence of cirrhosis, history of depression, or clinically overt
fatigue was strong and consistent across all four domains
(Table 2; Fig. 2). In fact, patients with cirrhosis had, on average,
a change of 0.33 to 0.50 in their CLDQ-HCV domain scores (all
P o 0.0001). Similarly, treatment-experienced patients’ scores
were lower by 0.08 to 0.19 (all P o 0.01). In patients with
depression, the average domain score impairment ranged from
0.41 to 0.82 (all P o 0.0001) and a similar difference in patients
with fatigue ranged from 0.43 to 0.96 (Table 2; Fig. 2).
Another round of validity assessment included correlations
with the SF-36 (Table 3). As expected, the highest correlated item
for the activity/energy domain was vitality (R ¼ 0.82) closely
followed by role physical (R ¼ 0.80), and the lowest correlated
item was, again expectedly, mental health (R ¼ 0.65). For the
emotional domain, the highest correlated item was, in turn,
Adjustment for overlap for the domain’s items
Item-to-own-dimension correlations Cronbach α
0.704–0.845 0.905–0.923
0.734–0.829 0.929–0.934
0.543–0.824 0.884–0.909
0.448–0.685 0.806–0.851
 VALUE IN HEALTH 19 (2016) 544–551 547

Fig. 1 – The histograms of the domain and total scores of the CLDQ-HCV (the sample includes all pre-, on-, and post-treatment
time points). CLDQ-HCV, Chronic Liver Disease Questionnaire-Hepatitis C Version; IQR, interquartile range.

mental health summary score (R ¼ 0.81) and the lowest corre-
lated item was physical functioning (R ¼ 0.47). By design, the
worry domain was unique for the CLDQ-HCV and, therefore, had
relatively low correlations with all the SF-36 domains: from 0.39
to 0.56. The systemic domain was the most correlated with the
bodily pain domain of the SF-36 (R ¼ 0.72).
There were 29 patients with perfect SF-36 scores (all domains
equal to 100). Of these, 10 also had perfect CLDQ-HCV scores (the
total score equal to 7). At the same time, there were 4 patients
with the worst possible SF-36 scores (all domains equal to 0), and
none of them scored 1 (the lowest possible value) in the CLDQ-
HCV. The association between extreme values of other related
scores of the CLDQ-HCV and the SF-36 was similar. In particular,
there were 89 patients with the perfect physical scores of the SF-
36; 44 of them also had a perfect activity/energy score, and 38 had
a perfect systemic score. However, there were 41 patients with 0
physical scores of the SF-36, but only 2 of them had the lowest
possible scores in activity/energy or systemic domains of the
CLDQ-HCV. Similarly, out of 93 patients with perfect SF-36 mental
scores, 52, 48, and 49, respectively, also had the highest possible
scores in the emotional, worry, and systemic domains of the
CLDQ-HCV. However, out of 13 patients with 0 SF-36 mental
scores, only 1 also had the lowest possible value in the emotional,
systemic, or worry domain.
Responsiveness of the CLDQ-HCV was assessed as the mag-
nitude of change in the CLDQ-HCV scores from baseline in
patients who developed severe anemia during treatment (pri-
marily because of the use of ribavirin) and in patients who
achieved SVR after treatment cessation (regardless of the treat-
ment regimen used). As shown in Figure 3, the average drop in
the CLDQ-HCV domain scores associated with the development
of severe anemia was from 0.33 to 0.74 (all P o 0.0001). The
only exception was the worry domain, the score for which
increased regardless of the presence of the treatment-emergent
adverse event and was likely driven solely by the fact of receiving
treatment. However, SVR-associated increases in the CLDQ-HCV
domains were between þ0.30 and þ0.70 at post-treatment week
12 and between þ0.39 and þ0.85 by post-treatment week 24
(Fig. 3; all P o 0.0001). As shown, the greatest post-SVR increases
were in the worry domain of the CLDQ-HCV and the lowest were
in the systemic domain.
Test-Retest Reliability
For this part of the study, a total of 36 patients who completed
the CLDQ-HCV at least twice while not receiving treatment were
enrolled. The patients were, on average, 54.8 ⫾ 9.9 years old; 56%
were women, 75% white, and 26% cirrhotic. Patients completed
the paper and the electronic administrations of the CLDQ-HCV.
The total number of entries was 123, or 3.4 per patient.
The correlations between the individual items of the two
paper administrations of the CLDQ-HCV ranged from 0.05 to 0.94
 548 VALUE IN HEALTH 19 (2016) 544–551

Table 2 – Validity and discriminatory power of the CLDQ-HCV (N ¼ 4142 baseline entries; activity/energy ¼ 5.33
⫾ 1.36, emotional ¼ 5.42 ⫾ 1.21, worry ¼ 5.49 ⫾ 1.26, systemic ¼ 5.07 ⫾ 1.29, total ¼ 5.33 ⫾ 1.13).
Item Male Female P Treatment-naive Experienced P

N 2756 1386 2589 1553

Activity/energy 5.42 ⫾ 1.31 5.15 ⫾ 1.44 o0.0001 5.40 ⫾ 1.36 5.21 ⫾ 1.36 o0.0001
Emotional 5.47 ⫾ 1.18 5.32 ⫾ 1.26 0.0011 5.45 ⫾ 1.21 5.37 ⫾ 1.20 0.0073
Worry 5.53 ⫾ 1.24 5.42 ⫾ 1.31 0.0195 5.56 ⫾ 1.26 5.39 ⫾ 1.26 o0.0001
Systemic 5.15 ⫾ 1.24 4.90 ⫾ 1.36 o0.0001 5.13 ⫾ 1.29 4.97 ⫾ 1.28 o0.0001
Total 5.39 ⫾ 1.10 5.20 ⫾ 1.18 o0.0001 5.38 ⫾ 1.13 5.23 ⫾ 1.13 o0.0001

Item Cirrhosis No cirrhosis P Anemia No anemia P

N 796 3338 163 3966

Activity/energy 5.00 ⫾ 1.43 5.41 ⫾ 1.34 o0.0001 5.09 ⫾ 1.52 5.34 ⫾ 1.36 0.0680
Emotional 5.15 ⫾ 1.23 5.48 ⫾ 1.20 o0.0001 5.22 ⫾ 1.35 5.43 ⫾ 1.20 0.1027
Worry 5.09 ⫾ 1.33 5.59 ⫾ 1.23 o0.0001 5.35 ⫾ 1.40 5.50 ⫾ 1.26 0.3081
Systemic 4.76 ⫾ 1.30 5.14 ⫾ 1.27 o0.0001 4.79 ⫾ 1.39 5.08 ⫾ 1.28 0.0101
Total 5.00 ⫾ 1.17 5.40 ⫾ 1.11 o0.0001 5.12 ⫾ 1.27 5.34 ⫾ 1.13 0.0409

Item Elevated ALT level Normal ALT level P HIV No HIV P

N 2171 1971 810 3332

Activity/energy 5.36 ⫾ 1.35 5.30 ⫾ 1.38 0.2292 5.28 ⫾ 1.29 5.34 ⫾ 1.38 0.0172
Emotional 5.42 ⫾ 1.21 5.42 ⫾ 1.20 0.9467 5.40 ⫾ 1.15 5.42 ⫾ 1.22 0.1584
Worry 5.44 ⫾ 1.29 5.55 ⫾ 1.24 0.0102 5.57 ⫾ 1.23 5.47 ⫾ 1.27 0.0368
Systemic 5.08 ⫾ 1.26 5.06 ⫾ 1.32 0.8657 5.00 ⫾ 1.27 5.08 ⫾ 1.29 0.0630
Total 5.32 ⫾ 1.13 5.33 ⫾ 1.13 0.8410 5.31 ⫾ 1.08 5.33 ⫾ 1.14 0.2572

Item Fatigue No fatigue P Depression No depression P

N 490 3652 1112 3030

Activity/energy 4.48 ⫾ 1.44 5.44 ⫾ 1.31 o0.0001 4.73 ⫾ 1.47 5.55 ⫾ 1.25 o0.0001
Emotional 4.94 ⫾ 1.32 5.48 ⫾ 1.18 o0.0001 4.89 ⫾ 1.32 5.61 ⫾ 1.11 o0.0001
Worry 5.11 ⫾ 1.36 5.54 ⫾ 1.24 o0.0001 5.19 ⫾ 1.37 5.60 ⫾ 1.20 o0.0001
Systemic 4.44 ⫾ 1.33 5.15 ⫾ 1.26 o0.0001 4.56 ⫾ 1.34 5.26 ⫾ 1.22 o0.0001
Total 4.74 ⫾ 1.19 5.41 ⫾ 1.10 o0.0001 4.84 ⫾ 1.21 5.50 ⫾ 1.05 o0.0001
ALT, alanine aminotransferase; CLDQ-HCV, Chronic Liver Disease Questionnaire-Hepatitis C Version.

(average R ¼ 0.66) and the correlations for the domain scores
ranged from 0.71 to 0.89 (Table 4). The value of correlation of 0.05
for Q20 (“o…4 concerned about the availability of a liver trans-
plant”) was, in fact, driven by one subject who changed the
response from 7 to 1 in 23 days, and if that subject was excluded,
then that item’s correlation magnitude would increase to 0.50.
The correlations between the electronic administrations of
the CLDQ-HCV were generally higher (from 0.19 to 0.95 for the
individual items with an average R ¼ 0.76, and from 0.77 to 0.95
for the domain scores) (Table 4).
Finally, the correlations between the electronic and paper
administrations were the highest (on average, R ¼ 0.91, with a
range from 0.83 to 0.97 for the individual items and from 0.97 to
0.98 for the domain scores) (Table 4). The same was confirmed by
the relatively high values of ICC: 0.84 to 0.93 for the domain
scores. Furthermore, as expected, no association between the
CLDQ-HCV scores and the administration ID was found to be
significant or close to significance—all P 4 0.25 (Table 4).
The summary of differences in scores between the CLDQ-HCV
administrations is presented in Table 5. For the individual items,
the greatest average difference between paper administrations
was for the Q23 item (“o…4 feeling uncomfortable in social
situations?”): mean 0.59, ranged from 1 to þ6, P ¼ 0.11. That P
value was the lowest among all 29 items, four domain scores, and

Fig. 2 – The distribution of the total CLDQ-HCV score (baseline time point only) by the presence of cirrhosis (A) and depression
(B). CLDQ-HCV, Chronic Liver Disease Questionnaire-Hepatitis C Version.
 VALUE IN HEALTH 19 (2016) 544–551 549

Table 3 – Validity of the CLDQ-HCV: Correlations between the domains of the CLDQ-HCV and the domains of the
SF-36.
SF-36 domain CLDQ-HCV domain, R (95% CI)

Activity/energy Emotional Worry Systemic Total

Physical functioning 0.66 (0.64–0.67) 0.47 (0.45–0.49) 0.39 (0.37–0.42) 0.57 (0.55–0.59) 0.59 (0.57–0.61)
Role physical 0.80 (0.78–0.81) 0.62 (0.60–0.64) 0.52 (0.49–0.54) 0.64 (0.62–0.66) 0.73 (0.72–0.74)
Bodily pain 0.69 (0.67–0.70) 0.52 (0.50–0.54) 0.43 (0.41–0.46) 0.72 (0.70–0.73) 0.67 (0.65–0.69)
General health 0.67 (0.66–0.69) 0.61 (0.59–0.63) 0.55 (0.53–0.57) 0.58 (0.56–0.60) 0.69 (0.67–0.70)
Vitality 0.82 (0.81–0.83) 0.70 (0.69–0.72) 0.52 (0.50–0.55) 0.67 (0.65–0.69) 0.77 (0.76–0.78)
Social functioning 0.74 (0.73–0.76) 0.73 (0.72–0.75) 0.53 (0.51–0.56) 0.61 (0.59–0.62) 0.74 (0.73–0.75)
Role emotional 0.68 (0.67–0.70) 0.70 (0.68–0.72) 0.53 (0.50–0.55) 0.57 (0.55–0.59) 0.70 (0.69–0.72)
Mental health 0.65 (0.63–0.66) 0.80 (0.79–0.81) 0.54 (0.52–0.56) 0.57 (0.55–0.59) 0.72 (0.71–0.74)
Physical summary 0.74 (0.72–0.75) 0.47 (0.45–0.50) 0.44 (0.42–0.47) 0.67 (0.65–0.68) 0.66 (0.64–0.68)
Mental summary 0.70 (0.68–0.72) 0.81 (0.80–0.82) 0.56 (0.54–0.58) 0.57 (0.55–0.59) 0.75 (0.73–0.76)
Note. All P o 0.0001. Numbers in bold indicate highly correlated domains and summary scores of the SF-36 for the CLDQ-HCV domains.
CI, confidence interval; CLDQ-HCV, Chronic Liver Disease Questionnaire-Hepatitis C Version; SF-36, 36-item short form health survey.

the total score, so neither item showed a statistically significant
difference between two rounds of paper-based administrations of
the CLDQ-HCV (Table 5). Similarly, the greatest difference
between two electronic administrations of the CLDQ-HCV was,
on average, 0.66 (from 5 to þ4; P ¼ 0.03) for the Q29 item (“o…
4 concerned about the cost of your medication?”). A potentially
significant P value of 0.03 was also found for the Q2 item (“o…4
suffered from bodily pain o…4”), whereas the remaining 27
items and all domain scores did not change significantly (all
P 4 0.09) (Table 5). Finally, significant differences between the
paper and electronic administrations were found for Q25, Q27,
and the systemic domain score (all P o 0.05) (Table 5).
Discussion
We have previously developed and fully validated a disease-
specific HRQOL instrument for patients with chronic liver disease
(the CLDQ) using a comprehensive methodological framework
that had been established for the development of these types of
instruments [37]. Because patients with HCV may have different
issues impacting their HRQOL, we subsequently developed the
HCV-specific version of the CLDQ [36,38], that is, the CLDQ-HCV.
For the development of the CLDQ-HCV, we used the original
CLDQ tool as well as various other sources including input from
Fig. 3 – The anemia- and SVR-associated changes in the
CLDQ-HCV scores. All P o 0.001. CLDQ-HCV, Chronic Liver
Disease Questionnaire-Hepatitis C Version; SVR, sustained
virologic response.
patients with HCV to ensure that all items important to their
HRQOL were captured [36].
This study provides strong evidence for systematic validation
of the CLDQ-HCV. We used data from large clinical trials to
provide the most extensive evidence for validity and internal
consistency of the CLDQ-HCV. In particular, we have shown that
patients with HCV infection with advanced liver disease (cirrho-
sis) have the most impaired HRQOL. In fact, having cirrhosis was
one of the best-discriminated outcomes out of a number of
studied demographic and clinical parameters, which is exactly
what would be expected for a disease-specific instrument. This is
also consistent with previously published data [41–43]. Other
clinical and demographic factors associated with HRQOL impair-
ment captured by the CLDQ-HCV are also consistent with pre-
viously published literature and make sense from the clinical
point of view [30,33,44–54]. All these data support the validity of
the CLDQ-HCV.
Our data assessing HRQOL during treatment of HCV provides
evidence for the responsiveness of the CLDQ-HCV. In fact, for
patients with HCV who had a clinically important change in their
health status as measured by the development of treatment-
induced anemia, the CLDQ-HCV was able to capture the respec-
tive change in HRQOL. Furthermore, for patients who had their
HCV infection cured (achieving an SVR), a significant improve-
ment in their CLDQ-HCV scores is demonstrated. These two
examples suggest the responsiveness of the CLDQ-HCV in cap-
turing clinically important changes in the health status of
patients with HCV infection and the impact on their HRQOL.
Our internal consistency testing and test-retesting of the
CLDQ-HCV provide sufficient evidence for the reliability of this
instrument. High correlations between presumably related
domains of the CLDQ-HCV and the widely used and extensively
validated SF-36 provide additional confirmation for the validity of
the CLDQ-HCV. In fact, when matched against the SF-36, the
lowest correlated domain of the CLDQ-HCV is the worry domain,
which is expected because this domain is unique for the CLDQ-
HCV and may not be well captured by generic instruments such
as the SF-36.
Finally, our data suggest high correlation between the elec-
tronic version and the paper version of the CLDQ-HCV. Given the
ease of use of ePRO instruments, the e-CLDQ-HCV version may be
preferable to some investigators [51].
These data have important implications for the field of HCV.
As the knowledge about the impact of HCV expands, there is
better understanding of its clinical significance as not only an
550 VALUE IN HEALTH 19 (2016) 544–551

Table 4 – Test-retest reliability of the CLDQ-HCV: Correlations between multiple paper administration, multiple
electronic administrations, and between paper and electronic administrations.
Domain Paper administrations (N ¼ 22) Electronic administrations (N ¼ 29)

R (95% CI) P R (95% CI) P

Activity/energy 0.78 (0.53–0.90) o0.0001 0.95 (0.89–0.97) o0.0001

Emotional 0.80 (0.58–0.91) o0.0001 0.89 (0.78–0.95) o0.0001
Worry 0.71 (0.40–0.87) 0.0002 0.77 (0.56–0.89) o0.0001
Systemic 0.89 (0.75–0.95) o0.0001 0.93 (0.86–0.97) o0.0001
Total 0.80 (0.57–0.91) o0.0001 0.94 (0.87–0.97) o0.0001

Domain Paper and electronic administrations (N ¼ 36 patients; n ¼ 123 entries)

R (95% CI) P ICC& (95% CI) Padm

Activity/energy 0.97 (0.94–0.98) o0.0001 0.91 (0.86–0.95) 0.25

Emotional 0.97 (0.94–0.98) o0.0001 0.90 (0.84–0.94) 0.59
Worry 0.97 (0.94–0.98) o0.0001 0.84 (0.75–0.91) 0.72
Systemic 0.98 (0.96–0.99) o0.0001 0.93 (0.89–0.96) 0.32
Total 0.98 (0.97–0.99) o0.0001 0.92 (0.87–0.95) 0.73
CI, confidence interval; GLM, general linear model; CLDQ-HCV, Chronic Liver Disease Questionnaire-Hepatitis C Version; ICC&, intraclass
correlation by a two-factor GLM; ID, identification; Padm, P value for the administration ID parameter in a GLM (supposed to be 440.05).

important cause of liver disease but also an important etiologic
agent for a number of extrahepatic manifestations [1–4]. In this
context, it is critical to comprehensively capture both clinical and
HRQOL outcomes of HCV with valid, reliable, and responsive
HRQOL disease-specific instruments. Our data provide strong
evidence that the CLDQ-HCV is such an instrument. Therefore,
the CLDQ-HCV tool could be used in clinical trials involving
patients with chronic HCV infection.
It is important to note that the outcomes obtained in the
clinical trial setting are not always replicated in the real-world
setting of clinical practice. These potential differences between
efficacy and effectiveness apply not only to clinical outcomes
such as SVR but also to PROs such as the CLDQ-HCV. Although
the validity of the CLDQ-HCV is now established in the clinical
trial setting, its performance in the real-world setting of clinical
practices must be established.
Conclusions
We provide extensive evidence for the validity of the CLDQ-HCV
as an important instrument to measure the disease-specific
aspect of HRQOL in patients with HCV infection. Addition of the
CLDQ-HCV to other clinical parameters of clinical trials will
ensure that we measure not only clinically important variables
but also issues that are important to patients with HCV infection
about their experiences with their disease and its treatment.

Table 5 – Test-retest reliability of the CLDQ-HCV:

The distribution of differences between multiple
paper and electronic administrations. Supplementary Materials
Domain Mean ⫾ SD Min. Max. P Supplemental material accompanying this article can be found in
the online version as a hyperlink at http://dx.doi.org/doi:10.1016/
Two paper administrations at different time points (N ¼ 22) j.jval.2016.02.005 or, if a hard copy of article, at www.valuein
Activity/energy 0.30 ⫾ 0.93 0.67 3.17 0.14 healthjournal.com/issues (select volume, issue, and article).
Emotional 0.06 ⫾ 0.72 1.11 2.00 0.91
Worry 0.13 ⫾ 0.92 1.25 3.13 0.96
Systemic 0.11 ⫾ 0.74 1.17 2.00 0.62
Total 0.15 ⫾ 0.75 0.66 2.57 0.65
Two electronic administrations at different time points (N ¼ 29) R EF E R EN C ES
Activity/energy 0.07 ⫾ 0.53 0.67 1.33 0.81
Emotional 0.13 ⫾ 0.60 1.22 1.33 0.19
Worry 0.09 ⫾ 0.74 1.88 1.38 0.53
[1] Rodger AJ, Roberts S, Lanigan A, et al. Assessment of long-term
Systemic 0.10 ⫾ 0.56 1.00 1.33 0.33 outcomes of community-acquired hepatitis C infection in a cohort with
Total 0.01 ⫾ 0.44 0.61 1.14 0.50 sera stored from 1971–1975. Hepatology 2000;32:582–7.
Paper and electronic administration means (N ¼ 36) [2] Zarski JP, McHutchison J, Bronowicki JP, et al. Rate of natural disease
progression in patients with chronic hepatitis C. J Hepatol 2003;38:307–14.
Activity/energy 0.06 ⫾ 0.40 0.83 1.17 0.69
[3] Hajarizadeh B, Grebely J, Dore GJ. Epidemiology and natural history of
Emotional 0.03 ⫾ 0.33 0.50 0.89 0.84 HCV infection. Nat Rev Gastroenterol Hepatol 2013;10:553–62.
Worry 0.08 ⫾ 0.34 1.06 0.50 0.21 [4] Leone N, Rizzetto M. Natural history of hepatitis C virus infection: from
Systemic 0.09 ⫾ 0.27 1.08 0.67 0.0293 chronic hepatitis to cirrhosis, to hepatocellular carcinoma. Minerva
Gastroenterol Dietol 2005;51:31–46.
Total 0.02 ⫾ 0.22 0.59 0.50 0.44
[5] Younossi ZM, Kanwal F, Saab S, et al. The impact of hepatitis C burden:
CLDQ-HCV, Chronic Liver Disease Questionnaire-Hepatitis C Ver- an evidence-based approach. Aliment Pharmacol Ther 2014;39:518–31.
sion; Max., maximum; Min., minimum. [6] El-Serag HB, Hampel H, Yeh C, Rabeneck L. Extrahepatic manifestations
of hepatitis C among United States male veterans. Hepatology
2002;36:1439–45.
 VALUE IN HEALTH 19 (2016) 544–551
[7] Louie KL, Forssen UM, Mundy LM, Pimenta JM. The high comorbidity
burden of the hepatitis C virus infected population in the United States.
BMC Infect Dis 2012;12:86.
[8] John-Baptiste AA, Tomlinson G, Hsu PC, et al. Sustained responders
have better quality of life and productivity compared with treatment
failures long after antiviral therapy for hepatitis C. Am J Gastroenterol
2009;104:2439–48.
[9] Bronowicki JP, Ouzan D, Asselah T, et al. Effect of ribavirin in genotype
1 patients with hepatitis C responding to pegylated interferon alfa-2a
plus ribavirin. Gastroenterology 2006;131:1040–8.
[10] Häuser W, Holtmann G, Grandt D. Determinants of health-related
quality of life in patients with chronic liver diseases. Clin Gastroenterol
Hepatol 2004;2:157–63.
[11] Laouénan C, Guedj J, Peytavin G, et al. A model-based illustrative
exploratory approach to optimize the dosing of peg-IFN/RBV in
cirrhotic hepatitis C patients treated with triple therapy. CPT
Pharmacometrics Syst Pharmacol 2015;4:e00008.
[12] Matsushita H, Ikeda F, Iwasaki Y, et al. Assessment of health-related
quality of life and how it predicts the outcome of pegylated interferon
and ribavirin therapy for chronic hepatitis C. J Gastroenterol Hepatol
2014;29:337–43.
[13] Hollander A, Foster GR, Weiland O. Health-related quality of life before,
during and after combination therapy with interferon and ribavirin in
unselected Swedish patients with chronic hepatitis C. Scand J
Gastroenterol 2006;41:577–85.
[14] Dan AA, Martin LM, Crone C, et al. Depression, anemia and
health-related quality of life in chronic hepatitis C. J Hepatol
2006;44:491–8.
[15] Dan AA, Crone C, Wise TN, et al. Anger experiences among hepatitis C
patients: relationship to depressive symptoms and health-related
quality of life. Psychosomatics 2007;48:223–9.
[16] Nyalakonda H, Utay NS. A new era of therapy for hepatitis C virus
infection. Curr Opin Infect Dis 2015;28:471–8.
[17] Lawitz E, Mangia A, Wyles D, et al. Sofosbuvir for previously untreated
chronic hepatitis C infection. N Engl J Med 2013;368:1878–87.
[18] Poordad F, Lawitz E, Kowdley KV, et al. Exploratory study of oral
combination antiviral therapy for hepatitis C. N Engl J Med 2013;368:45–53.
[19] Jacobson IM, Gordon SC, Kowdley KV, et al. POSITRON Study; FUSION
Study. . Sofosbuvir for hepatitis C genotype 2 or 3 in patients without
treatment options. N Engl J Med 2013;368:1867–77.
[20] Afdhal N, Zeuzem S, Kwo P, et al. ION-1 Investigators. Ledipasvir and
sofosbuvir for untreated HCV genotype 1 infection. N Engl J Med
2014;370:1889–98.
[21] Afdhal N, Reddy KR, Nelson DR, et al. ION-2 Investigators. . Ledipasvir
and sofosbuvir for previously treated HCV genotype 1 infection. N Engl J
Med 2014;370:1483–93.
[22] Zeuzem S, Dusheiko GM, Salupere R, et al. VALENCE Investigators. .
Sofosbuvir and ribavirin in HCV genotypes 2 and 3. N Engl J Med
2014;370:1993–2001.
[23] Kowdley KV, Gordon SC, Reddy KR, et al. ION-3 Investigators.
Ledipasvir and sofosbuvir for 8 or 12 weeks for chronic HCV without
cirrhosis. N Engl J Med 2014;370:1879–88.
[24] Ferenci P, Bernstein D, Lalezari J, et al. ABT-450/r-ombitasvir and dasabuvir
with or without ribavirin for HCV. N Engl J Med 2014;370:1983–92.
[25] Kowdley KV, Lawitz E, Poordad F, et al. Safety and efficacy of interferon-
free regimens of ABT-450/r, ABT-267, ABT-333 ⫾ ribavirin in patients
with chronic HCV GT1 infection: results from the Aviator study.
Presented at: The 48th Annual Meeting of the European Association for
the Study of the Liver, April 24–28, 2013, Amsterdam, The Netherlands.
[26] Lawitz E, Sulkowski MS, Ghalib R, et al. Simeprevir plus sofosbuvir, with
or without ribavirin, to treat chronic infection with hepatitis C virus
genotype 1 in non-responders to pegylated interferon and ribavirin and
treatment-naive patients: the COSMOS randomised study. Lancet
2014;384:1756–65.
[27] Younossi ZM, Singer ME, Mir HM, et al. Impact of interferon free
regimens on clinical and cost outcomes for chronic hepatitis C
genotype 1 patients. J Hepatol 2014;60:530–7.
[28] Younossi ZM, Stepanova M, Nader F, et al. Patient-reported outcomes
in chronic hepatitis C patients with cirrhosis treated with sofosbuvir-
containing regimens. Hepatology 2014;59:2161–9.
[29] Younossi ZM, Stepanova M, Henry L, et al. Minimal impact of
sofosbuvir and ribavirin on health related quality of life in chronic
hepatitis C (CH-C). J Hepatol 2014;60:741–7.
[30] Younossi ZM, Stepanova M, Zeuzem S, et al. Patient-reported outcomes
assessment in chronic hepatitis C treated with sofosbuvir and ribavirin:
the VALENCE study. J Hepatol 2014;61:228–34.
551
[31] Younossi ZM, Jiang Y, Smith NJ, et al. Ledipasvir/sofosbuvir regimens
for chronic hepatitis C infection: insights from a work productivity
economic model from the United States. Hepatology 2015;61:1471–8.
[32] Younossi Z, Henry L. The impact of the new antiviral regimens on
patient reported outcomes and health economics of patients with
chronic hepatitis C. Dig Liver Dis 2014;46(Suppl. 5):S186–96.
[33] Younossi ZM, Stepanova M, Afdhal N, et al. Improvement of health-
related quality of life and work productivity in chronic hepatitis C
patients with early and advanced fibrosis treated with ledipasvir and
sofosbuvir. J Hepatol 2015;63:337–45.
[34] Wells GA, Russell AS, Haraoui B, et al. Validity of quality of life
measurement tools—from generic to disease-specific. J Rheumatol
Suppl 2011;88:2–6.
[35] Wiebe S, Guyatt G, Weaver B, et al. Comparative responsiveness of
generic and specific quality-of-life instruments. J Clin Epidemiol
2003;56:52–60.
[36] Younossi ZM, Boparai N, McCormick M. A disease-specific health-
related quality of life instrument for chronic hepatitis C: CLDQ-HCV.
Hepatology 2000;32(369A):838.
[37] Younossi ZM, Guyatt G, Kiwi M, et al. Development of a disease specific
questionnaire to measure health related quality of life in patients with
chronic liver disease. Gut 1999;45:295–300.
[38] Loria A, Escheik C, Gerber L, et al. The Chronic Liver Disease
Questionnaire-Hepatitis C (CLDQ-HCV): a sensitive and valid health
related quality of life instrument. Hepatology 2012;56:646A.
[39] Brazier JE, Harper R, Jones NM, et al. Validating the SF-36 health survey
questionnaire: new outcome measure for primary care. BMJ
1992;305:160–4.
[40] Medline Plus. Hemoglobin levels. Available from: http://www.nlm.nih.
gov/medlineplus/ency/article/003645.htm. [Accessed August 6, 2015].
[41] Younossi ZM, Kiwi ML, Boparai N, et al. Cholestatic liver diseases and
health-related quality of life. Am J Gastroenterol 2000;95:497–502.
[42] Popovic DDJ, Kovacevic NV, Kisic Tepavcevic DB, et al. Validation of the
chronic liver disease questionnaire in Serbian patients. World J
Gastroenterol 2013;19:4950–7.
[43] Loria A, Escheik C, Gerber NL, Younossi ZM. Quality of life in cirrhosis.
Curr Gastroenterol Rep 2013;15:301.
[44] Sanyal A, Younossi ZM, Bass NM, et al. Randomised clinical trial:
rifaximin improves health-related quality of life in cirrhotic patients
with hepatic encephalopathy—a double-blind placebo-controlled
study. Aliment Pharmacol Ther 2011;34:853–61.
[45] Two R, Verjee-Lorenz A, Clayson D, et al. A methodology for
successfully producing global translations of patient reported
outcome measures for use in multiple countries. Value Health
2010;13:128–31.
[46] Schulz KH, Kroencke S, Ewers H, et al. The factorial structure of the
Chronic Liver Disease Questionnaire (CLDQ). Qual Life Res
2008;17:575–84.
[47] Bondini S, Kallman J, Dan A, et al. Health-related quality of life in
patients with chronic hepatitis B. Liver Int 2007;27:1119–25.
[48] Dan AA, Kallman JB, Wheeler A, et al. Health-related quality of life in
patients with non-alcoholic fatty liver disease. Aliment Pharmacol
Ther 2007;26:815–20.
[49] Kallman J, O’Neil MM, Larive B, et al. Fatigue and health-related quality
of life (HRQL) in chronic hepatitis C virus infection. Dig Dis Sci
2007;52:2531–9.
[50] Hirano H, Saito M, Yano Y, et al. Chronic liver disease questionnaire
would be a primary screening tool of neuropsychiatric test detecting
minimal hepatic encephalopathy of cirrhotic patients [published online
ahead of print July 31, 2015]. Hepatol Res. http://dx.doi.org/10.1111/
hepr.12442.
[51] Reeve, Bryce B, PROMIS Cooperative Group. Psychometric evaluation
and calibration of health-related quality of life item banks: plans for
the Patient-Reported Outcomes Measurement Information System
(PROMIS). Med Care 2007;45:S22–31.
[52] Younossi ZM, Stepanova M, Pol S, et al. The impact of ledipasvir/
sofosbuvir on patient-reported outcomes in cirrhotic patients with
chronic hepatitis C: the SIRIUS study [published online ahead of print
June 9, 2015]. Liver Int. http://dx.doi.org/10.1111/liv.12886.
[53] Younossi ZM, Stepanova M, Marcellin P, et al. Treatment with
ledipasvir and sofosbuvir improves patient-reported outcomes: results
from the ION-1, -2, and -3 clinical trials. Hepatology 2015;61:1798–808.
[54] Ong JP, Oehler G, Krüger-Jansen C, et al. Oral L-ornithine-L-aspartate
improves health-related quality of life in cirrhotic patients with hepatic
encephalopathy: an open-label, prospective, multicenter observational
study. Clin Drug Investig 2011;31:213–20.