fibrillation based on the analysis of a single symptom, i.e., an
irregular heart rhythm [16,23]. As an alternative, we con-
structed a questionnaire that covers the main clinical mani-
festations (palpitation, breathlessness, chest pain, and dizziness)
and usual therapies for this arrhythmia (cardioversion, drugs
and ablation). We compared our questionnaire with the SF-36,
which is cited in more than 4000 publications and has been
translated in more than 50 countries, with at least three publi-
cations on its scientific validity [24,25].
AF is a complex arrhythmia in terms of both symptoms
and therapeutic approach; its impact on quality of life involves
clinical manifestations and treatment modalities, so assessment
of patient quality of life must include these aspects. Therefore,
the questionnaires used for this purpose in large trials are not
satisfactory.
4.2. Reproducibility
Analysis of reproducibility, reliability, or precision is a
necessary but not sufficient part of instrument validation. Not
only should an instrument measure what it is supposed to, but
results must be replicable.
In the present study, the coefficients obtained in repro-
ducibility analysis of time taken to administer QLAF were
adequate (intra- and inter -observ er analysis alpha coef ficients
were 0.95 and 0.77). A better alpha coefficient for intra-observer
analysis is an initial sine qua non condition for validating a test
but inter-observer agreement is the basis for scientific inter-
objectivity [26].
On analyzing the reproducibility of inter and intra-observer
results using mean total QLAF scores (Tables 4 and 5), excellent
coeffici ents were obtained with both methods — internal con-
sistency (Cronbach= 0.98 and 0.96,) and test–retes t (Bartko = 0.95
and 0.91). Results show excellent reliability; the QLAF is highly
reproducible. Analyzing inter and intra-observer reproduci-
bility for each of the seven QLAF domains by both methods
also yielded very good results, with excellent reproducibility
coefficients for most domains. However, we found slightly
lower coefficients for Drugs (Cronbach=0.58 and Bartko=0.41)
and Palpitation (Bartko=0.53) domains on inter-observer anal-
ysis. We believe that compliance with drug use was a con-
founding factor in the Drugs domain [27]. Patients may tell one
observer they take their medications, and then confess they do
not to another. This would explain the lack of a high repro-
ducibility coefficient for this domain. A lower Bartko intraclass
correlation coefficient was also obtained for the Palpitation
domain, although the alpha coefficient was adequate (0.70).
Analysis of reproducibility by the internal consistency method
(Cronbach=0.70) showed that this domain is reproducible
between distinct observers. The test–retest method measure-
ments showed stable characteristics over time. Palpitation is
a dynamic symptom, which represents the variation in fre-
quency and intensity of an arrhythmia; it is affected by cog-
nitive interference and due to hemodynamic adaptation may go
unnoticed in chronic AF patients. Therefore, we must ask how
well questions in a domain fit together conceptually using the
internal consistency method. Stability is not a determinant of
reproducibility for a domain characterized by “instability”, even
between morning and afternoon periods.
4.3. Responsiveness
In addition to AF patient quality of life measured by differ-
ent persons at distinct times and places, the QLAF also detected
changes in clinical conditions. Analysis of responsiveness
over time confirmed this property, with significant differences
in mean domain scores, particularly Palpitation, Breathlessness,
Dizziness and Drugs, and total score on comparing baseline
and end of follow-up (Table 6). The ability to detect variations
in AF patients' clinical conditions is even more interesting since
variable symptoms occur almost randomly over time.
4.4. Limitations
One limitation of the present study was that we did not
interview the target population to determine content validity, an
approach rarely used but frequently cited when assessing quality
of life. In addition to the literature and expert opinions, con-
firmation of questions corresponding to patient expectations
would be useful. Five AF patients interviewed surprisingly said
that fear was the worst symptom and prolonging life was more
important than improving current health, which is an interesting
topic for future research.
The present study offers a reliable tool for quality-of-life
assessment through different types of clinical interventions.
Further studies involving a larger number of patients will be
necessary for the separate analysis of different AF subgroups,
as well as to measure final outcomes. Furthermore, separate
analysis of quality of life and hospitalization are needed for
direct-current cardioversion, ablation or changes in drug thera-
py in order to obtain reliable responses.
4.5. Clinical implications
When evaluating quality of life, the SF-36 and QLAF com-
plement one another since the generic questionnaire is important
for the comparison with other health states and provides an
overall assessment of any health condition or disease, but is
unable to detect certain particular or specific aspects related to
the quality of life associated with a specific disease.
Considering the complexity of AF and its therapeutic arsenal,
developing an instrument to capture quality of life is a challenge.
The QLAF's substantial clinical content makes it intuitive for
professionals and no other AF-specific questionnaire covering
the main clinical manifestations and treatments is found in
the literature. Simple and direct language facilitates use with
patients of different educational levels, and easy-to-use scoring
provides quick results. Detailed analysis of measurement prop-
erties showed consistent evidence for validity, reproducibility ,
and responsiveness. QLAF takes just three minutes and may be
used
as part of an AF patient's routine visit to provide overall
assessment of the disease and individual.
6 É.O.V. Braganca et al. / International Journal of Cardiology xx (2009) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Braganca ÉOV, et al, Validating a new quality of life questionnaire for atrial fibrillation patients, Int J Cardiol (2009), doi:10.1016/
j.ijcard.2009.03.087