10
a given medical condition. In the value-based healthcare framework hierarchy of
outcome domains, patient health status achieved is the highest tier. Consequently,
PROs will play an important role in the assessment of delivered quality of care.
Population Studies
Finally, PROs are frequently used in population studies to describe the burden of
illness faced in a certain condition and to provide information about the health pro-
file and healthcare needs of the population. The objective of this type of studies is
frequently to compare the burden of illness between populations. Generic instru-
ments are typically most suitable for this. Comparisons across different diseases,
population groups, or interventions require instruments that are reliable.
Single- and Multi-Item Scales
Single-Item Scales
PROs can be either single-item scales or multi-item instruments. Single-item instru-
ments are typically used for the assessment of specific symptoms/impairments (i.e.,
pain or stiffness) or otherwise simple concepts that require patients to report on
information that is readily retrieved from memory and requires relatively little cogni-
tive processing [ 41 ]. For instance, most people can give sufficiently accurate reports
regarding their disease duration or employment status and therefore such variables
are typically assessed in a single-item format. The most commonly employed single-
item instruments in rheumatology are numerical rating scales (NRS) and visual ana-
logue scales (VAS). A VAS is comprised of a line, 100 mm in length, anchored by 2
verbal descriptorsre presenting the domain extremes. The NRS is a numeric version
of the visual analogue scale in which patients are asked to select the integer (typi-
cally 0–10) that best reflects their standing on the measured trait. Measurement prop-
erties of NRS and VAS are generally similar [ 11 ]. Less frequently used single-item
instruments are Likert scales. These are dissimilar to NRS in that they typically
contain fewer response options (usually 5) and are characterized by an equal num-
bers of positive and negative positions around a neutral response option. Although it
has been said that patients might prefer the clarity provided by Likert scales, they
provide less statistical information than NRS/VAS and have consistently been shown
to be inferior in comparative studies [ 42 ]. General advantages of single-item scales
over multi-item instruments are that they are easy to implement and to interpret, are
least burdensome to patients, and in many cases provide relevant information for
monitoring outcomes over time. However, limitations of single-item instruments are
that they provide less statistical information compared with more elaborate tools,
which undermines their reliability. Moreover, it is frequently difficult to fully char-
acterize a domain using single-item instruments.
M.A.H.O. Voshaar and M.A.F.J. van de Laar