Taking the PACIC back to basics: the structure of the Patient Assessment of Chronic Illness Care.

Rationale, Aims And Objectives: The Patient Assessment of Chronic Illness Care (PACIC) is a widely used 20-item measure consisting of five subscales. Published factor analyses of PACIC scores have produced conflicting results on the measure's factorial validity, and therefore some confusion as to the utility of its subscales. We aim to reduce this confusion by reviewing the evidence on the PACIC's factorial validity, exploring the statistical issues it raises, and considering more broadly what such analyses can reveal about the validity of the PACIC.

Methods: To achieve these aims we review six published studies on the PACIC's factorial validity, present confirmatory factor analyses of our own PACIC data from 251 chronic care patients, and assess the PACIC with respect to its status as a reflective or a formative measure.

Results: Our statistical analyses support the view that a 5-factor model does not fit the structure of the PACIC, and highlight a variety of technical issues that confront researchers who wish to factor analyse the measure. However, we argue that, as the PACIC is more accurately seen as a formative measure, such analyses do not provide information that should be used to assess the PACIC's validity.

Conclusions: We conclude that, while it is important to continue examining the reliability and validity of the PACIC in a variety of ways, traditional analyses of its factorial validity (and internal consistency) are inappropriate. Meanwhile, use of the subscales is defensible as long as they continue to meet other types of reliability and validity requirements.