Quality of physical therapy from a patient's perspective; factor analysis on web-based survey data revealed three
Marijn Scholte1, Hilly Calsbeek, Maria W G Nijhuis-van der Sanden
1Scientific Institute for Quality of Healthcare, Radboud university medical centre, Geert Grooteplein 21, 6525 EZ Nijmegen, The Netherlands. marijn.scholte@radboudumc.nl.
Background:
Assessing quality of care from the patient's perspective has changed from patient satisfaction to the more general term patient experience, as satisfaction measures turned out to be less discriminative due to high scores. Literature describes four to ten dimensions of patient experience, tailored to specific conditions or types of care. Given the administrative burden on patients, less dimensions and items could increase feasibility. Ten dimensions of patient experiences with physical therapy (PT) were proposed in the Netherlands in a consensus-based process with patients, physical therapists, health insurers, and policy makers. The aim of this paper is to detect the number of dimensions from data of a field study using factor analysis at item level.
Methods:
A web-based survey yielded data of 2,221 patients from 52 PT practices on 41 items. Principal component factor analysis at item level was used to assess the proposed distinction between the ten dimensions.
Results:
Factor analysis revealed two dimensions: 'personal interaction' and 'practice organisation'. The dimension 'patient reported outcome' was artificially established. The three dimensions 'personal interaction' (14 items) (median(practice level) = 91.1; IQR = 2.4), 'practice organisation' (9 items) (median(practice level) = 88.9; IQR = 6.0) and 'outcome' (3 items) (median(practice level) = 80.6; IQR = 19.5) reduced the number of dimensions from ten to three and the number of items by more than a third.
Conclusions:
Factor analysis revealed three dimensions and achieved an item reduction of more than a third. It is a relevant step in the development process of a quality measurement tool to reduce respondent burden, increase clarity, and promote feasibility.
