Patient activation in Europe: an international comparison of psychometric properties and patients' scores on the short form Patient Activation Measure (PAM-13)

Abstract

Background

To allow better assessment of patients’ individual competencies for self-management, the Patient Activation Measure (PAM) has been developed in the USA. Because the American studies have shown the PAM to be a valuable tool, several European countries have translated the instrument into their native languages (Danish, Dutch, German, Norwegian). The aim was to compare the psychometric properties in studies from the different countries and establish whether the scores on the PAM vary between the studies. Methods

Data from the four separate studies were subjected to the same data cleaning procedures and statistical analyses. The psychometric properties of the instruments were established with measures of data quality and scale structure. The mean patient activation score and distribution across four predefined activation levels were described and the differences between the four studies were tested with ANOVA (unadjusted and adjusted) followed by a post-hoc Tukey HSD test and the Pearson chi-squared test respectively. Results

The total N of the four studies was 5184. The percentage of missing values was low in all datasets, confirming the good quality of the datasets. Factor analyses revealed moderate to strong factor loadings on the first factor in all datasets. Cronbach’s α was high for all version, ranging from .80 (German) to .88 (Dutch). Item-rest correlations varied between .32 and .66, indicating a moderate to strong correlation of the individual items to the sum scale.

Both the mean PAM score and the distribution across activation levels differed between the four datasets. After adjustment of the PAM score, patients in Norway in particular had a higher patient activation level. Conclusions

The European translations of PAM-13 (into Danish, Dutch, German and Norwegian) resulted in four instruments with good psychometric capabilities for measuring patient activation. The mean PAM score and the distribution across activation levels differed between the four datasets.