Testing the validation and reliability of the Caregiver-Patient Activation Measure (CG-PAM)

Katherine Carleton-Eagleton; Iain Walker; Diane Gibson; Nicole Freene; Stuart Semple

doi:10.1016/j.pecinn.2022.100098

Testing the validation and reliability of the Caregiver-Patient Activation Measure (CG-PAM)

PEC Innov. 2022 Nov 9:1:100098. doi: 10.1016/j.pecinn.2022.100098. eCollection 2022 Dec.

Authors

Katherine Carleton-Eagleton¹, Iain Walker², Diane Gibson³, Nicole Freene³, Stuart Semple⁴

Affiliations

¹ Faculty of Health, University of Canberra, 11 Kirinari Street, Bruce, ACT 2617, Australia.
² School of Psychological Sciences, University of Melbourne, Level 7, Redmond Barry Building, The University of Melbourne, Victoria, 3010, Australia.
³ Health Research Institute, Faculty of Health, University of Canberra, 11 Kirinari Street, Bruce, ACT 2617, Australia.
⁴ Health Research Institute and Research Institute for Sport & Exercise, Faculty of Health, University of Canberra, 11 Kirinari Street, Bruce, ACT 2617, Australia.

Abstract

Objective: The objective of this study was to test the reliability and validity of the Caregiver-Patient Activation Measure (CG-PAM).

Methods: Based on the psychometric testing of the original Patient Activation Measure (PAM), three assessments of reliability and validity were completed on the CG-PAM. Test-retest reliability was assessed across two weeks (n = 23). Criterion validity was assessed by interviewing participants from the test-retest cohort (n = 10), with transcripts assessed by subject matter experts (n = 3) to classify activation levels of the interviewee. Construct validity was assessed through a survey (n = 179) consisting of demographic questions, the CG-PAM and concepts hypothesised to be related to caregiver activation.

Results: There was strong test-retest reliability (r = 0.893), but poor criterion validity. Assessment o;f construct validity demonstrated significant relationships found between caregiver activation and weekly hours of care provided (p < 0.001), relationship satisfaction (p < 0.001), and dyad typology (p < 0.001), but not with perceived levels of stress or social support.

Conclusion: The CG-PAM was found to have strong reliability, but there were inconsistent results across the validation tests conducted.

Innovation: Future research must consider the dynamic nature of caring and the importance of the relationship between the caregiver/recipient when defining activation levels within the CG-PAM.

Keywords: Activation; CG-PAM; Caregivers; Chronic disease; Validity testing.