Development of a determinant scale of pharmacist’s care for herbal and dietary supplement users

Abstract

Objective: There is a lack of instrument that is valid and reliable to measure the determinant of pharmacist’s care (PCare) for herbal and dietary supplement (HDS) users. The present study aimed to develop scales to measure the determinant of PCare for HDS users and to quantify the practices. The study utilized qualitative and quantitative approaches.

Methods: Three scales were developed: Direct TPB scale based on a modified Theory of Planned Behaviour (m-TPB) framework, Indirect TPB scale based on the m-TPB and beliefs elicited from community pharmacists (CPs), and PCare-HDS scale based on findings from the qualitative study. This study involved three phases: (1) qualitative study to explore CPs’ salient practices of PCare for HDS users, and the beliefs associated with the practices; (2) generation of item pools; and (3) psychometric evaluation of the scales.

Results: The qualitative study identified eight domains of PCare for HDS users, divided into two dimensions of direct and non-direct customer/patient care activities, and identified relevant beliefs underlying the practices. Item generation procedures produced pools of items for the scales. These items were examined and refined by a group of pharmacy experts in a content validity study. A face validity study established the clarity and comprehensibility of the items. For the third phase, a cross-sectional survey was carried out among CPs in Bangkok. The sample was divided into two datasets: sample 1, to explore the factor structures and to refine the scales; and sample 2, to cross-validate the factor structures. Additionally, Rasch analysis and criterion validity tests were performed on the entire sample of CPs. Exploratory factor analysis (EFA) supported the factor structures of the Direct TPB and PCare-HDS scales, and identified an additional factor for the Indirect TPB scale. Both the EFA and a preliminary confirmatory factor analysis (CFA) guided the refinement of the scales. Data from the second dataset for all three scales fitted well with the models using CFA. Discriminant and convergent validity were shown for the scales. Rasch analysis showed no substantial misfit and the category functioning followed monotonic increases in mean and step measures for all domains of the scales. All constructs of the scales had good to excellent internal consistency reliability. The scales had criterion validity to some extent.

Conclusion: The scales were shown to have validity and reliability and can be of interest to researchers aiming to understand PCare for HDS users. Further study can validate the scales in different sample of CPs such as in the Northern and Southern regions of Thailand.