Adaptation of the Texas Christian University Organisational Readiness for Change Short Form (TCU-ORC-SF) for use in primary health facilities in South Africa

Objectives: The Texas Christian University Organisational Readiness for Change Scale (TCU-ORC) assesses factors influencing adoption of evidence-based practices. It has not been validated in low- and middle-income countries (LMIC). This study assessed its psychometric properties in a South African setting with the aim of adapting it into a shorter measure.

Methods: This study was conducted in 24 South African primary health care clinics in the Western Cape Province. The TCU-ORC and two other measures, the Organisational Readiness to Change Assessment (ORCA) and the Checklist for Assessing Readiness for Implementation (CARI) were administered. The questionnaire was re-administered after 2 weeks to obtain data on test-retest reliability. Three hundred and ninety-five surveys were completed: 281 participants completed the first survey, and 118 re-completed the assessments.

Results: We used exploratory factor analysis (EFA) to identify latent dimensions represented in the data. Cronbach's alpha for each subscale was assessed and we examined the extent to which the subscales and total scale scores for the first and retest surveys correlated. Convergent validity was assessed by the correlation coefficient between the TCU-ORC, ORCA and CARI total scale scores. EFA resulted in a three-factor solution. The three subscales proposed are Clinic Organisational Climate (8 items), Motivational Readiness for Change (13 items), and Individual Change Efficacy (5 items) (26 items total). Cronbach's alpha for each subscale was >.80. The overall shortened scale had a test-retest correlation of r = .80, p < 0.01, acceptable convergent validity with the ORCA scale (r = .56, p < .05), moderate convergence with the CARI (r = 39, p<.05) and strong correlation with the original scale (r = 0.79, p < .05).

Conclusions: This study presents the first psychometric data on the TCU-ORC from a LMIC.  The proposed shortened tool may be more feasible for use in LMICs.