Testing the psychometric properties of a Chinese version of Dyadic Relationship Scale for families of people with hypertension in China

Abstract Background Interventions for people with chronic illness have increasingly got involvement and partnership with family members in China and worldwide. The patient-family interactions in chronic illness care can greatly influence not only family dyadic relationship or collaboration in caregiving but also both patient’s and caregiver’s health and well-being. To date, very few instruments have been developed to measure the family dyadic relationship; and none has been found in Chinese language. This study aimed to translate the original English Dyadic Relationship Scales (DRS), including DRS-patient and DRS-caregiver, into simplified Chinese language (DRS-C) and examine their psychometric properties in Chinese people with hypertension in a rural community. Methods The Brislin’s model of translation was adopted for translation. Face and content validity and semantic equivalence of the translated Chinese version of the two DRS-patient and DRS-caregiver scales were examined. A sample of 132 adults with hypertension and their family caregivers were recruited to test the psychometric properties of the DRS-C scales. Results The DRS-C scales indicated very satisfactory face validity with 10 family dyads in hypertension care, content validity rated by five experts (Item CVI = 0.8–1.0; Scale CVI = 0.98) and semantic equivalence rated by 22 panel members (i.e., all items were rated as 3 = relevant or 4 = very relevant by > 18 members). A well-fitting model of DRS-C-patient was identified with χ2/df = 1.47, p = 0.04, RMSEA = 0.06, GFI = 0.941, CFI = 0.961, TLI = 0.947, and SRMR = 0.019. A well-fitting model of DRS-C-caregiver was identified with χ2/df = 1.340, p = 0.092, RMSEA = 0.039, GFI = 0.940, CFI = 0.975, TLI = 0.965, and SRMR = 0.014. The Chinese DRS-patient and DRS-caregiver had satisfactory internal consistency with Cronbach’s α coefficients of 0.82 and 0.83, respectively, and test–retest reliabilities with intra-correlation coefficients of 0.97 and 0.96, respectively. The convergent validities of the Chinese versions of the DRS-patient and DRS-caregiver were very satisfactory with the self-efficacy subscale of Hypertension Self-Care Profile, and Zarit Burden Interview, respectively (Pearson’s r = − 0.70 and 0.79; both p