Lack of patient education is risk factor of disease flare in patients with systemic lupus erythematosus in China

Self-reported adherence was assessed using Compliance Questionnaire for Rheumatology (CQR). It was proved of good reliability and validity in lupus patients by previous study [15]. The CQR consists of 19 items about taking medicine, in which patients were asked the degree of agreement with statements. Answers are based on four-point Likert scales from 4 to 1 [16] (4: agree very much;3: agree; 2: do not agree; 1: do not agree at all;). The final point allows the identification of non-adherent patients (defined as poor compliance ≤80%) with a small false-positive rate.

Patient beliefs about medicines were assessed using the Beliefs about Medicines Questionnaire, BMQ. This questionnaire has been confirmed of its reliability and validity before [17], and we also found Cronbach’s α of 0.88 for necessity scale and 0.77 for concern scale in this study (Additional file 1). The BMQ-specific quantifies patients’ beliefs about the necessity of a prescribed medication to control their disease, as well as their concerns about the potential side effects of the medication [18]. Both beliefs about necessity and concerns about side effects were measured in terms of 5 items rated on a 5-point Likert scale and the total scores of the Necessity and Concerns scales range from 5 to 25. By subtracting a patient’s concerns score from the his/her necessity score, a necessity–concerns differential was obtained, ranging from − 20 to 20 [19]. Higher differential scores indicate the stronger perceived necessity about medication necessity and/or lower concerns about its side effects.

Levels of patients’ satisfaction with information on treatment were evaluated using the Chinese version of the validated Satisfaction with Information about Medicines Scale (SIMS). It investigates whether patients feel that they have been given adequately informed on prescribed drugs [20]. The questionnaire comprises 18 items, each relating to a particular aspect of drug use. An overall satisfaction rating was obtained by summing the scores of all 18 items, ranging from 0 to 18 [21]. The higher the score, the greater satisfaction a patient felt with information. Patients were asked to make satisfaction ratings on the information provided for them using the 5 items: “too much”, “about right”, “too little”, “none received”, and “none needed”. Ratings of “about right” or “none needed” suggested patients’ satisfaction with the information and were given a score of 1, while ratings of “too much”, “too little” or “none received” suggested dissatisfaction and were scored as 0. The internal reliability of the scale scores was good with a Cronbach’s α coefficient of 0.95. Total score of items 1–9 was used to measure patients’ satisfaction with information on action and usage (score ranges from 0 to 9) (Cronbach’s α 0.94); total score of items 10–18 was used to measure the satisfaction with information on potential problems (score ranges from 0 to 9) (Cronbach’s α 0.94) (Additional file 1). For the three SIMS scales, median scores were used to define dissatisfaction (< 16 of 18 items for overall satisfaction rating; < 8 of 9 items for subscale of action and usage and potential problems). In consideration of the left-skewed distribution of the three scales, they were dichotomized into satisfaction coded as 1 versus dissatisfaction as 0.

The quality of life was accessed using the Chinese version of the general population-based three-level EuroQol five-dimensions questionnaire [EQ-5D-3 L] [22, 23]. Each EQ-5D-3 L health state was scored as 1 (no problems), 2 (some/moderate problems), or 3 (extreme problems) to indicate functional levels in five dimensions: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. The EQ-5D-index was calculated by “time-trade-off”, to assess the patients’ quality of life.