Study design and participants
We conducted a cross-sectional study from September to November 2017 in Liangshan. Hospitals in China are designated as primary, secondary and tertiary hospitals based on a hospital’s capacity of providing medical care, education and research. Primary hospitals have less than 100 beds and focus on focus on health education, disease prevention and health care, rehabilitation, family planning, and common and frequent disease management. Secondary hospitals usually have beds between 100 and 500 and provide comprehensive health care, medical education and research. Tertiary hospitals typically have more than 500 beds and provide comprehensive health services at the city, provincial or national level and carry a more important role in medical education and research [31].
Using a stratified, random cluster sampling method, our study sites included 3 primary, 5 secondary and 2 tertiary hospitals in Liangshan. Our target participants were registered nurses aged 18 or older who had worked in the selected hospitals for at least 6 months and consented to participate.
Data Collection
In each hospital, we first contacted the head nurse of its nursing department for approval and assistance in recruiting nurses. Once we received approval, we sent eligible nurses an online survey link, including a consent form describing the purpose, procedures, the potential risks and benefits of the study. Nurses who consented to participate received a link to access the online anonymous survey.
Variables and Instruments
Our survey consisted of variables shown in the literature to be related to HIV-related stigma in healthcare providers. It included questions about individual and workplace characteristics, HIV-related stigma scale for health care workers [32], and HIV knowledge scale[33]. Both the stigma and knowledge scales have been tested in Chinese population [32,33].
Individual and workplace characteristics
Individual characteristics included age, gender, ethnicity, educational level, years of experience working as a nurse, experience of providing care to PLWH, prior experience of HIV-related training, and willingness to receive HIV-related training. Workplace characteristics consisted of questions regarding level of hospital s/he worked, HIV prevalence area (high vs. low; high prevalence was defined as an area with >1% HIV prevalence), prior experience working in AIDS specialized hospitals, universal precaution supplies and policies in workplace that protect PLWH from stigma.
HIV-related stigma
We used a 17-item “HIV-related stigma scale for health care workers” developed by Stein and Li [32] to assess the HIV-related stigma among nurses. This multidimensional scale includes 5 subscales: Discrimination Intent at Work (4 items; e.g., you would be willing to work with HIV positive patients), Opinion about Health Care for HIV-infected Patients (3 items; e.g., people who got infected with HIV/AIDS through drug use deserve good quality medical care), Prejudiced Attitudes (4 items; e.g., people who got HIV/AIDS through sex and drug use got what they deserved), Internalized Shame (3 items; e.g., if you worked with HIV positive patients, you would feel embarrassed to tell other people about it), and Fear of PLWHA (3 items; e.g., you feel afraid of PLWH). Scores for each item ranges from 1(strongly agree) to 5 (strongly disagree). Some items were reverse-scored (e.g., the Opinion about Health Care for HIV-infected Patients subscale) when appropriate, and a higher total score suggested higher levels of HIV-related stigma toward PLWH. Stein and Li [32] reported adequate internal consistency (alpha = 0.68-0.82) and construct validity of the scale.
HIV knowledge
We measured HIV knowledge using a Chinese version HIV knowledge scale [33] adapted from Jemmott and colleagues’ work [34]. It is a 24-item scale assessing nurses’ knowledge of basic characteristic and transmission routes of HIV. Responses of each scale item were: “true”, “false” or “not sure”. Participants received 1 point if their answers were correct, and they received 0 points if their answers were incorrect or “not sure.” A higher total score indicated higher levels of HIV knowledge.
Statistical analyses
We used IBM SPSS 22.0 [35] to manage and analyze data. We used the mean, standard deviation (SD), frequency and percentage to describe distributions of variables. Pearson's r correlations, t-tests and one-way analyses of variance (ANOVA) were used to examine relationships between key variables and HIV-related stigma toward PWLH. We further conducted multiple stepwise regression analysis to investigate factors associated with HIV-related stigma toward PLWH. A p-value less than 0.05 was used to indicate statistical significance.
Power estimate: We calculated the required sample size based on the formula below [36] The required study size was 1,225 (S = 8.92; effective sizeδ= 0.50; a= 0.05).