Reliability and validity of the Chinese version of the HIV Treatment Adherence Self-Efficacy Scale in mainland China

Abstract

The aim of this study was to translate the HIV Treatment Adherence Self-Efficacy Scale (HIV-ASES) and evaluate its reliability and validity in people living with HIV (PLWHIV) in mainland China. The original HIV-ASES was translated from English into Chinese and adapted for cultural context according to cross-cultural adaptation guidelines. A cross-sectional survey investigating 1742 PLWHIV receiving antiretroviral treatment was conducted. Exploratory factor analysis with varimax rotation confirmed the presence of two factors with an accumulated contribution rate of 58.357%. Furthermore, confirmatory factor analysis revealed the model’s goodness-of-fit index. The comparative fit index was greater than 0.9 and the overall efficacy of the model was satisfactory. The Cronbach α coefficient of the questionnaire was 0.876 (95% CI: 0.868–0.885), and the correlation coefficient of each item and the total was 0.536–0.660 (P < 0.01). The Chinese version of the HIV-ASES has high reliability and validity; however, the individual item requires some adjustment.

Keywords

Chinese HIV AIDS adherence self-efficacy

Introduction

AIDS is a chronic infectious disease affecting people worldwide. There were approximately 36.9 million people living with HIV at the end of 2014 with 2.0 million people becoming newly infected with HIV in 2014 globally.¹ In the same year, China reported 104,000 new HIV/AIDS cases in 2014, up 14.8% from the previous year.² The number of people living with HIV (PLWHIV) currently receiving antiretroviral therapy (ART) increased from 227,489 in 2013 to 295,358 in 2014 in China.³

The emergence of ART has been proven to be beneficial to PLWHIV^4,5; however, ART cannot completely eliminate HIV from the body. In addition, the side effects of ART drugs considerably affect PLWHIV health and treatment adherence.^6,7 Studies have shown that high adherence prevents the spread of diseases, reduces the morbidity and mortality of HIV/AIDS, effectively prolongs the life of PLWHIV, and reduces the resistance of the virus. Low adherence may lead to viral mutation, a decreased quality of life among PLWHIV, decreased ART drug efficiency, and premature death.^8–11 Therefore, high treatment adherence has become crucial in the treatment of HIV/AIDS.

Self-efficacy is the determination of whether a person has the ability to complete a certain behavior. It consists mainly of two parts: self-efficacy beliefs and outcome expectation. Self-efficacy beliefs, also known as efficacy expectation, refer to the ability of individuals to complete a specific behavior; these beliefs are expected to predict behavior that will produce positive results.^12,13 Also, self-efficacy has a potential intervening role among PLWHIV^14,15 and can affect their daily decision-making; people with higher self-efficacy tend to have stronger confidence in overcoming difficulties. Because of an incomplete understanding and misunderstandings of the disease, PLWHIV experience feelings of inferiority, fear, and pessimism; however, enhancing self-efficacy can stimulate their self-care motivation, thereby improving medication adherence.

Current study evaluating medication adherence quantitatively of PLWHIV is few in China. Huang et al.¹⁶ used the translated version of HIV Self-Efficacy Questionnaire comprised of 34-item made by Shively et al.¹⁷ to measure the self-efficacy of PLWHIV. Zhang et al.¹⁸ assessed adherence self-efficacy with a scale based on their qualitative interviews; however, the validity of the scale couldn’t be determined. Beyond that, most medication adherence studies are qualitative evaluation or used common quantitative scales in China, such as the Morisky scale,¹⁹ but the reliability and validity of the scale were not high.^20,21 The reliability and validity of the HIV Treatment Adherence Self-efficacy Scale (HIV-ASES) with 12-item in American sample are very high.²² It is more convenient and feasible in the follow-up work. However, at present, no report has addressed a Chinese translation. Therefore, the purpose of this paper is to translate this scale into Chinese and evaluate the reliability and validity of the Chinese version of HIV-ASES.

Methods and materials

Participants

This survey was administered at regular follow-up examinations. Eligible participants met the following criteria: inclusion criteria: (a) be at least 18 years old, (b) have received a confirmed diagnosis of HIV infection, (c) had been prescribed and undergoing ART for at least the previous month, (d) have provided written informed consent. The exclusion criteria: have severe neuropsychological impairment or psychosis. Questionnaires were completed by the participants themselves. For those who were unable to fill in the questionnaire for reading difficulties or personal preference, the participants were conducted by highly trained interviewers. The research protocol was approved by the ethical committee of Zhengzhou University in China.

Procedures

Multistage stratified random cluster sampling was used to recruit participants receiving ART at the Henan Province Center for Disease Control and Prevention in February 2014. The cities or counties with PLWHIV in Henan province are divided into high incidence areas, middle areas, and low incidence areas. The details are as follows: in the first stage, two cities or counties were selected randomly from high incidence areas (Shangcai, Shenqiu), two from middle incidence areas (Queshan, Yongcheng), and two from low incidence areas (Jiyuan, Sanmenxia). In the second stage, two towns were randomly selected from Shangcai and Shenqiu, respectively, three towns randomly selected from Queshan and Yongcheng, respectively, and all PLWHIV were enrolled from Sanmenxia and Jiyuan according to probability proportional to size (PPS) sampling by a random number generator. Finally, 1742 participants completed the questionnaires. A total of 300 participants from the 1742 participants completed the questionnaires by stratified random sampling according to PPS for assessing the test–retest reliability at three and 12 months.

Translation of the HIV-ASES

The translation process comprised the following six stages: translation, integration, back-translation, expert consultation, pretesting, and cultural adjustment of the scale. First, two English language professionals independently translated the scale into Chinese. The first professional was familiar with the scale content, whereas the second was blinded to the content. According to the translation results, H1 and H2 were formed. Second, another English language professional integrated H1 and H2, and H3 was formed after revision of the first draft of the Chinese version of the HIV-ASES. Third, two native English-speaking bilingual translators who were unfamiliar with the scale back-translated H3 into English in two rounds, thus forming the translated versions of BH1 and BH2. Fourth, we invited a team of five experts, comprising three HIV/AIDS research experts, a psychology expert, and a linguistics expert, to evaluate word appropriateness, determine whether the wording accorded with applicable Chinese expressions, and evaluate the degree to which the scale items of the Chinese version reflected the measured concepts. According to the evaluation, the experts modified the items to ensure that the Chinese translation closely resembled the original English scale in content, semantics, and format.²³ Ultimately, the new self-efficacy scale measuring the treatment adherence of PLWHIV self-efficacy scale was determined to be equivalent to the HIV-ASES and was named HIV-ASES Chinese Version-2015. Finally, the translated instrument was assessed employing 120 respondents at the pretest stage. All items were clearly explained to the respondents, and the formal measurement scale of the HIV-ASES Chinese Version-2015 was implemented.

HIV/AIDS follow-up treatment questionnaire

A self-designed HIV/AIDS follow-up treatment questionnaire included items on respondent age, gender, race, educational level, income status, marital status, employment, most recent CD4 cell count, route of HIV infection, time of starting antiretroviral treatment, and medication adherence rate in the past month.

Medication adherence to ART was measured with a visual analog scale developed by Walsh et al.²⁴ in the past month for drugs along a continuum anchored by 0–100%. Participants marked the point along the line that most closely reflected daily dose HIV medications they have taken in the last month. This measure was dichotomized as 100 versus <100%.²⁵

HIV-ASES

The HIV-ASES comprises 12 items, used to evaluate participant’s confidence and integration in the implemented treatment plan. According to the study conducted by Johnson et al.,²² the scale is categorized into two dimensions: integration and perseverance. The participants rate their agreement with the items from 0 (cannot do it at all) to 10 (can certainly do it). The highest total score is 120 points; a higher score corresponds to higher adherence self-efficacy. The items of the scale are shown in Table 2 and Appendix 1.

Community Programs for Clinical Research on AIDS Antiretroviral Medication Self-Report (CPCRA)

Adherence self-report questionnaire

A translated version of the CPCRA questionnaire was used to evaluate the medication adherence of participants undergoing ART.^26–28 The specificity and sensitivity of the questionnaire have been previously demonstrated.^29–31 The participants were asked how often they took tablets on time and how many they took at one time within the seven days prior to the survey. For each antiretroviral drug, participants indicated the number of tablets taken within one week according to five options in the questionnaire. The options of ‘all’ and ‘most’ indicated taking 100 and 80% of the medication, respectively, whereas ‘half’ indicated taking 50%, ‘a little’ indicated taking 20%, and ‘none’ indicated not taking any medication. We learned the number of antiretroviral tablets actually taken by the participants in the past seven days. The adherence rate of the participants was calculated according to the number of tablets taken in one week. It equals to the number of doses of the drug taken within the prior week divided by the number of tablet doses prescribed for a week multiplied by 100%.³²

Social Support Rating Scale (SSRS)

The SSRS is a 14-item scale designed by Xiao.³³ The SSRS measures the objective support received by respondents from their family, friends, colleagues, as well as their subjective feelings regarding this support and the extent of their utilization of the support. Social support comprises both material support and emotional support. The two-month test–retest reliability is 0.92.³⁴

Perceived stress scale

This study used the Chinese version of the 10-item Perceived Stress Scale (CPSS) which is a simplified version of the original 14-item scale developed by Cohen et al.,³⁵ designed to evaluate the stress experienced by respondents in the current environment. The total score was calculated by summing ratings on a 5-point Likert scale (0 = never, 1 = scarcely, 2 = sometimes, 3 = often, 4 = very often) according to the calculation rules for the concise scale. A higher score corresponds to higher levels of stress. Alpha reliability was shown to be 0.78 in Moscow, 0.82 in Taiwan, and 0.86 in the United States.³⁶

Morisky scale

The 4-item Morisky scale, developed by Morisky, is an effective method for measuring the short-term medication adherence rate. It is a generic scale comprising four questions: ‘Do you have a history of forgetting to take medicine?’, ‘Do you sometimes stop taking the medicine?’, ‘Do you stop when you feel better?’ and ‘If you feel worse, will you stop?’ The reported Cronbach α is 0.61.¹⁹

Statistical analyses

The content validity and convergent validity were evaluated using Spearman rank correlation coefficients. The convergent validity of the HIV-ASES was evaluated according to the applied CPCRA Adherence Self-Report questionnaire, Morisky scale, SSRS, CPSS, and self-reported one-month adherence rate, which are measures that assess underlying phenomena similar to those assessed through the HIV-ASES. To analyze the factorial validity of the HIV-ASES, we randomly divided the study sample into two subsamples to conduct exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). Data from the first half of the split sample (N = 902) were entered into SPSS Version 21.0 (SPSS, Inc., Chicago, IL, USA). EFA was conducted using principal component factoring with varimax rotation for factor extraction to investigate the underlying factor structure of the scale. The Kaiser–Meyer–Olkin (KMO) and Bartlett’s spherical check were carried out to text whether the sample is suitable for factor analysis. From the factors that had eigenvalues greater than 1, principal components were extracted after reviewing the scree plot. CFA was performed to confirm the factor structure obtained using robust maximum likelihood estimation in AMOS Version 17.0 (SPSS Inc., Chicago, IL, USA) to verify the revealed factor structures for the second half of the spilt sample (N = 840). The internal consistency of the questionnaire was evaluated by the Cronbach α coefficient, and correlations among subscales and other scales of the HIV-ASES were examined. Moreover, Pearson correlation coefficient was used to calculate the test–retest reliability.

Results

Sociodemographic background characteristics of the participants

A total of 1742 PLWHIV (mean age: 48 years) taking antiretroviral drugs were enrolled in this study. More than half of the participants were female (50.5%); most participants were married or living with a partner (67.3%), and most had a primary school or lower education (60.1%). The mean CD4 cell count of the participants was 411.5 ± 235.8. Eighteen percent of the participants completed the questionnaire themselves; the remaining 81.5% completed the questionnaire through one-to-one interviews because of reading difficulties or personal preference. Additional participants’ demographic characteristics are shown in Table 1.

Table 1.

Demographic characteristics of the participants.

Characteristics	N = 1742
Age – mean years (SD)	47.8 (11.0)
Gender – n (%)
Male	862 (49.5)
Female	880 (50.5)
Marital status – n (%)
Single	124 (7.1)
Married or with partner	1172 (67.3)
Separated or divorced	77 (4.4)
Widowed	369 (21.2)
Education level – n (%)
No formal education	460 (26.4)
Elementary school	587 (33.7)
Junior high school	507 (29.1)
Senior high school	154 (8.8)
College or higher	34 (2.0)
Occupation – n (%)
farmer	1362 (78.2)
Worker (including migrant worker)	142 (8.2)
Businessman	50 (2.9)
Student	21 (1.2)
Civil servant/cadre	14 (0.8)
Retired worker	20 (1.1)
Other	70 (4.0)
Unemployed	63 (3.6)
Ways to fill in the scale – n (%)
Fill by oneself	322 (18.5)
Interview	1420 (81.5)
CD4 cell count – mean (SD)	411.5 (235.8)

Notes: A total of 1742 PLWH were involved in our study. All values, except for age and CD4 (mean ± standard deviation), are expressed as the number of participants, with the percentage in parentheses.

Content validity

Differences between each item regarding the dimension correlation and total score coefficient test were statistically significant (P < 0.01), and each item score was positively correlated with the total score. The correlation coefficient value of each item score and total score was 0.536–0.660 (Table 2), indicating that the scale had favorable content validity.

Table 2.

HIV-ASES content and construct validity analysis table.

Item	Integration	Loading in integration	Perseverance	Loading in perseverance	Common factor variance	Correlation coefficient
1. Stick to your treatment plan even when side effects begin to interfere with daily activities?	0.693	0.480	0.236	0.056	0.536	0.582
2. Integrate your treatment into your daily routine?	0.595	0.354	0.378	0.143	0.496	0.536
4. Stick to your treatment schedule even when your daily routine is disrupted?	0.828	0.686	0.178	0.032	0.718	0.621
5. Stick to your treatment schedule when you aren’t feeling well?	0.756	0.572	0.262	0.069	0.641	0.632
6. Stick to your treatment schedule when it means changing your eating habits?	0.735	0.540	0.168	0.028	0.568	0.604
7. Continue with your treatment even if doing so interferes with your daily activities?	0.793	0.629	0.291	0.085	0.713	0.629
3. Integrate your treatment into your daily routine even if it means taking medication or doing other things in front of people who don’t know you are HIV infected?	−0.055	0.003	0.577	0.333	0.336	0.660
10. Continue with your treatment even when getting to your clinic appointments is a major hassle?	0.456	0.208	0.617	0.381	0.589	0.576
11. Continue with your treatment even when people close to you tell you that they don’t think that it is doing any good?	0.328	0.108	0.681	0.464	0.572	0.602
12. Get something positive out of your participation in treatment, even if the medication you are taking does not improve your health?	0.360	0.130	0.719	0.517	0.646	0.633
8. Continue with the treatment plan your physician prescribed even if your T-cells drop significantly in the next three months?	0.439	0.193	0.600	0.360	0.552	0.616
9. Continue with your treatment even when you are feeling discouraged about your health?	0.411	0.169	0.682	0.465	0.633	0.652

HIV-ASES: HIV Treatment Adherence Self-Efficacy Scale.

Construct validity

EFA

Bartlett test results revealed that the correlation matrix in the samples had common factors (χ² = 5295.612, P < 0.01); the KMO was 0.909, indicating that the samples were highly suitable for factor analysis. EFA with varimax rotation confirmed the presence of the two factors, with eigenvalues higher than 1.00 and an accumulated contribution rate of 58.36%. The Factor 1 (eigenvalue = 5.93) comprised six items and accounted for 33.92% of the variance in measuring the participants’ integration of treatment into their daily lives. Factor 2 (eigenvalue = 1.07) also comprised six items accounting for 24.43% of the variance in measuring the participants’ treatment perseverance when faced with life difficulties. The factor loadings of all items were within 0.577–0.828. HIV-ASES item-to-total score correlations and standardized factor loadings analysis are shown in Table 2.

CFA

Figure 1 shows the path model of the CFA. The two-factor model revealed in the EFA had an adequate data fit. The fit indices, namely χ²(N = 871, df = 53) = 297.207, χ²/df = 5.608, P < 0.001, CFI = 0.950, RMSEA = 0.074, GFI = 0.944, AGFI = 0.918, PGFI = 0.641, IFI = 0.950, and SRMR = 0.037, were all adequate in accordance with the cutoff criteria; however, RMR = 0.125 > 0.05 was inadequate. The overall fit of the model was satisfactory. Table 3 presents the standardized factor loadings between EFA and CFA.

Table 3.

HIV-ASES standardized factor loadings.

Factor	EFA	CFA
Integration
1	0.693	0.676
2	0.595	0.707
4	0.828	0.811
5	0.756	0.725
6	0.735	0.721
7	0.793	0.826
Perseverance
3	0.577	0.276
8	0.600	0.703
9	0.682	0.769
11	0.681	0.678
12	0.719	0.786
10	0.617	0.687

CFA: confirmatory factor analysis; EFA: exploratory factor analysis; HIV-ASES: HIV Treatment Adherence Self-Efficacy Scale.

Convergent validity

The correlations between the HIV-ASES and the Morisky scale, CPSS, SSRS, one-month adherence rate, and CPCRA Adherence Self-Report questionnaire are presented in Table 4. The findings revealed that the two dimensions of the HIV-ASES, namely integration and perseverance, were positively correlated with the self-reported one-month adherence rate, social support, and CPCRA Adherence Self-Report questionnaire, respectively. The correlation coefficients of integration and self-reported one-month adherence rate, SSRS, CPSS, Morisky scale, and CPCRA Adherence Self-Report questionnaire were 0.292, 0.062, −0.096, 0.371, 0.325, respectively, all P < 0.05. The correlation coefficients of perseverance and self-reported one-month adherence rate, SSRS, CPSS, Morisky scale, CPCRA Adherence Self-Report questionnaire were 0.145, 0.126, −0.087, 0.414, 0.249, respectively, all P < 0.05. Integration and perseverance in the HIV-ASES were both negatively correlated with perceived stress, indicating that higher perceived stress is associated with lower adherence self-efficacy.

Table 4.

Convergent validity of HIV-ASES.

Validation scale	Adherence self-efficacy scale factors
Validation scale	Integration	Perseverance
Adherence: one month	0.292**	0.145**
Social support rating scale	0.062*	0.126**
Chinese perceived stress scale	−0.096**	−0.087**
Morisky scale	0.371**	0.414**
CPCRA Adherence Self-Report questionnaire	0.325**	0.249**

CPCRA: Community Programs for Clinical Research on AIDS Antiretroviral Medication Self-Report; HIV-ASES: HIV Treatment Adherence Self-Efficacy Scale.

Note: *P < 0.05; **P < 0.01.

Reliability

The internal consistency Cronbach α coefficient of the full scale was 0.876 (95% CI: 0.868-0.885). The Cronbach α values of the interview and self-administered questionnaire were 0.857 (95% CI: 0.846–0.868) and 0.919 (95% CI: 0.905–0.931), respectively. The α coefficients of the dimensions of integration (factor 1) and perseverance (factor 2) were 0.884 (95% CI: 0.876–0.892) and 0.702 (95% CI: 0.679–0.723), respectively. We also measured the test–retest reliability at three and 12 months, which were, respectively, 0.768 (95% CI: 0.758–0.792) and 0.464 (95% CI: 0.459–0.477). The reliability of the scale was 0.912 after item 3 was omitted.

Discussion

This study focused on the translation, cross-cultural adaptation, and assessment of the reliability and validity of the HIV-ASES Chinese Version-2015 among Chinese PLWHIV, providing strong support for the evaluation of this translated scale and confirming that it can be applied in assessing the adherence self-efficacy of PLWHIV in mainland China. Targeted interventions for low-scoring items can be implemented to improve medication adherence.

The overall alpha coefficient of the scale, alpha coefficients of the two dimensions, and three-month test–retest reliability were all higher than 0.7, showing the reliability of the scale satisfactory. However, the common factor variance of item 3 was low, and the reliability of the scale was higher after this item was omitted. This may be attributable to differences between China and the West, such as differences in economic conditions, customs, habits, and cultures. Some farmers in rural areas of China are migrant workers or have experience as migrant workers, and unstable working conditions often expose them to new environments. Therefore, psychological stress among these PLWHIV is more apparent when they meet strangers; some PLWHIV fear that diseases exposure disclosing their condition will cost them their jobs.³⁷ In addition, because of the influence of the traditional Chinese Confucian collectivist culture, people are often reluctant to reveal personal information to strangers. This suggests that social groups still discriminate against PLWHIV, which affects the self-efficacy of these participants. Therefore, society should strengthen support for PLWHIV, which could improve their self-efficacy and medication adherence.¹¹ Item 3 is an expression of the degree of discrimination against PLWHIV; if the item is more directly and clearly presented, it may be more beneficial for the Chinese population. Therefore, before a large-scale application of the HIV-ASES, item 3 should be replaced with statements with a more appropriate form and content.

In this study, the χ² value increased with the increasing of sample size; therefore, the model can easily be rejected. Therefore, Marsh et al. proposed using the chi square criterion of N < 1000.³⁸ This study calculated the χ² value without further analysis. The common factor variance of all items in the questionnaire was higher than 0.4, except item 3, indicating that the validity of the scale is generally high. During factor analysis, two factors emerged, namely integration and perseverance, which is consistent with the original scale results. However, in the factor attribution of individual items, the results were inconsistent with the original English scale analysis.²² In the original English scale, items 8, 9, and 12 are attributed to perseverance; however, in our study, 3, 8, 9, 10, 11, and 12 were all attributed to perseverance. Items 3, 10, and 11 were, respectively, as follows: ‘Integrate your treatment into your daily routine even if it means taking medication or doing other things in front of people who don’t know you are HIV-infected?’ ‘Continue with your treatment even when getting to clinic appointments is a major hassle?,’ and ‘people close to you tell you that they don’t think that it is doing any good?’ These three items described circumstances in which people close to PLWHIV or in the external environment may view disease treatment as unfavorable determining participants’ confidence in treatment; they are reasonable according to Chinese cultural understanding. The convergent validity of the questionnaire relative to other related questionnaires was moderate. The HIV-ASES scores were not correlated with CD4 cell count (P > 0.05); therefore, in accordance with previous studies,^39,40 it was determined that the life quality of PLWHIV is not necessarily correlated with CD4 cell count. The present study sample only included people from the Chinese province of Henan, and the main factors influencing the scale comprised the time of starting ART, route of HIV infection, and population culture. The employment of participants from only one province did not considerably affect the outcome of the scale. In brief, the translated scale is comparable with the original scale.

The interview questionnaires may have resulted in selection bias and reporting bias, and interviewers had received intensive training to avoid bias. The reliability of interviewed questionnaires is 0.857 and satisfactory. It indicates that interviewing can also be used when PLWHIV cannot fill it by themselves because of limitations in education, age, physical condition, and other factors.

The study has some limitations. First, some data in the study depended on participants self-reporting, such as the one-month medication adherence rate and CPCRA Adherence Self-Report questionnaire. If an electronic drug detection system was used for data recording, the results could have been more objective and the effects more favorable. Second, this study had a cross-sectional design, which measured participants’ adherence within a certain time point; however, the adherence level varies with time and other factors.⁴¹ Finally, the geographical distribution of samples was not diversified.

Conclusion

The HIV-ASES Chinese Version-2015 is a reliable and valid instrument for measuring the medication adherence rate of PLWHIV. The present findings provide a foundation for future studies on HIV/AIDS medication adherence and estimating PLWHIV adherence quantitatively. Also, it can play a predictive role in preventing nonadherence to ART. Prospective studies can be conducted to investigate items and PLWHIV with low adherence to improve their medication adherence rate and the efficiency of HIV follow-up studies.

Footnotes

Acknowledgments

We are grateful to the staff of the Sanmenxia, Queshan, Shangcai, Shenqiu, Jiyuan and Yongcheng Centers for Disease Control and Prevention for participating in the study and to all the study participants for willingly providing the study data. We also thank all of the translators.

Authors’ contribution

LS and S-MY contributed equally to this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by national ‘12th Five Year Plan’ science and technology project (Grant number: 2012ZX10004905-004-002) and foundation and advanced technology research project in Henan province (Grant number: 132300410271).

Appendix 1. HIV-ASES Items (section from Johnson et al. 22 )

References

World Health Organization. http://www.who.int/mediacentre/factsheets/fs360/en/ (2015, accessed 7 June 2016).

National Health and Family Planning Commission of the PRC. China reports over 100,000 new diagnosed HIV/AIDS cases in 2014, http://english.nhfpc.gov.cn/2015-01/20/c_45502.htm (2015, accessed 7 June 2016).

Ministry of Health of The People’s Republic of China: 2015 China AIDS Response Progress Report, http://www.unaids.org/sites/default/files/country/documents/CHN_narrative_report_2015.pdf (2015, accessed 5 February 2016).

Chen

Wantland

Reid

. Engagement with health care providers affects self- efficacy, self-esteem, medication adherence and quality of life in people living with HIV. J AIDS Clin Res 2013; 4: 256–256.

Mhaskar

Alandikar

Emmanuel

. Adherence to antiretroviral therapy in India: a systematic review and meta-analysis. Indian J Commun Med 2013; 38: 74–82.

Johnson

Dilworth

Taylor

. Improving coping skills for self-management of treatment side effects can reduce antiretroviral medication nonadherence among people living with HIV. Ann Behav Med 2011; 41: 83–91.

Chen

Shiu

Yang

. Antiretroviral therapy (ART) side effect impacted on quality of life, and depressive symptomatology: a mixed-method study. J AIDS Clin Res 2013; 4: 218–218.

Nokes

Johnson

Webel

. Focus on increasing treatment self-efficacy to improve human immunodeficiency virus treatment adherence. J Nurs Scholarship 2012; 44: 403–410.

Adefolalu

Nkosi

Olorunju

. Self-efficacy, medication beliefs and adherence to antiretroviral therapy by patients attending a health facility in Pretoria. South Afr Fam Pract 2015; 56: 281–285.

10.

Ingersoll

Cohen

. The impact of medication regimen factors on adherence to chronic treatment: a review of literature. J Behav Med 2008; 31: 213–224.

11.

Bangsberg

Acosta

Gupta

. Adherence-resistance relationships for protease and non-nucleoside reverse transcriptase inhibitors explained by virological fitness. Aids 2006; 20: 223–231.

12.

Erlen

Cha

Kim

. The HIV medication taking self-efficacy scale: psychometric evaluation. J Adv Nurs 2010; 66: 2560–2572.

13.

Bandura

. Self-efficacy: toward a unifying theory of behavioral change. Psychol Rev 1977; 84: 191–215.

14.

Johnson

Chesney

Goldstein

. Positive provider interactions, adherence self-efficacy, and adherence to antiretroviral medications among HIV-infected adults: a mediation model. AIDS Patient Care STDs 2006; 20: 258–268.

15.

Simoni

Frick

Huang

. A longitudinal evaluation of a social support model of medication adherence among HIV-positive men and women on antiretroviral therapy. Health Psychol 2006; 25: 74–81.

16.

Huang

Zhang

. Self-efficacy, medication adherence, and quality of life among people living with HIV in Hunan Province of China: a questionnaire survey. J Assoc Nurses AIDS Care 2013; 24: 145–153.

17.

Shively

Smith

Bormann

. Evaluating self-efficacy for HIV disease management skills. AIDS Behav 2002; 6: 371–379.

18.

Zhang

Lin

. Side effects, adherence self-efficacy, and adherence to antiretroviral treatment: a mediation analysis in a Chinese sample. AIDS Care 2016; 28: 919–926.

19.

Morisky

Green

Levine

. Concurrent and predictive validity of a self-reported measure of medication adherence. Med Care 1986; 24: 67–74.

20.

Trimble

Nava

McFarlane

. Intimate partner violence and antiretroviral adherence among women receiving care in an urban Southeastern Texas HIV clinic. J Assoc Nurses AIDS Care 2013; 24: 331–340.

21.

Cha

Erlen

Kim

. Mediating roles of medication-taking self-efficacy and depressive symptoms on self-reported medication adherence in persons with HIV: a questionnaire survey. Int J Nurs Stud 2008; 45: 1175–1184.

22.

Johnson

Neilands

Dilworth

. The role of self-efficacy in HIV treatment adherence: validation of the HIV Treatment Adherence Self-Efficacy Scale (HIV-ASES). J Behav Med 2007; 30: 359–370.

23.

Guillemin

Bombardier

Beaton

. Cross-cultural adaptation of health-related quality of life measures: literature review and proposed guidelines. J Clin Epidemiol 1993; 46: 1417–1432.

24.

Walsh

Pozniak

Nelson

. Virologic rebound on HAART in the context of low treatment adherence is associated with a low prevalence of antiretroviral drug resistance. J Acquir Immune Defic Syndr 2002; 30: 278–287.

25.

Johnson

Dilworth

Taylor

. Primary relationships, HIV treatment adherence, and virologic control. AIDS Behav 2012; 16: 1511–1521.

26.

Mannheimer

Friedland

Matts

. The consistency of adherence to antiretroviral therapy predicts biologic outcomes for human immunodeficiency virus-infected persons in clinical trials. Clin Infect Dis 2002; 34: 1115–1121.

27.

Wang

Zhou

Huang

. Effects of nurse-delivered home visits combined with telephone calls on medication adherence and quality of life in HIV-infected heroin users in Hunan of China. J Clin Nurs 2010; 19: 380–388.

28.

Wang

. Self-Reported adherence to antiretroviral treatment among HIV-infected people in Central China. AIDS Patient Care STDs 2008; 22: 71–80.

29.

Paterson

Swindells

Mohr

. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med 2000; 133: 21–30.

30.

Gebo

Keruly

Moore

. Association of social stress, illicit drug use, and health beliefs with nonadherence to antiretroviral therapy. J Gen Intern Med 2003; 18: 104–111.

31.

Bangsberg

Perry

Charlebois

. Non-adherence to highly active antiretroviral therapy predicts progression to AIDS. AIDS 2001; 15: 1181–1183.

32.

Bangsberg

Hecht

Clague

. Provider assessment of adherence to HIV antiretroviral therapy. J Acquir Immune Defic Syndr 2001; 26: 435–442.

33.

Xiao

. Social support rating scale. Chin Ment Health J 1999, pp. 127–131.

34.

Liu

. Social support and quality of life: a cross-sectional study on survivors eight months after the 2008 Wenchuan earthquake. BMC Public Health 2010; 10: 573–573.

35.

Cohen

Kamarck

Mermelstein

. A global measure of perceived stress. J Health Soc Behav 1983; 24: 385–396.

36.

Glei

Goldman

Shkolnikov

. Perceived stress and biological risk: is the link stronger in Russians than in Taiwanese and Americans? Stress 2013; 16: 411–420.

37.

Corrigan

Tsang

Shi

. Chinese and American employers’ perspectives regarding hiring people with behaviorally driven health conditions: the role of stigma. Soc Sci Med 2010; 71: 2162–2169.

38.

Marsh

Wen

Hau

. Structural equation models of latent interactions: evaluation of alternative estimation strategies and indicator construction. Psychol Methods 2004; 9: 275–300.

39.

Lau

Tsui

Patrick

. Validation of a Chinese version of the Medical Outcomes Study HIV Health Survey (MOS-HIV) among Chinese people living with HIV/AIDS in Hong Kong. Qual Life Res 2006; 15: 1079–1089.

40.

Scott-Lennox

Boyer

. Reliability and validity of French, German, Italian, Dutch, and UK English translations of the Medical Outcomes Study HIV Health Survey. Med Care 1999; 37: 908–925.

41.

Wasti

Simkhada

Randall

. Factors influencing adherence to antiretroviral treatment in Nepal: a mixed-methods study. PLoS One 2012; 7: e35547–e35547.