Back to Journals » Patient Preference and Adherence » Volume 14

Psychometric Evaluation of Chinese Version of Adherence to Refills and Medications Scale (ARMS) and Blood-Pressure Control Among Elderly with Hypertension

Authors Chen YJ, Chang J, Yang SY

Received 27 October 2019

Accepted for publication 25 December 2019

Published 10 February 2020 Volume 2020:14 Pages 213—220

DOI https://doi.org/10.2147/PPA.S236268

Checked for plagiarism Yes

Review by Single anonymous peer review

Peer reviewer comments 2

Editor who approved publication: Dr Naifeng Liu



Yi-Jing Chen,1,2 Jing Chang,1 Si-Yu Yang3

1Department of Health Care Clinic, First Affiliated Hospital of Zhengzhou University, Henan, People’s Republic of China; 2Academy of Medical Science, Zhengzhou University, Henan, People’s Republic of China; 3Xiangya Medical School, Central South University, Changsha, People’s Republic of China

Correspondence: Jing Chang Email [email protected]

Objective: This study aimed to develop the ARMS-C and test its psychometric properties in hypertensive patients, to assess the level of medication adherence and to identify associated predictors for medication adherence and blood-pressure control among Chinese hypertensive patients.
Methods: Hypertensive elderly who met inclusion criteria were recruited from an aged-care facility in Henan Province between January 2019 and July 2019. The patients completed the adapted ARMS-C. The scale’s factor structure, internal consistency, and construct validity were tested.
Results: Good internal consistency (Cronbach’s α=0.89) and test–retest reliability (r=0.86, p< 0.01) were obtained. Item-total correlation coefficients for the ten-item ARMS-C were 0.505– 0.801. Factor analysis of construct validity identified two factors, explaining a total variance of 63.3%. Binary regression showed that patients with scores at level 2 (≤ 20 ARMS-10 scores < 30) were six times as likely to have blood pressure uncontrolled as those at level 1 (ARMS-10 scores < 20, OR 6.6, 95% CI 1.7– 25.1; p=0.006), and patients with scores at level 3 (ARMS-10 scores > 30) were 115 times as likely to have blood pressure uncontrolled as those at level 1 (ARMS-10 scores < 20,OR 115, 95% CI 9– 1,470; p=0).
Conclusion: The ten-item ARMS-C is a reliable and valid self-reporting screening tool for adherence to medication and refills in elderly hypertensive Chinese patients.

Keywords: adherence, 10-item ARMS, hypertension, blood-pressure control

Background

Hypertension is a common disease in the elderly, and its prevalence increases with age.1 It has been calculated that people aged 60 years and over will reach a third of the hypertensive population in Europe and North America in 20302 and that nearly half of Chinese adults have hypertension at 35–75 years of age.3 Hypertension is a key factor in developing cardiovascular disease,4 and uncontrolled blood pressure (BP) has been observed to be associated with cognitive decline, stroke, renal failure, increased risk of cardiovascular events, and mortality.5 The benefit of BP lowering has already been proved, and it has been reported that a reduction of 10 mmHg in systolic BP can lower the risk of cardiovascular disease by as much as 20%, stroke by 27%, heart failure by 28%, and all-cause mortality by up to 13%.6

Pharmacotherapy with lifestyle change remains the main component in hypertension management.7,8 A China-based screening study revealed that less than a third of hypertensive patients were being treated and fewer than one in twelve in control of their BP.3 Medication nonadherence is the leading cause of inadequate hypertension management and the major patient-related cause of inadequate BP control.9 Improving medication adherence will increase treatment effectiveness and decrease the burden of health-care usage.10 Identifying adherence-related behavior is a crucial step in improving medication intake and patient education,11 and more targeted and specific measures can be developed through this process.

The Adherence to Refills and Medications Scale (ARMS) is a valid and reliable medication-adherence scale specially designed for the chronic-disease population. It is highly correlated with the most commonly used self-reported measure (the Morisky scale) and measures of refill adherence.12 Currently, the ARMS is available in multiple languages and is used among patients with chronic illness, including diabetes, coronary heart disease, breast cancer, and hypertension.1316 However, all studies have been conducted in hospitals, and none have been applied to the elderly in the daily practice setting, especially for the elderly who live alone with a high risk of medication nonadherence.17 Therefore, this study aimed to validate a Chinese translation and culturally adapted version of the ARMS-10 and to evaluate its psychometric properties among elderly hypertensive Chinese patients. Secondary objectives were to identify demographic and clinical factors that affect levels of medication adherence and BP control.

Methods

Participants

This was a randomized trial conducted in an aged-care facility in Zhengzhou, China. The study was performed between January 2019 and July 2019. Sociodemographic data were obtained from the facility. Patients who met the criteria were invited to participate. After that, questionnaires were administered and collected by staff working there without access to data. Oral explanations and handwriting assistance were offered by these staff. Participants were required to finish the questionnaires after they had signed a written consent form. Participation in the study was voluntary, and patients were free to withdraw from the study at any stage.

Inclusion and Exclusion Criteria

Patients included were aged 65 years or above, had been treated with at least one antihypertensive medicine for 6 months or more, and had been diagnosed with stage I or stage II hypertension (140–159/90–99 mmHg or 160–180/100–110 mmHg, respectively) by a professional physician. Patients who were conscious and able to communicate, took medications by themselves, and were willing to participate in the investigation were included. Exclusion criteria were accelerated hypertension, secondary hypertension, dementia, hemorrhagic stroke in the previous 6 months, estimated life expectancy of <6 months, and a requirement for daily nursing care. For the purposes of this study, the goal of controlled BP was defined in accordance with the 2011 NICE guideline for hypertension in adult patients. Patients’ BP was considered controlled if those <80 years old with treated hypertension had BP <140/90 mmHg, those <80 years old with treated hypertension had BP <150/90 mmHg, or those with both hypertension and diabetes mellitus had BP <140/80 mmHg or <130/80 mmHg (in cases of any kidney, eye, or cerebrovascular damage).18

Sample Size

In accordance with the recommendation of five to ten participants per questionnaire item for assurance of accuracy19 in estimating model parameters, a minimum sample of 100 was needed for the ten‐item ARMS. A total of 120 participants were selected to participate in this study eventually to allow for a maximum of 20% of missing or incomplete questionnaires.

Ethical Considerations

Ethical approval was obtained from the First Affiliated Hospital of Zhengzhou University (approval 2019-KY-143). The study was performed in accordance with the Helsinki Declaration and the principles of good clinical practice, with respect for the rights and dignity of the participants. All participants provided written informed consent.

Questionnaire

The ARMS was developed by Kripalani et al, and its initial version was developed based on the Morisky and Hill–Bone questionnaires. The 12 items in the original questionnaire comprise two subscales: adherence to taking medications (eight items) and adherence to refilling prescriptions (four items). Each item is structured for response on a Likert scale, with responses of “none”, “some”, “most”, or “all the time”, which are given values of 1–4. Item scores are summed to produce an overall adherence score of 10–40, with lower scores indicating better adherence.12

Adaptation of the Scale

The original 12-item ARMS has already been translated into Chinese by Emory University, and both the original English version and translated Chinese version of the instrument (ARMS-C) were deemed equally important for verifying that scale items in the two languages closely resembled each other.12 A pilot study was conducted to evaluate patients’ understanding of the ARMS-C, and 20 hypertensive elderly subjects were included from an aged-care facility in Henan Province, China. The results were reflected in the final version of the questionnaire and some modifications made, given the outpatient nature of the study and target population.20 Items asking about refill medicine were adjusted: according to our investigation and patient feedback, a doctor’s prescription is not required to refill hypertensive medicine and patients can easily get the medicine at a pharmacy with medical insurance. As such, we deleted items 4 and 11 from the original ARMS-12 after obtaining permission from the original author. The final version of the ARMS-10 was used for psychometric testing among elderly hypertensive Chinese patients in this study. A total of ten items in AMRS-C comprised two subscales: adherence to taking medication (seven items) and adherence to refilling medication (three items). Each item was structured for response on a Likert scale, with responses of “none”, “some”, “most”, or “all the time”, which were given values of 1–4. Item scores were summed to produce an overall adherence score of 10–40. Most items were written so that a lower score indicated better adherence.12

Statistical Analysis

Statistical analysis was conducted with SPSS 24. Differences in demographic data and outcome variables were assessed using frequencies and percentages. Internal consistency was assessed by Cronbach's α-coefficient and item-total correlation.19 In general, items with a correlation of no less than 0.3 on the total scale are considered conceptually similar, and an α-value no less than 0.7 is considered to indicate adequate internal consistency.21 The structure validity of the questionnaire was analyzed by confirmatory factor analysis with a varimax rotation–component matrix. Test–test reliability was conducted by calculating Spearman’s correlation coefficient between the baseline test and retest after 10 days. Correlations of 0.7–0.9 were considered very good. Goodness-of-fit criteria, including χ2:df ratio, root mean–square error of approximation, and goodness-of-fit index, adjusted goodness-of-fit index, and comparative fit index, were used to determine the overall data-model fitness.22 Binary logistic regression analysis using the enter method was conducted to determine factors that could predict medication adherence and BP control. Correlation and Hosmer–Lemeshow goodness-of-fit tests were done to select the best prediction model.23

Results

Participant Characteristics

A total of 120 patients were included in the study: 55.8% female with a mean age of 81.9±5.2 years. The majority of the study population were living alone (93.3%) with an elementary education level (43.3%). Demographic and medication-related characteristics of the subjects are shown in Table 1.

Table 1 Demographic and Clinical Characteristics of Subjects

Internal Consistency and Reliability

Table 2 shows reliability-analysis results for the ARMS-C questionnaire. Item-total correlation coefficients for the ARMS-C were 0.505–0.801, indicating moderate–strong correlation. Cronbach's α for the ten-item scale was 0.89, which demonstrated good internal consistency.

Table 2 Reliability Analysis for ARMS-10 Items in the Study Group (n=120)

Validity Analysis

Table 3 shows the results of factor analysis. The Kaiser–Meyer–Olkin test of sampling adequacy and Bartlett’s test of sphericity were performed before exploratory factor analysis. Results were 0.79 for the Kaiser–Meyer–Olkin test, while Bartlett’s test was significant (χ2=707.3, p<0.001). Through principal-component analysis, two factors were identified, with a total variance of 63.3%. Factor 1 had an an eigenvalue of 1.91 and explained 37.59% of the variance. Factor 2 had an eigenvalue of 1.31 and explained 25.72% of the variance. The rotated factor analysis varied between 0.57 and 0.86. Test–retest reliability of the Chinese ARMS-10 was good at 10 days, with an ICC of 0.86 (p<0).

Table 3 Factor Analysis (Varimax-Rotated Component Matrix) for ARMS in the Study Group (n=120)

Risk Factors for Adherence Levels

High levels of adherence were present in 14.3% of the population. Patient sociodemographic characteristics were divided into three adherence categories: level 1 (ARMS-10 scores <20), level 2 (ARMS-10 scores 20–<30), and level 3 (ARMS-10 scores ≥30). Table 4 displays the results of binary logistic regression analysis assessing factors related to medication adherence. This model explained 17% of the change in levels of adherence. No variable was found to be an independent predicator of medication adherence (p>0.05).

Table 4 Binary Logistic Regression Analysis Identifying Factors for ARMS Score

Blood-Pressure Control

BP was controlled in 33.3% of the study participants. Table 5 shows the results of binary logistic regression analysis identifying factors predicting BP control. About 35% of the change in BP and 73% accuracy of classification was explained by this model. Patients with ARMS-10 scores at level 2 were six times as likely to have BP uncontrolled as those at level 1 (OR 6.6, 95% CI 1.7–25.1; p=0.006), and patients with ARMS-10 scores at level 3 were 115 times as likely to have BP uncontrolled as level 1 (OR 115, 95% CI 9–1,470; p=0).

Table 5 Binary Logistic Regression Analysis Identifying Factors for BP Control

Discussion

This study focused on the psychometric properties of the ARMS-10 adapted to Chinese and assessed the level of medication adherence among elderly hypertensive patients at an aged-care facility. The ARMS is one of the best self-reported questionnaires measuring medication adherence in elderly hypertensive patients.12 It is crucial for both clinicians and health policy-makers to identify patients with poor medication adherence and BP control so that more specific interventions can be designed and implemented.

This study was conducted on elderly hypertensive patients. It demonstrated that the ten-item ARMS-C questionnaire had good psychometric properties and could be used to assess medication adherence in chronically ill patients as well as the original 12-item ARMS. Psychometric analysis revealed high consistency, reliability, and validity. The internal consistency of the ten-item ARMS-C for all levels of education was 0.89 compared to a Cronbach α of 0.75 in the ARMS-7 and 0.82 for a high level of literacy in the original ARMS-12. Item-total correlation coefficients for the ten-item ARMS-C were 0.505–0.801. On the original scale, item-total correlation coefficients were 0.34–0.59, so the ten-item ARMS-C had better psychometric properties than the original ARMS-12 in English and ARMS-7 in Turkish, regardless of patient education.12,15 Health literacy is an important risk factor affecting adherence to treatment.24 The original ARMS-12 included only patients with low literacy. In this study, nearly half the participants had no education or only primary education, which extends the original work so that it can be applied to patients with all level of literacy.

For the original English version of the ARMS-12, two factors were identified for factor analysis of construct validity. Factor 1 comprised ten items assessing adherence to taking medication with an eigenvalue of 6.67 and accounted for 55.6% of variance. Factor 2 had two items with an eigenvalue of 2.69 and accounted for 22.4% of variance. In the validation of the 10-item ARMS-C, factor 1 (refilling medicine and dosage change) comprised five items, had an eigenvalue of 1.91, and explained 37.59% of the variance. Factor 2 (intentional nonadherence) comprised five items with an eigenvalue of 1.31 and explained 25.72% of the variance. Cronbach's α for factor 1 was 0.90, with an item-total correlation of 0.63–0.82. For factor 2, the item-total correlation was 0.5–0.6 with Cronbach's α of 0.77. Values computed from he ten-item ARMS-C were higher than those from the original version of the ARMS-12 (factor 1, α=0.79; factor 2, α=0.64).

Identifying predictors of poor adherence is crucial for disease management in elderly.25 In this study, the overall percentage of medication adherence in hypertension was less than studies conducted in Saudi Arabia (22.5%)23 and Buea, Cameroon (33.3%).26 One possible explanation for this discrepancy is the existence of cultural differences between China and other countries,27 and different self-reported adherence scales may also account for this.

We found that the majority of participants had poor medication adherence, while age, sex, education level, duration, and number of medications were not found to be correlated with medication adherence. Previous studies have reported inconsistent results between demographic variables and medication adherence,23,2830 while some support age, sex, number of medications, and medical conditions being independent predicators of medication adherence.23,30,31 In line with our study, a study conducted in China also reported that education level was the only predicator of medication adherence,27 while another study reported that age was the only factor that correlated with medication adherence among Chinese patients.32 Age was not statistically significantly related to medication adherence (p=0.05) in this study, and larger samples may be needed to detect an association between age and medication adherence.

Regarding independent predicators for BP control, medication adherence was found to be the only factor related to BP management: patients with higher adherence were more likely to have their BP controlled. Previous studies have reported similar correlations between medication adherence and BP control.12,23,31 Kripalani et al revealed that higher adherence was in line with better diastolic BP control and more optimal overall BP. Therefore, the ten-item ARMS-C questionnaire can be regarded as a reliable and powerful self-report instrument in assessing medication adherence.

There were several limitations in this study. Firstly, it was conducted at just one aged-care facility, which may limit application to other settings. Secondly, only one questionnaire was used, so there is a lack of comparison between self-report and other measures, which may affect the reliability of our findings. Thirdly, only patients with hypertension were included, which may restrict the generalizability of these results. Future research should apply these scales to more diverse chronically ill patients.

Conclusion

The results of this study proved the used of the ten-item ARMS-C in Chinese population, which is highly recommended to hypertensive patients as a reliable screening tool for adherence to medication and refills. Future studies are need in Chinese population with other chronic disease, besides, more lager samples are required and its ability to identify predicators that might associate with enhanced adherence.

Acknowledgments

The authors would like to thank the authors of the original ARMS questionnaire for their kind permission, and all the patients in this study.

Disclosure

The authors report no conflicts of interest in this work.

References

1. Cohen DL, Townsend RR. Update on pathophysiology and treatment of hypertension in the elderly. Curr Hypertens Rep. 2011;13(5):330–337. doi:10.1007/s11906-011-0215-x

2. Pol LG. Rapid growth in the elderly population of the world. In: Berhouma M, Krolak-Salmon P, editors. Brain and Spine Surgery in the Elderly. Cham: Springer; 2017:3–15

3. Lu J, Lu Y, Wang X, et al. Prevalence, awareness, treatment, and control of hypertension in China: data from 1.7 million adults in a population-based screening study (China PEACE million persons project). Lancet. 2017;390(10112):2549–2558. doi:10.1016/S0140-6736(17)32478-9

4. Del PR, Ferri C. Hypertension management at older age: an update. High Blood Press Cardiovasc Prev. 2019;26(1):27–36. doi:10.1007/s40292-018-0290-z

5. Pengo MF, Rossitto G, Bisogni V, et al. Systolic and diastolic short-term blood pressure variability and its determinants in patients with controlled and uncontrolled hypertension: a retrospective cohort study. Blood Press. 2015;24(2):124–129. doi:10.3109/08037051.2014.992187

6. Ettehad D, Emdin CA, Kiran A, et al. Blood pressure lowering for prevention of cardiovascular disease and death: a systematic review and meta-analysis. Lancet. 2016;387(10022):957–967. doi:10.1016/S0140-6736(15)01225-8

7. James PA, Oparil S, Carter BL, et al. 2014 evidence-based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA. 2014;311(5):507–520. doi:10.1001/jama.2013.284427

8. Frisoli TM, Schmieder RE, Grodzicki T, et al. Beyond salt: lifestyle modifications and blood pressure. Eur Heart J. 2011;32(24):3081–3087. doi:10.1093/eurheartj/ehr379

9. Ferdinand KC, Yadav K, Nasser SA, et al. Disparities in hypertension and cardiovascular disease in blacks: the critical role of medication adherence. J Clin Hypertens(Greenwich). 2017;19(10):1015–1024. doi:10.1111/jch.13089

10. Bosworth HB, Granger BB, Mendys P, et al. Medication adherence: a call for action. Am Heart J. 2011;162(3):412–424. doi:10.1016/j.ahj.2011.06.007

11. Korb-Savoldelli V, Gillaizeau F, Pouchot J, et al. Validation of a French version of the 8-item morisky medication adherence scale in hypertensive adults. J Clin Hypertens(Greenwich). 2012;14(7):429–434. doi:10.1111/j.1751-7176.2012.00634.x

12. Kripalani S, Risser J, Gatti ME, et al. Development and evaluation of the Adherence to Refills and Medications Scale (ARMS) among low-literacy patients with chronic disease. Value Health.. 2009;12(1):118–123. doi:10.1111/j.1524-4733.2008.00400.

13. Kim CJ, Park E, Schlenk EA, et al. Psychometric evaluation of a Korean version of the Adherence to Refills and Medications Scale (ARMS) in adults with type 2 diabetes. Diabetes Educ. 2016;42(2):188–198. doi:10.1177/0145721716632062

14. Lomper K, Chabowski M, Chudiak A, et al. Psychometric evaluation of the Polish version of the Adherence to Refills and Medications Scale (ARMS) in adults with hypertension. Patient Prefer Adherence. 2018;12:2661–2670. doi:10.2147/PPA.S185305

15. Gokdogan F, Kes D. Validity and reliability of the Turkish Adherence to Refills and Medications Scale. Int J Nurs Pract. 2017;23(5):1–7. doi:10.1111/ijn.2017.23.issue-5

16. Jaam M, Mohamed Ibrahim MI, Kheir N, et al. Assessing prevalence of and barriers to medication adherence in patients with uncontrolled diabetes attending primary healthcare clinics in Qatar. Prim Care Diabetes. 2018;12(2):116–125. doi:10.1016/j.pcd.2017.11.001

17. Park HY, Seo SA, Yoo H, et al. Medication adherence and beliefs about medication in elderly patients living alone with chronic diseases. Patient Prefer Adherence. 2018;12:175–181. doi:10.2147/PPA.S151263

18. National Institute for Health and Care Excellence (NICE) [Internet]. Manchester, UK: Hypertension in adults: diagnosis and management NICE guidelines [CG127]. [reviewed 2013 October; published 2011 August].

19. Devellis R. Scale Development: Theory and Applications. 4th ed. Chapel Hill, USA: SAGE Publications; 2016

20. Covert KL, Mardis CR, Fleming JN, et al. Development of a predictive model for drug-related problems in kidney transplant recipients. Pharmacotherapy. 2017;37(2):159–169. doi:10.1002/phar.2017.37.issue-2

21. Nunnally JC, Bernstein IH. Psychometric Theory. 3rd ed. New York: McGraw-Hill; 1994

22. Byrne BM. Structural equation modeling with LISREL, PRELIS and SIMPLIS: basic concepts, applications, and programming. In: New Jersey: Lawrence Erlbaum Associates. Psychology Press: 1998

23. Khayyat SM, Khayyat SM, Hyat Alhazmi RS, et al. Predictors of medication adherence and blood pressure control among saudi hypertensive patients attending primary care clinics: a cross-sectional study. PLoS One. 2017;12(1):e0171255. doi:10.1371/journal.pone.0171255

24. Gazmararian JA, Kripalani S, Miller MJ, et al. Factors associated with medication refill adherence in cardiovascular-related diseases. J Gen Intern Med. 2006;21(12):1215–1221. doi:10.1111/j.1525-1497.2006.00591.x

25. Nieuwlaat R, Wilczynski N, Navarro T, et al. Interventions for enhancing medication adherence. Cochrane Database Syst Rev. 2014;20(11):CD000011

26. Adidja NM, Agbor VN, Aminde JA, et al. Non-adherence to antihypertensive pharmacotherapy in Buea, Cameroon: a cross-sectional community-based study. BMC Cardiovasc Disord. 2018;18(1):150. doi:10.1186/s12872-018-0888-z

27. Du X, Chen H, Zhuang Y, et al. Medication Adherence in Chinese Patients with systemic lupus erythematosus. J Clin Rheumatol. 2018.

28. Ehret MJ, Shelton D, Barta W, et al. Medication adherence among female inmates with bipolar disorder: results from a randomized controlled trial. Psychol Serv. 2013;10(1):106–114. doi:10.1037/a0031433

29. Mahmoud Manal IH. Compliance with treatment of patients with hypertension in Almadinah Almunawwarah: a community-based study. J Taibah Univ Med Sci. 2012;7(2):92–98

30. AlHewiti A. Adherence to long-term therapies and beliefs about medications. Int J Family Med. 2014;2014:479596. doi:10.1155/2014/479596

31. Boima V, Ademola AD, Odusola AO, et al. Factors associated with medication nonadherence among hypertensives in Ghana and Nigeria. Int J Hypertens. 2015;2015:205716. doi:10.1155/2015/205716

32. Wong MC, Wu CH, Wang HH, et al. Association between the 8-item Morisky medication adherence scale (MMAS-8) score and glycaemia control among Chinese diabetes patients. J Clin Pharmacol. 2015;55(3):279–287. doi:10.1002/jcph.408

Creative Commons License © 2020 The Author(s). This work is published and licensed by Dove Medical Press Limited. The full terms of this license are available at https://www.dovepress.com/terms.php and incorporate the Creative Commons Attribution - Non Commercial (unported, v3.0) License. By accessing the work you hereby accept the Terms. Non-commercial uses of the work are permitted without any further permission from Dove Medical Press Limited, provided the work is properly attributed. For permission for commercial use of this work, please see paragraphs 4.2 and 5 of our Terms.