A New Lebanese Medication Adherence Scale: Validation in Lebanese Hypertensive Adults

Effect of psychosocial aspects on medication adherence in patients with heart failure amid socioeconomic challenges

OBJECTIVE

To evaluate stress, depression and quality of life among community-dwelling patients with heart failure (HF) and evaluate their effect on perceived medication adherence in a socioeconomically challenged setting.

DESIGN

A cross-sectional design with self-administered questionnaire with data collected between October 2021 and September 2022.

METHODS

Patients with confirmed diagnosis of HF were sought for data collection in the community and cardiology clinics through an electronic platform. Confirmation of cases was done through the ejection fraction, medication list and frequent symptoms of the patients. The Patient Health Questionnaire-9, the COVID-19 Stress Scale, the Minnesota Living with HF Questionnaire and the Lebanese Medication Adherence Scale were used to evaluate depression, stress, quality of life and medication adherence, respectively. Univariate analysis was done to present the descriptive statistics, whereas bivariate and multivariate analyses were done to evaluate the relationship between the variables.

RESULTS

A total of 237 participants were included in the final analysis. The mean age was 61.3±17.36 years, and the majority (57.8%) were male participants. Only 44.7% were on ACE inhibitors/angiotensin receptor blockers and 54.9% on beta-blockers. The mean scores for stress, depression, quality of life and medication adherence were 75.86 (SD=24.5), 14.03 (SD=5.7), 55.73 (SD=23.05) and 6.79 (SD=6.93), respectively, indicating high stress levels, depression, poor quality of life and medication adherence. Those with a history of hypertension and depression were significantly more adherent to their medications than those who were not. Multivariate analysis showed that anxiety, medical follow-up, quality of life and functionality class were predictors of medication adherence.

CONCLUSION

The study showed the population with HF in Lebanon to have psychological health problems with these variables acting as predictors for medication adherence. Sociodemographic characteristics also played a role on the outcome, which can be targeted when planning interventions to improve outcomes. Future studies should compare prescribed medication with consumed medication through longitudinal approaches and medical refilling techniques when possible.

Adherence to treatment and harmful effects of medication shortages in the context of severe crises: scale validation and correlates

BACKGROUND

Medication shortage is a public health problem, affecting patients' outcomes mainly through the difficulty in maintaining adherence, particularly in the context of a severe economic crisis. There is a need for a new scale that assesses the effect of medication shortage on adherence.

AIM

To develop and validate a scale to evaluate the harmful impact of medication shortage among the general Lebanese population and assess its correlates and association with medication adherence.

METHODS

A questionnaire was used to assess medication shortage harmful effects and patients' adherence, allowing to generate the Harmful Impact of Medication Shortage scale (HIMS). The factor analysis, convergent validity and reliability of the generated scale were assessed, followed by multivariable regressions to evaluate its correlates.

RESULTS

The developed HIMS scale is a 9-item tool, used to assess how difficult it was for people to deal with medication shortages and their harmful effects on treatment. It was significantly and inversely linked to treatment adherence and affected by the patients' socioeconomic status and the type of chronic disease.

CONCLUSION

The Harmful Impact of Medication Shortage scale could be an efficient tool to measure the detrimental effects of medication shortages among the Lebanese adult population with chronic diseases, particularly affecting treatment adherence. Future studies and evidence are still needed to confirm our findings and help build global mitigation policies addressing medication shortages.

Quality of measurement properties of medication adherence instruments in cardiovascular diseases and type 2 diabetes mellitus: a systematic review and meta-analysis

BACKGROUND

Medication adherence has a major impact on reducing mortality and healthcare costs related to the treatment of cardiovascular diseases and diabetes mellitus. Selecting the best patient-reported outcome measure (PROM) among the many available for this kind of patient is extremely important. This study aims to critically assess, compare and synthesize the quality of the measurement properties of patient-reported outcome measures to assess medication adherence among patients with cardiovascular diseases and/or type 2 diabetes mellitus.

METHODS

This review followed the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) guidelines and was reported according to the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA). The searches were performed in Web of Science, SCOPUS, PubMed, CINAHL, EMBASE, LILACS, PsycINFO, and ProQuest (gray literature).

RESULTS

A total of 110 records encompassing 27 different PROMs were included in the review. The included records were published between 1986 and 2023, most of which reported studies conducted in the United States and were published in English. None of the PROMs were classified in the category "a", thus being recommended for use due to the quality of its measurement properties. The PROMs that should not be recommended for use (category "c") are the MTA, GMAS, DMAS-7, MALMAS, ARMS-D, and 5-item questionnaire. The remaining PROMs, e.g., MMAS-8, SMAQ, MEDS, MNPS, ARMS-12, MGT, MTA-OA, MTA-Insulin, LMAS-14, MARS-5, A-14, ARMS-10, IADMAS, MAQ, MMAS-5, ProMAS, ARMS-7, 3-item questionnaire, AS, 12-item questionnaire, and Mascard were considered as having the potential to be recommended for use (category "b").

CONCLUSION

None of the included PROMs met the criteria for being classified as trusted and recommended for use for patients with cardiovascular diseases and/or type 2 diabetes mellitus. However, 21 PROMs have the potential to be recommended for use, but further studies are needed to ensure their quality based on the COSMIN guideline for systematic reviews of PROMs.

SYSTEMATIC REVIEW REGISTRATION

PROSPERO CRD42019129109.

Assessment of medication adherence among Lebanese adult patients with non-communicable diseases during COVID-19 lockdown: a cross-sectional study

Background

Medical treatment is considered a cornerstone in non-communicable diseases (NCDs) management, lack of adherence remains the main challenge that may compromise optimal therapeutic outcome achievement.

Purpose

This study aimed to evaluate treatment adherence levels and associated factors among Lebanese adult patients with non-communicable diseases.

Materials and methods

A cross-sectional survey conducted during the COVID-19 lockdown imposed by the Lebanese Government (between September 2020 and January 2021) enrolled 263 adult patients through an anonymous online questionnaire to assess adherence to medications using the Lebanese Medication Adherence Scale (LMAS-14).

Results

Of the total sample, 50.2% showed low adherence with a total mean adherence score of 4.41 ± 3.94. The results showed that depression (β = 1.351) and peptic ulcer (β = 1.279) were significantly associated with higher LMAS scores (lower adherence). However, age between 50 and 70 (β = -1.591, p = 0.011), practicing physical exercise (β = -1.397, p = 0.006), having kidney disease (β = -1.701, p = 0.032), and an intermediate (β = -1.336, p = 0.006) to high income (β = -3.207, p < 0.001) were significantly associated with lower LMAS scores (higher adherence).

Conclusion

Our study shed light on the factors affecting medication adherence in patients with non-communicable diseases. It showed that depression and peptic ulcer were associated with lower adherence, contrary to older age, exercising, having chronic kidney disease, and a higher socioeconomic status.

Global Burden, Regional Differences, Trends, and Health Consequences of Medication Nonadherence for Hypertension During 2010 to 2020: A Meta-Analysis Involving 27 Million Patients

Background

Nonadherence to antihypertensive medications is the leading cause of poor blood pressure control and thereby cardiovascular diseases and mortality worldwide.

Methods and Results

We investigated the global epidemiology, regional differences, and trend of antihypertensive medication nonadherence via a systematic review and meta‐analyses of data from 2010 to 2020. Multiple medical databases and clinicaltrials.gov were searched for articles. Observational studies reporting the proportion of patients with anti‐hypertensive medication nonadherence were included. The proportion of nonadherence, publication year, year of first recruitment, country, and health outcomes attributable to antihypertensive medication nonadherence were extracted. Two reviewers screened abstracts and full texts, classified countries according to levels of income and locations, and extracted data. The Joanna Briggs Institute prevalence critical appraisal tool was used to rate the included studies. Prevalence meta‐analyses were conducted using a fixed‐effects model, and trends in prevalence were analyzed using meta‐regression. The certainty of evidence concerning the effect of health consequences of nonadherence was rated according to Grading of Recommendations, Assessment, Development and Evaluations. A total of 161 studies were included. Subject to different detection methods, the global prevalence of anti‐hypertensive medication nonadherence was 27% to 40%. Nonadherence was more prevalent in low‐ to middle‐income countries than in high‐income countries, and in non‐Western countries than in Western countries. No significant trend in prevalence was detected between 2010 and 2020. Patients with antihypertensive medication nonadherence had suboptimal blood pressure control, complications from hypertension, all‐cause hospitalization, and all‐cause mortality.

Conclusions

While high prevalence of anti‐hypertensive medication nonadherence was detected worldwide, higher prevalence was detected in low‐ to middle‐income and non‐Western countries. Interventions are urgently required, especially in these regions. Current evidence is limited by high heterogeneity.

Registration

URL: www.crd.york.ac.uk/prospero/; Unique identifier: CRD42021259860.

Adherence to Post-Stroke Pharmacotherapy: Scale Validation and Correlates among a Sample of Stroke Survivors

Background and Objectives: Adherence to post-stroke pharmacotherapy has been less studied compared with other cardiovascular diseases, and previous research in this context utilized generic tools without cross-validating for stroke specific factors and patient characteristics. This study aimed to validate the Lebanese Medication Adherence Scale (LMAS-14) among stroke survivors to assess adherence to post-stroke pharmacotherapy. It also aimed to determine the socioeconomic, clinical characteristics, and health related quality of life correlates of medication adherence among stroke survivors. Materials and Methods: This was a cross-sectional study that included stroke survivors from districts throughout Lebanon. A well-structured questionnaire consisting of three parts was developed and utilized to collect data. The first part included questions about the sociodemographic and socioeconomic characteristics. The second part included questions about medical history, current clinical characteristics of the patients, and use of medications. The third part included validated scales to assess stroke outcomes, daily performance and activities, and quality of life. Results: A total of 172 stroke survivors were included. The LMAS-14 structure was validated over a solution of three factors, with a Kaiser−Meyer−Olkin (KMO) measure of sampling adequacy = 0.836 and a significant Bartlett’s test of sphericity (p < 0.001). Severe difficulty in obtaining medications within the current Lebanese economic crisis was significantly associated with lower medication adherence (Beta = −8.473, p = 0.001). Lower medication adherence was also associated with poor stroke prognosis (Beta = −3.264, p = 0.027), higher number of used medications (Beta = −0.610, p = 0.034), and longer duration of stroke diagnosis (Beta = −4.292, p = 0.002). Conclusions: The LMAS−14 is a valid and reliable tool to assess medication adherence in stroke practice and research. Severe difficulty in obtaining medications due to unpredictable availability and shortage of supplies is associated with lower medication adherence, and thus places stroke survivors at higher risk of complications and morality. Additional measures and urgent action by stroke care providers and public health stakeholders are necessary to ensure adequate post-stroke management and outcomes.

Risk Factors for Coronary Heart Disease Among Lebanese Women: A Case–Control Study

Purpose

Women are increasingly concerned by coronary heart disease (CHD), with peculiarities of their own, particularly concerning risk factors. The aim of the study was to assess the risk factors for CHD in Lebanese women over forty.

Patients and Methods

A case–control study was carried out in 6 hospitals in Beirut and Mount-Lebanon, from December 2018 to December 2019 including 1500 patients (1200 controls and 300 cases). Women were stratified into pre- and post-menopausal groups. Personal and medical data were collected from hospital records and during an interview where validated questionnaires were used. Binary logistic regressions were performed to investigate potential predictors of CHD in the 2 groups.

Results

In post-menopausal women, dyslipidemia (adjusted odds ratio [aOR], 3.018; 95% confidence interval, 2.102–4.332), hypertension (aOR: 2.449, [1.386–4.327]), a family history of CHD (aOR: 2.724, [1.949–3.808]), cigarette smoking (aOR: 2.317, [1.574–3.410]) and common non-rheumatic joint pain (aOR: 1.457, [1.053–2.016]) were strongly associated with CHD. Conversely, living in Mount Lebanon seemed protective, compared to Beirut (aOR: 0.589, [0.406–0.854]), as well as having a moderate monthly income (aOR: 0.450, [0.220–0.923]), adhering to a Mediterranean diet (aOR: 0.965, [0.936–0.994]), and practicing physical activity [PA] (aOR: 0.396, [0.206–0.759] and 0.725, [0.529–0.992], respectively for high and moderate vs low PA). In pre-menopausal women, dyslipidemia (aOR: 6.938, [1.835–26.224]), hypertension (aOR: 6.195, [1.318–29.119]), family histories of dyslipidemia (aOR: 6.143, [1.560–24.191]) and CHD (aOR: 4.739, [1.336–16.805]) reached statistical significance.

Conclusion

The identification of factors associated with CHD in women, some of which are frequent and trivialized in post-menopause, underlines the need to put in place specific and dedicated CHD prevention strategies in women.

Qualitative Interviews to Support Development and Cognitive Debriefing of the Adelphi Adherence Questionnaire (ADAQ©): A Patient-Reported Measure of Medication Adherence Developed for Use in a Range of Diseases, Treatment Modalities, and Countries

PURPOSE

The Adelphi Adherence Questionnaire (ADAQ©) is a newly developed generic patient-reported outcome (PRO) assessment of medication adherence. The aim was to assess its content validity by conducting cognitive debriefing (CD) interviews with patients prescribed medication(s) of various treatment modalities in a range of therapy areas.

MATERIALS AND METHODS

Targeted literature/instrument review and concept elicitation interviews informed development of the ADAQ©. CD interviews were conducted with 57 adults from the United States of America (USA; n = 21), Spain (n = 18), and Germany (n = 18) who prescribed medication for hypertension, diabetes, depression, schizophrenia, asthma, multiple myeloma, psoriasis, and/or multiple sclerosis. Interviews were conducted in two rounds to explore the relevance and understanding of the item wording, instructions, recall period and response options. Verbatim transcripts were analysed in ATLAS.Ti using thematic analysis. Three expert clinicians provided guidance throughout the study.

RESULTS

ADAQ© items/instructions were well understood and relevant to participants. Key modifications following round 1 included revising instructions to refer to current medication(s) for one condition to reduce cognitive burden, removing two items with lower relevance (specifically those assessing running out of medication and social discouragement), and adding a response option for participants to indicate if they had stopped taking a medication. Minor wording modifications were made following round 2. Subgroup differences in item relevance were explored based on clinical characteristics. Cost of medication was more relevant amongst US participants.

CONCLUSION

Content validity of the ADAQ© was confirmed in demographically and clinically diverse participants. Psychometric properties of the ADAQ© will be explored in future studies.

Measurement Properties of Patient-Reported Outcome Measures for Medication Adherence in Cardiovascular Disease: A COSMIN Systematic Review

BACKGROUND AND OBJECTIVE

Several medication adherence patient-reported outcome measures (MA-PROMs) are available for use in patients with cardiovascular disease (CVD); however, little evidence is available on the most suitable MA-PROM to measure medication adherence in patients with CVD. The aim of this systematic review is to synthesise the measurement properties of MA-PROMs for patients with CVD and identify the most suitable MA-PROM for use in clinical practice or future research in patients with CVD.

METHODS

An electronic search of nine databases (PubMed, MEDLINE, CINAHL, ProQuest Health and Medicine, Cochrane Library, PsychInfo, Scopus, Embase, and Web of Science) was conducted to identify studies that have reported on at least one of the measurement properties of MA-PROMs in patients with CVD. The methodological quality of the studies included in the systematic review was evaluated using the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) checklist.

RESULTS

A total of 40 MA-PROMs were identified in the 84 included studies. This review found there is a lack of moderate-to-high quality evidence of sufficient content validity for all MA-PROMs for patients with CVDs. Only eight MA-PROMs were classified in COSMIN recommendation category A. They exhibited sufficient content validity with very low-quality evidence, and moderate-to-high quality evidence for sufficient internal consistency. The 28 MA-PROMs that meet the requirements for COSMIN recommendation category 'B' require further validation studies. Four MA-PROMs including Hill-Bone Compliance Medication Scale (HBMS), the five-item Medication Adherence Report Scale (MARS-5), Maastricht Utrecht Adherence in Hypertension (MUAH), and MUAH-16 have insufficient results with high quality evidence for at least one measurement property and consequently are not recommended for use in patients with CVD. Two MA-PROMs (Adherence to Refills and Medications Scale [ARMS] and ARMS-7) are comprehensive and have moderate to high quality evidence for four sufficient measurement properties.

CONCLUSION

From the eight MA-PROMs in COSMIN recommendation category A, ARMS and ARMS-7 were selected as the most suitable MA-PROMs for use in patients with CVD. They are the most comprehensive with be best quality evidence to support their use in clinical practice and research.

Validation of the Lebanese Medication Adherence Scale among Lebanese diabetic patients

Background Adherence to oral antidiabetics plays a pivotal role in controlling diabetes. Healthcare workers evaluate this adherence when visited by patients. Objectives The primary objective of this study was to validate the existing LMAS-14 (Lebanese Medication Adherence Scale) in Lebanese diabetic adults. The secondary objective was to evaluate factors affecting adherence among this population. Setting The main five Lebanese governorates. Methods This study was conducted between May and September 2019, and community dwelling adult participants were enrolled using a proportionate random sample. The scale was validated using factor analysis and reliability testing, while bivariate and multivariable analyses assessed correlates of adherence. Main outcomes measures Validity of LMAS-14. Results A total of 308 participants were included, and the response rate was 91.25%. All LMAS-14 items converged over a solution of four factors, explaining a total of 64.39% of the variance (α = 0.859). The cutoff point between controlled and uncontrolled patients was set at 11. The sensitivity and specificity were good at this cutoff (71.1% and 94.74%, respectively). Results showed that 167 (57.2%) patients had good medication adherence. Advanced age (Beta = 0.046; p = 0.001) and having medical coverage (Beta = 1.452; p = 0.005) were significantly associated with higher adherence. Furthermore, adherence to oral antidiabetic drugs (Beta = 1.197; p = 0.018), female gender (Beta = 2.695; p = 0.011), and taking dyslipidemia medication (Beta = 3.527; p = 0.005) predicted higher diabetes control. Conclusion This study validated the LMAS among Lebanese adult diabetic patients taking oral antidiabetic drug. Advanced age and having medical coverage were associated with higher medication adherence. Further national studies are warranted to corroborate our findings.

Evaluation of risk factors and drug adherence in the occurrence of stroke in patients with atrial fibrillation

Background:

Atrial fibrillation (AF) patients are at high risk of developing a stroke and anticoagulant medications are generally prescribed to prevent stroke in AF population.

Objective:

This study aims to evaluate stroke risk factors among hospitalized patients with AF and to assess the level of adherence to medications in AF patients and their relation with stroke.

Methods:

This is a case-control study conducted between June 1^st, 2018 and December 31^th, 2018 among AF patients admitted to seven tertiary Lebanese hospitals. Data were collected using a standardized questionnaire. Adherence to medications was assessed using the Lebanese Medication Adherence Scale-14. Odds ratios (OR) expressed the strength of association between the independent variables and the dependent variable and were estimated using unconditional logistic regression adjusted for confounding factors. P<0.05 determined statistical significance.

Results:

In total, 174 cases of AF patients were included with 87 cases and 87 controls. The risk of stroke among AF significantly increased with the presence of a history of hypertension, aOR 16.04 (95%CI, 2.27-113.37; p=0.005), history of coronary heart disease/myocardial infarction, and history of obesity. Anticoagulant medication significantly decreased the risk of stroke among AF patients, aOR 0.27 (95%CI, 0.07-0.98; P=0.047). High adherence to medications was significantly associated with a reduced risk of stroke, aOR 0.04 (95%CI, 0.01-0.23; p<0.001).

Conclusions:

Having a history of hypertension is one of the strongest risk factors for stroke among AF patients in Lebanon. While anticoagulant medication use was associated with a reduced risk for stroke, high adherence to medications is critical for stroke prevention. Public health interventions are needed to tackle low-adherence to medication and prevent stroke among AF patients.

Validation of diabetes medication adherence scale in the Lebanese population

AIM

To validate the Diabetes Medication Adherence Scale (DMAS-7), determine its concordance with another validated scales and to assess factors affecting medication adherence.

METHODS

A cross-sectional study was conducted on a sample of Lebanese patients with diabetes using a questionnaire. The level of adherence was measured using the DMAS-7 and the Lebanese Medication Adherence Scale (LMAS-14). Bivariate and multivariate analyses were conducted, and the scale was validated in terms of reliability, predictive ability, and construct validity using SPSS version 19.

RESULTS

Out of 300 eligible patients, the rate of adherence was 33.7%. Measures of validity showed good reliability (Cronbach alpha = 0.627), and good construct validity with LMAS-14 (Spearman's rho = 0.846; Cohen's kappa = 0.711). DMAS-7 was found to be both correlated with LMAS-14 (ICC average measure = 0.675; p-value <0.001) in addition to possessing a better predictive value. Thus, DMAS-7 showed to have good concordance and increased validity compared to LMAS-14. Having an optimal glycated hemoglobin (HbA1C) (OR = 0.779; p = 0.001) and performing regular physical activity (OR 2.328; p = 0.002) increased medication adherence.

CONCLUSION

The DMAS-7 showed to be reliable and valid instrument superior to LMAS-14 in predicting adherence levels to oral anti-diabetic medications, and thus can be used to achieve better glycemic outcomes.

Evaluation of antidiabetic medication adherence in the Lebanese population: development of the Lebanese Diabetes Medication Adherence Scale

Objective

To evaluate factors affecting adherence to oral antidiabetic treatment in the Lebanese population and to develop the Diabetes Medication Adherence Scale (DMAS) based on these factors.

Methods

A cross-sectional study was conducted on a sample of Lebanese diabetic patients. Data were collected using a structured questionnaire. The level of adherence was measured using the Lebanese Medication Adherence Scale (LMAS-14). Bivariate analyses and multivariable analysis was done using SPSS. Psychometric evaluation of DMAS included an assessment of internal consistency, factor analysis, evaluation of sensitivity and specificity. Criterion-related validity was assessed by comparison with LMAS-14 measure of adherence.

Key findings

A total of 500 patients were recruited. 39.2% were adherent to treatment. Long working hours, increased number of oral antidiabetic medication per day, drug discontinuation when travelling, longer duration of diabetes and treatment burden were among factors that decreased adherence. While understanding the treatment regimen, following up physician recommendations and following up the recommended diet contributed to good medication adherence. The final 7-item scale (DMAS) had a good internal consistency (Cronbach’s α = 0.612) and a good correlation and agreement with LMAS-14 (Spearman’s rho = 0.699, Cohen’s kappa = 0.566). Patients with high DMAS scores were significantly more likely to have controlled glycaemia (P &lt; 0.05). Sensitivity and specificity reached 70.39% and 51.47%, respectively.

Conclusion

Adherence to oral antidiabetic treatment is suboptimal in Lebanon. The DMAS is a reliable instrument for assessing adherence and predicting poor glycaemic control in clinical practice, but requires further validation in other populations.

Evaluation of medication adherence among Lebanese diabetic patients

Background:

Diabetes type 2 is considered one of the main public health concerns. Lack of adherence to treatment leads to poor therapeutic outcome, poor glycemic control, and high risk for developing diabetes complications.

Objectives:

The aim of this study is to evaluate adherence to oral antidiabetic medication in Diabetes type 2 Lebanese patients, and to evaluate factors leading to low adherence.

Methods:

A cross-sectional study was conducted in outpatients endocrinology clinics of two hospitals and four private clinics located in Beirut-Lebanon. Data was collected using a well-structured questionnaire by trained pharmacists. Adherence level was measured by the Lebanese Medication Adherence Scale (LMAS-14). Bivariate and multivariate analyses were conducted using SPSS version 20.

Results:

Overall, 245 patients were included in the study with the majority being females (54.3%) and obese (47.8%). Only 29% of the participants had controlled glycemia (HbA1c <7%) with 31.8% of subjects had high adherence to their medication compared to 68.2% with low adherence. Increased working hours/day was associated with a decrease in adherence to oral antidiabetic medication (OR=0.31; 95% CI 0.11:0.88; p=0.029). Other factors significantly associated with decreased adherence to treatment were forgetfulness, high drug costs, complex treatment regimens, experiencing side effects, and perception of treatment inefficacy. Postponing physician office visits also decreased the probability of being adherent to oral antidiabetic medication (OR=0.36; 95% CI 0.15:0.86; p=0.022). Skipping or doubling the dose in case of hypo/hyperglycemia and the sensation of treatment burden also decreased medication adherence (OR=0.09; 95% CI 0.02:0.34; p=0.001, and OR=0.04; 95% CI 0.01:0.13; p<0.001 respectively).

Conclusions:

Adherence to oral antidiabetic medication is low for Lebanese patients, which leads to a poor glycemic control and increases the diabetes complications. Intervention programs including patient education strategies are essential to improve medication adherence.