Interventions to improve medication adherence in adults with mental–physical multimorbidity in primary care: a systematic review

Background Medication non-adherence is a notable contributor to healthcare inefficiency, resulting in poor medication management, impaired patient outcomes, and ineffective symptom control. Aim To summarise interventions targeting medication adherence for adults with mental–physical multimorbidity in primary healthcare settings. Design and setting A systematic review of the literature — published in any language and with any country of origin — was conducted. Method MEDLINE, EMBASE, PsycInfo, Web of Science, Cochrane Library, and the Cumulated Index to Nursing and Allied Health Literature — more commonly known as CINAHL — were searched for relevant studies. Data were extracted and synthesised using narrative synthesis. The Effective Practice and Organisation of Care (EPOC) taxonomy was used to classify intervention types. Risk of bias was assessed using the National Heart, Lung, and Blood Institute’s quality assessment tool for controlled intervention studies. Results Eleven studies, representing 2279 patients, were included. All interventions examined were classified into one EPOC domain, namely ‘delivery arrangements’. All included studies examined patients who had a physical condition and depression. Seven studies examining interventions focused on coordination of care and management of care processes reported statistically significant improvements in medication adherence that were attributed to the intervention. Four studies considering the use of information and communication technology observed no changes in medication adherence. Conclusion Interventions that coordinate and manage healthcare processes may help improve patients’ adherence to medication regimes in those with mental–physical multimorbidity. However, it is still necessary to better understand how digital health technology can support patients in following their medication regimes. As the growing challenges of treating multimorbidity are faced, everyone involved in health services — from providers to policymakers — must be receptive to a more integrated approach to healthcare delivery.


Introduction
Medication adherence 1,2 is 'the degree to which the person's behaviour corresponds with the agreed recommendations from a healthcare provider'. 1This description encompasses several behaviours, from seeking medical attention to consuming medication as prescribed. 2,3Nonadherence, therefore, presents health systems with a multifaceted challenge, imposing a significant economic burden globally. 4Five interacting dimensions are recognised to affect the ability to adhere to medication, namely: • social and economic factors; • healthcare provider-related factors; • condition-related factors; • therapy-related factors; and • patient-related factors. 3,5,6cioeconomic status is one of the most frequently explored contributors to poor medication adherence, because of illness severity and primary care accessibility. 2,7][10] Multimorbidity is the coexistence of at least two long-term conditions; 11,12 it poses unique clinical challenges as patients have coexisting, and potentially interacting, diseases, which may amplify symptoms and discomfort. 13,14Managing multimorbidity is complex and interventions to improve outcomes must be multifaceted. 15onsequently, individuals living with multimorbidity often require multiple medications to achieve optimal treatment. 16,17This may result in a considerable burden on these individuals -for example, those with five or more Research chronic conditions could spend 5-8 hours a day managing these, which could contribute to treatment burden and reduce medication adherence. 18,19Patients may also prioritise certain medications over others, according to disease progression and severity, acceptability or tolerability, and perceived importance. 2,7,14][22] Additionally, psychiatric treatments have a lower adherence rate than those for physical conditions. 23[25][26] Previous studies 16,27,28

Method
This review was conducted in line with recommendations in the Cochrane Handbook for Systematic Reviews of Interventions 29 and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) Guidelines. 30The protocol for this systematic review was preregistered on PROSPERO (registration number: CRD42022332974).

Eligibility criteria
The population, intervention, comparison, outcome, and study design -often referred to as PICOS -framework 31 was used to formulate eligibility criteria.Studies were not excluded based on published language or country of origin.Studies with a quantitative randomised controlled trial design were eligible for inclusion if they considered adult populations with multimorbidity, including at least one chronic condition and at least one CMD comorbidity, presenting to primary care.Results from pilot studies were also eligible.Individuals with major depression -namely, a severe mental illness -were included in this systematic review for comprehensiveness.Studies were required to compare interventions targeting medication adherence in primary care with usual or standard care, or care without therapeutic components, and had to consider patient medication/ treatment/therapy adherence rate as a primary or secondary outcome.Measures of effect included: • self-reported questionnaires or structured interviews; • therapeutic drug monitoring; • electronic devices; and • prescription pick-up/refill rates.

Search strategy
Information sources.Articles were identified through searches of the electronic databases MEDLINE, EMBASE, PsycInfo, Web of Science, Cochrane Library, and Cumulated Index to Nursing and Allied Health Literature -more commonly known as CINAHL -from January 2000 until May 2022.The search strategy was initially constructed for MEDLINE (Supplementary Table S1), but was later appropriately adapted for use with the other databases searched.
Selection process.After deduplication, the titles and abstracts of the eligible articles were independently screened against the inclusion and exclusion criteria by two reviewers.Conflicts were resolved by discussion.This process was repeated for full-text examination.

Data collection process
Data relevant to the study question were independently extracted from each qualified study by one of the reviewers who had conducted the screening and validated by the other.These data were summarised in a tabular format.Data extracted included the intervention under investigation, sociodemographic factors, and intervention effectiveness.

Study risk of bias assessment
The same two reviewers independently assessed the included studies for risk of bias using the National Heart, Lung, and Blood Institute's study quality assessment tool for controlled intervention studies. 32Studies were not excluded based on quality assessment.

Synthesis methods
A meta-analysis was considered impractical, because of the anticipated heterogeneity of interventions.Consequently, a narrative synthesis approach was conducted, utilising the Effective Practice and Organisation of Care (EPOC) taxonomy 33 to classify the interventions reported.There are four categories: • delivery arrangements; • financial arrangements; • governance arrangements; and • implementation strategies.
Three studies 40,42,44 were deemed as being of good quality, five studies 36,37,39,41,43 were considered to be of fair quality, and three studies 34,35,38 were assessed as being of poor quality (Supplementary Table S3).Reasons for the poorer ratings included a failure to meet an adequate randomisation requirement, not specifying the randomisation method, not specifying the treatment allocation concealment, or not reporting the blinding of the researchers.

Intervention classification
All 11 trials [34][35][36][37][38][39][40][41][42][43][44] assessed interventions that could be classified as delivery arrangement interventions, according to the EPOC taxonomy.Of the interventions reported, the most frequent EPOC subcategories of the delivery arrangement domain that were explored were the: • coordination of care and management of care processes; and • use of information and communication technology (ICT).

Interventions focused on coordination of care and management of care processes
Seven studies [34][35][36]40,41,43,44 examined interventions that targeted the coordination of care and the management of care processes. These interentions were multifaceted, with several components spanning EPOC subcategories.These components included: • case management; • integration; • shared care; • shared decision making; and • teams.
For instance, three studies [34][35][36] assessed an integrated depression and chronic-disease treatment intervention by utilising an integrated care manager as a liaison between the physician and patient.The role of this intermediary case management was to work with the patient to address the factors of adherence and effectively communicate the rationale behind antidepressant and antihypertension medication use.Similarly, the multicondition collaborative care management intervention proposed by Lin et al 41 aimed to coordinate care using a care manager.
The nurse-led intervention detailed by Safren et al 43,44 integrated adherence counselling with traditional cognitive behavioural therapy (CBT) techniques for treating depression and antiretroviral therapy adherence.This programme commenced with problem solving, employing case studies (Life-Steps) to prepare participants for subsequent CBT sessions.Additional resources offered to participants included reminders, graphic aids, and self-care and physical-activity techniques.
Katon et al 40 focused on shared care, decision making, and team collaboration.Nurses collaborated with other health professionals, such as psychiatrists, who would co-review medication use, provide recommendations, and work closely with patients to establish patient-centred goals.A stepped-care algorithm was also employed, whereby participants who did not initially meet goals had to choose their next treatment plan during a second phase of the intervention.

ICT interventions
Four studies [37][38][39]42 utilised ICT and telemedicine to deliver care. Clake et al 37 assessed the delivery of their self-management CBT programme through a web-based platform.Targeting the improvement of social and occupational functioning, participants were required to complete a minimum set of CBT modules.Moreover, the platform offered self-monitoring reminders, home practice activities, and motivational statements. Ths intervention was entirely self-guided; the other ICT interventions were guided by health professionals.
Both Himelhoch et al 39

Research
telehealth management programme involved the deployment of an electronic device (Health Buddy ® ) to facilitate communication -it sent individualised daily reminders to patients to take their medications.

Discussion
Summary Current interventions show a potential to improve medication adherence in adults with mental-physical multimorbidity in primary care, despite substantial heterogeneity in participant and intervention characteristics.Trials that assessed interventions focused on the coordination of care and care processes showed improvements across all outcomes, supporting the implementation of this intervention type for this complex patient population.In contrast, studies examining the efficacy of ICT interventions reported conflicting findings.

Strengths and limitations
To the authors' knowledge, this is the first systematic review to examine the effectiveness of interventions to improve medication adherence for adults with multimorbidity including CMD.This review addresses a crucial evidential gap for this group, presenting valuable insight to improve service delivery.Moreover, the focus on RCTs ensured that the evidence generated was of high quality, as RCTs are recommended when establishing a causal relationship. 45spite an extensive search, only 11 articles were included because of strict eligibility criteria.Only depression plus a limited number of physical comorbidities were trialled in the studies eligible, restricting applicability across other chronic diseases and mental disorders.Few countries were represented, with all trials conducted in countries with English as their primary language; however, it is important to note that variation in terminology and intervention description, due to differences in language or culture, may have led to some studies being overlooked in the search.This review is also vulnerable to the limitations of the various methods used to quantify adherence, such as self-report measures. 46ere are recognised limitations in employing the EPOC taxonomy as a guide in synthesis, 33

Coordination of care and management of care processes
Three studies [34][35][36] reported that, for their respective outcome assessments, the intervention group exhibited a higher proportion of patients who had ≥80% adherence to medication, compared with their control participants.Three studies 40,43,44 reported that intervention participants had significantly greater medication adherence than the control group.One of these studies 44 recorded that adherence gains were not maintained at 8 months.A further study 41 reported that, although there were no differences observed for medication adherence, adjustment rates were higher among the intervention group, relative to the control Four studies [34][35][36]44 reported significant reductions in depression status and depressive symptoms, compared with their control participants Three studies [34][35][36] reported better physical chronic-disease indices compared with their control participants

ICT
Two studies 38,39 observed that medication adherence was maintained in the intervention participants, in comparison with the controls.Two studies 37,40 reported no differences in medication adherence between intervention and control participants Two studies 37,42 reported significant reductions in depressive symptoms.
In contrast, one study 39 observed that both intervention and control group participants experienced significant reductions in depressive symptoms.
One study 38 reported that there were no significant changes for both groups in terms of depression scores One study 37 observed that intervention participants experienced no differences in blood glucose monitoring.Another study 42 reported no differences in Hba1c levels between the intervention and control groups EPOC = Effective Practice and Organisation of Care.Hba1c = glycated haemoglobin.ICT = information and communication technology.Some of the interventions examined that were reported as successful at improving medication adherence have also been successful when assessed for suitability in other contexts across a variety of outcomes -for example, team care has been used in the treatment and management of other conditions, such as depression and anxiety, fostering improved outcomes in participants, and contributing to the potential usefulness of the intervention throughout the healthcare system. 51revious research 52,53 has explored the relationship between care coordination and medication adherence; although the exact mechanism of association is not fully understood, it is likely influenced by fostering a robust relationship between the patient and the professional. 54,55n addition, evidence indicates that an ongoing partnership may improve patient satisfaction, trust, and communication. 56plications for research and practice This systematic review identified that the dominant intervention type assessed targeted service delivery arrangements.Trials assessing coordination of care interventions reported improvements in medication adherence as well as mental and physical health outcomes.In contrast, ICT interventions failed, overall, to improve outcomes for trial participants.Additional resources must, consequently, be allocated to foster the creation, testing, and implementation of interventions that aim to integrate care to improve outcomes for this complex group of patients.
To mitigate the impending challenges associated with multimorbidity, health services, providers, and policymakers must be receptive to adopting alternative approaches to care.National guidelines and policies should be reviewed to ensure alignment with best practices for this complex patient population and to support the implementation of novel ideas for change.This review presents several interventions that could be utilised to improve medication adherence, either as a reference for care delivery or as a foundational basis for additional development.Primary care providers should embrace the evidence presented in this review to reinforce care management processes and coordinate efforts to improve outcomes.The patient-provider relationship should also be prioritised when formulating strategies to enhance medication adherence for this complex patient population.The benefit of investment in medication non-adherence strategies could markedly outweigh any associated short-term expenses associated with staffing, cost, and capacity constraints. 4cision makers should also grasp the opportunity to use digital health technologies and integrate them into usual service delivery.Despite conflicting evidence, the MyCompass 37 platform and the Health Buddy ®38 device represent how digital health technologies may improve outcomes and alleviate associated burdens for the provider.However, comprehensive training, acceptance, communication, and organisational stability must be given special consideration to support the successful implementation of digital health technologies and ensure optimal outcomes. 57ere is a lack of evidence concerning which existing interventions improve medication adherence for adults with mental-physical multimorbidity in primary care and to what extent they do so.Further research is urgently required to expand the sparse evidence base for interventions supporting the care of this complex group of patients.Most of the studies collated in this review focused solely on a specific combination of chronic disease and CMD; subsequently, more trials should actively include participants with various comorbidity combinations.This would provide greater insight into the viability of widespread implementation, whereby adults with multimorbidity present to primary care with varied combinations of chronic disease.Future research should also conduct subgroup analyses to assess how variables associated with adults with multimorbidity affect the efficacy of these interventions.As an example, only one study in this review recruited African American participants.Special focus should be given to age, ethnicity, and socioeconomic status, as the literature on these factors is limited.By neglecting to incorporate these variables into the analysis, appropriate evidence will not be generated, potentially resulting in inadequate care.Furthermore, the interventions trialled in this review may not be suitable for implementation across all global contexts.It would be advantageous if trials examined efficacy in other contexts and countries to ensure the interventions' contextual adaptability.Economic evaluations to determine costeffectiveness would also be beneficial by providing decision makers with additional operational information for reaching a consensus on feasibility.