Introduction

Rheumatic disease is an umbrella term that encompasses the spectrum of musculoskeletal, arthritic and connective tissue disorders, which include conditions such as inflammatory bowel disease (IBD), psoriasis (PS), rheumatoid arthritis (RA), ankylosing spondylitis (AS), and psoriatic arthritis (PsA) [1]. Rheumatic diseases are a huge burden on the healthcare system worldwide. This group of diseases accounts for reduction in productivity, significant disability and affects quality of life [3]. A systematic review published in 2014 revealed that about 3% of worldwide population is suffering from inflammatory arthritis [4]. This group of diseases is commonly treated with disease-modifying anti-rheumatic drugs (DMARDs), corticosteroids and nonsteroidal anti-inflammatory drugs (NSAIDs) [5]. Thus, adherence to the prescribed drugs is important in the prognosis of the disease.

Adherence is defined as a patient’s agreement with the recommendations from physician. It is closely linked with the effectiveness of all pharmacological therapies. As such it is particularly crucial for medications prescribed for chronic diseases. Of all medication-related hospitalizations in the United States, around one-third to two-thirds are the result of poor medication adherence [2]. Non-adherence not only causes therapeutic failure but may also worsen the disease state. Hence, it is believed that increasing adherence may have a far greater impact on health outcomes than advances in medical treatments [6, 7].

According to Arthritis Research UK, drugs that are used for rheumatic diseases are broadly classified into two groups: drugs that treat the symptoms of arthritis (NSAIDs, steroids) and treatments that suppress inflammatory disease and may improve the outcome (DMARDs and Biologics). As biologics are newer class of drugs, it is commonly used in the treatment when other types of DMARDs have not been effective. Owing to these two reasons, adherence to DMARDs and NSAIDs is of utmost importance in order to slow down the progression of the disease, saving patient’s money as well as minimizing risk of side effects.

Adherence is influenced by many factors [8]. The most commonly studied factors are medication characteristic, perceptions and cognitions about illness and medication, socioeconomic and demographic factors, disease features and doctor–patient relationship [9]. As such, we based our factors on the WHO model of medication adherence; this model forms the basis of out theoretical framework and anchor. The variety of factors makes it difficult to understand the complex problems and formulate recommendations or to develop interventions that could help to increase adherence in patients.

Currently, there are systematic reviews looking at adherence rates of DMARDs in various rheumatic diseases [10, 11]. However, there is no succinct summary that seeks to identify, group and present factors affecting medication adherence in Rheumatology as a group. Therefore, the aim of this paper is to identify factors influencing adherence and to assess the association between these factors and adherence.

Methods

A search of literature was performed in online databases: PubMed®, PsychINFO® and Embase®. Keywords used for the search were: (adherence OR compliance OR persistence) AND (rheumatic OR rheumatism OR rheum* OR rheumatology OR lupus OR ankylosing spondylitis OR rheumatoid arthritis OR psoriatic arthritis OR scleroderma OR systemic sclerosis) AND factors AND (drugs OR medication OR medicine). Further hand searches were conducted using references of related articles. The literature review was current as of October 2016 and the start date for articles was unrestricted.

Two independent reviewers (H GOH and YH KWAN) independently reviewed the articles for inclusion and discussed when discrepancies arose.

Inclusion/exclusion criteria: We included full-text original studies published in English-language articles and included patients more than 18 years old with a diagnosis of any rheumatic conditions. Meta analysis, case series, case reports and reviews were excluded.

Subsequently, we identified relevant factors associated with medication adherence and the identified factors were then clustered according to World Health Organization (WHO) recommendations [10]. This includes patient-related factors, therapy-related factors, condition-related factors, health system factors and socioeconomic factors. Finally, a hand model of treatment adherence in rheumatic patients was proposed to help clinicians to better visualize these five categories that affected treatment adherence.

Results

In total, 1977 articles were retrieved from title and abstract screen. After removing 1840 articles that do not fulfill inclusion criteria from full article review and removing 32 duplicates, the final number of articles reviewed was 90 as shown in Fig. 1. The percentage agreement between Goh and Kwan was 90%. A total of 17 factors (38 sub-factors) were identified; we presented statistically significant factors and factors that are not statistically significant, separately and categorized, based on WHO’s five dimensions of adherence. In addition, the types and number of studies that support or do not support a particular factor were also identified. The characteristics of each study are presented in Supplementary Table 1.

Fig. 1
figure 1

Flow chart on selected articles for review

Full size image
Table 1 Patient-/caregiver-related factors associated with poorer medication adherence in rheumatologic patients
Full size table

Patient-related factors

A total of four main factors were identified under the patient-/caregiver-related category. A total of 29 studies (19 cohort studies, 7 cross-sectional studies, 3 qualitative studies) supported these four factors [12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46]. The number of studies and the types of studies that supported each factor are shown in Table 1.

The most cited factor in this category was ‘Patient’s Demographic Factors’, which encompassed patient’s age, gender, ethnicity and education level. ‘Patient’s Personality Factors’ includes forgetfulness, patient’s coping behaviours, personality traits and quality of life. ‘Patient’s Disease and Treatment Perceptions’ includes disease perception and knowledge, belief for the need of medication and perception towards diagnosis/treatment. ‘Caregiver Issues’ includes caregiver’s personality, dependence on caregiver and caregiver’s negligence.

Therapy-related factors

A total of five main factors were identified under the “Therapy-related factors” category. A total of 31 studies (22 cohort studies, 7 cross-sectional studies, 2 qualitative studies) supported these factors [13, 25, 30, 31, 42, 47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68]. The number of studies and the types of studies that supported each factor are shown in Table 2.

Table 2 Therapy-related factors affecting treatment adherence in rheumatologic patients
Full size table

The most cited factor in this category was ‘Choice of drugs’, which encompassed the dosage form of the drugs, properties of the drugs (taste, colour, smell, side) and how to administer the drugs. ‘Side Effects of Therapy’ includes the side effects experienced by patients during the course of treatment which lead to the discontinuation of drugs prescribed. ‘Length and Complexity of Treatment’ includes the duration of treatment, the number of drugs prescribed to the patients as well as dosing regimen. ‘Signs and Symptoms Severity’ suggests that patients would only take the medication regularly when there were worsening of signs and symptoms but will not take them when the signs and symptoms were insignificant. Lastly, ‘Changing of Medical Treatment’ refers to switching the current therapy that the patient was already receiving to another therapy. Patients might not be accustomed to the new therapy and hence were unable to follow the treatment plan completely.

Condition-related factors

A total of two main factors were identified under the “Condition-related factors” category. A total of 8 studies (5 cohort studies, 1 cross-sectional study, 2 qualitative studies) supported these two factors [29, 57, 61, 69,70,71,72]. The number of studies and the types of studies that supported each factor are shown in Table 3. The most cited factor in this category was ‘Depression’.

Table 3 Condition-related factors affecting treatment adherence in rheumatologic patients
Full size table

Health-system-related factors

A total of four main factors were identified under the “Health-system-related factors” category. A total of 10 studies (6 cohort studies, 3 cross-sectional studies, 1 qualitative study) supported these four factors [13, 16, 43, 51, 73,74,75,76,77,78,79]. The number of studies and the types of studies that supported each factor are shown in Table 4. The most cited factor in this category was ‘Poor healthcare provider communication and patient counseling’.

Table 4 Health-system-related factors affecting treatment adherence in rheumatologic patients
Full size table

Socioeconomic-related factors

A total of two main factors were identified under the “Condition-related factors” category. A total of 12 studies (9 cohort studies, 2 cross-sectional studies, 1 qualitative study) supported these two factors [16, 25, 78,79,80,81,82,83,84]. The number of studies and the types of studies that supported each factor are shown in Table 5. The most cited factor in this category was ‘Cost issues’.

Table 5 Socioeconomic-related factors affecting treatment adherence in rheumatologic patients
Full size table

Hand model of medication adherence in patients with rheumatic diseases

Through these five general categories of factors affecting adherence in rheumatic patients, we propose the hand model as shown in Fig. 2 to allow clinicians to remember these factors better and to apply these factors in their daily practice with rheumatic patients.

Fig. 2
figure 2

Hand model of medication adherence in rheumatic factors based on WHO recommendation [10]

Full size image

Discussion

This is a systematic review that involved a comprehensive review of literature involving factors affecting medication adherence in population suffering from rheumatological diseases. Although there are numerous studies done on various patient populations, the last systematic review on rheumatology patients was carried out about 7 years back [81]. The present review gives an overview of the types and the number of studies that supported or disagreed with the factors investigated.

There were five categories of factors presented in this hand model of adherence. Each finger represents a domain of factors that clinicians should take note of when assessing adherence in this group of patients. Hand model was designed to incorporate the five categories of factors identified in this review. During their daily practice, clinicians should focus and remember the five domains and apply this model as much as possible so that all factors of adherence can be addressed sufficiently.

In the “Patient-related factors” category, ‘Patient’s Demographic Factors’ was one of the common factors that was reported to cause non-adherence among patients in this category. It was stated that compliance with medication was higher in the older population and younger patients were at a higher risk of non-adherence [17,18,19,20, 35, 37, 38, 40, 41, 85,86,87]. Younger patients were more likely to experience non-adherence especially with regard to procedure-based therapies such as mouth care and central line care. It was also shown that older adolescents were less adherent to treatment as compared to younger adolescents, with the adolescent population defined as between 10 and 18 years old. Possible reasons include that adolescents were undergoing changes and face many problems such as autonomy, independence and separation. This could result in a transfer of medication responsibility between parent and children, thus resulting in disagreements on who was responsible for treatment adherence. It was also found out that different ethnic groups have a different compliance rate with the medication. This could be due to different perceptions on the necessity of medication. For example, three studies concluded that Whites have a higher compliance rate as compared to African American and Chinese [18, 29, 39]. However, as these studies were conducted in different settings, it was possible that there were other confounding factors (such as socioeconomic background) that were not well addressed. Lastly, patients’ belief in the need for medication was supported by ten studies and reinforces the importance of adequate patient education on the pathophysiology of disease, mechanism of action of therapeutics and most importantly patients buy-into adherence.

In the “Therapy-related factors” category, ‘Choice of Drugs and Route of Administration’ was one of the common factors that was reported to cause non-adherence among patients. It was found out adherence to drugs was higher when methotrexate (MTX) was prescribed as compared to treatment plan without methotrexate. A study revealed that concomitant MTX was associated with improved treatment continuation of biologics when compared with both TNF blockers as monotherapy and TNF blockers combined with other DMARDs [64]. We postulate that this was related to dosing frequency, side effects and ease of administration of the drugs. For example, side effects experienced during treatment were supported by nine studies to affect treatment compliance. Another study revealed that replacing hydrochloroquine and sulfasalazine in triple therapy with etanercept and MTX combination increases medication adherence [58]. Hence, these studies support that choice of drugs has a direct impact on affecting patient’s adherence to medication. As most of the biologic DMARDs were parenteral injection, it was found out that medication adherence in patients using subcutaneous injection was high due to ease in administration [64]. This suggests that self-injectable form was effective and well tolerated and was certainly a good addition to the therapeutic armamentarium for carefully selected patients with rheumatic diseases.

In the Condition related factors category, all of the studies we have identified support depression as an important reason behind medication adherence issues and difficulties with adherence were highly correlated with the severity of mental illnesses in which the patient was suffering from [57, 61, 69, 70]. For example, one study suggests that in general, depressed patient was 2.3 times more likely to self-report poor medication adherence than those who were non-depressed. As depression was a recognised comorbidity of people with rheumatic diseases, healthcare professionals supporting people with rheumatic diseases should have awareness of anxiety and depression and refer or screen as appropriate.

In the “Health-system-related factors” category, the majority of educational interventions have focused heavily on provision of information. However, raising the patient's level of understanding on the disease was not associated with increased medication adherence [88]. In fact these studies suggest that the goal of education should be to provide clear and comprehensible information to patients to allow them to make informed healthcare decisions [37]. This could be done by improving the communication between the physician and the patient. Once good rapport was established with the patient, it was more likely for him/her to follow the physician’s instruction. Clear and avoidance of medical jargons during patient counseling allow the patient to better understand the treatment regimen and drug usage [43, 74,75,76,77]. Hence, educational interventions should incorporate more clinician-patient interaction and include read-back to ensure patient understanding.

The “Socioeconomic-related factors” category highlighted that inability to afford the medications for long term has been a major and often neglected concern. One of the studies revealed the average annual cost of some commonly used biologics that a patient needs to pay yearly: adalimumab (US $25,960), abatacept (US $35,581), infliximab (US $36,107) [76]. Therefore, cost of medication was a profound issue that will affect adherence to medication, as most of the patients would require the use of these medications for a long period, if not throughout their remaining lifespan. In addition, multiple drug combinations were commonly used in treating rheumatic diseases, which will further increase the burden of cost of medications to the patient [84]. As such, patients who were at the borderline of poverty have the highest risk of discontinuing the medications due to financial constraints [16, 25, 78,79,80,81,82]. Thus, clinicians should discuss this issue with the patient so as to assess his financial capabilities and refer him for appropriate help when needed.

In terms of the types of studies, we found out that quantitative studies comprised of about 79% of the literature found, the rest being qualitative studies. A wide variety of study designs were used, including but not limited to case control studies, cohort studies, cross sectional studies and qualitative studies. Future studies should include quantitative components to measure the magnitude of effect that each factor has on medication adherence, in order to identify major causes and optimize resource allocation to improve adherence in those areas showing larger magnitude effect.

A weakness in this study was that this was a systematic review and not a meta-analysis of the various effects of factors on adherence. Although there were 17 factors in five categories and 38 sub-factors identified in this review, we do not know the aggregate magnitude of each factor on adherence rate in rheumatic patients. However, in our review, we have summarized as much as possible the available literature and presented the various factors succinctly into five major categories in order to allow clinicians to better understand the complex nature of adherence.

In conclusion, 17 factors in five categories and 38 sub-factors were found to be associated with the adherence in rheumatic patients. Clinicians would need to be cognizant of the complex nature of medication adherence in this population, and systematically address these risk factors in the patient in order to optimize therapeutic outcomes.