Improving adherence in paediatric respiratory disease

Introduction

To the frustration of many physicians, patients do not always follow their doctor's advice. In fact, it is quite unusual for patients to do exactly what their doctor recommends them to do. Historically, such noncompliance was considered to be due to irrational behaviour or wilful ignoring of instructions, and patients deviating from doctors' instructions were described with terms such as “unreliable”, “untrustworthy” and “faithless” [1]. This illustrates how doctors traditionally blamed the patients for not following expert advice. Over the past few decades, however, scientists have increasingly recognised that the patients' motives for deviation from their physician's recommendations are not irrational at all [2, 3]. On the contrary, they make perfect sense to the one person to whom they matter – the patient. Patients, it turns out, have strong and persistent cognitions about their health, about illness, and about the best way to prevent and treat illness. These cognitions often strongly deviate from the medical model of the chronic illness, such as asthma (see “Perceptions about asthma and treatment with inhaled corticosteroids inconsistent with the medical model of asthma” box).

This recognition has encouraged research into how these patient perceptions of illness and medications may be explored, and perhaps even modified to concord with the medical state of the art. The focus of research in this area is, therefore, shifting from blaming the patient to examining what doctors can do to help patients take the optimal decisions about the treatment of their illness. In this article, we will review the evidence on “compliance” and “noncompliance” in general, with an emphasis on paediatric respiratory disease. This will form the basis for a set of simple principles which will help establish concordance between patient and healthcare provider, which increases the likelihood that patients will follow their advice.

Terminology

The traditional term of “noncompliance”, along with its derogatory characterisation of patients, has largely been abandoned in research publications. In 2003, the World Health Organization (WHO) recommended to use the term “(non)adherence” instead, and presented a working definition of the concept (see definitions box). Despite this attempt to standardise terminology, the concept of (non)adherence remains problematic for several reasons, as noted by Steiner and Earnest [1]. “First, these words exaggerate the physician's control over the process of taking medications. Second, they imply that the patient must take the medication as prescribed to obtain benefit. Third, the terms "noncompliance" and "nonadherence" create a clinically unjustifiable distinction between persons who take all of their pills as prescribed and those who deviate from the prescription in any way. Finally, none of these terms accurately represents patients' motivations for choosing to take their medications a certain way.” [1].

Despite these shortcomings, the term “adherence” is widely used in research studies, and we will use it in this review according to the WHO definition.

Two additional terms need to be defined, both of which are related to, but not synonymous with, adherence (see definitions box) [3–5]. Many patients adhere to an initial treatment prescription, but fail to persist with it, for example because they do not receive a repeat prescription or because they are not followed-up. Nonpersistence is extremely common for inhaled corticosteroids (ICS) [6], in particular in primary care where scheduled follow-up visits for children with asthma are the exception rather than the rule [7, 8].

If patients and doctors explicitly agree on the choice, intensity, and duration of treatment (i.e. if they are in concordance), patients are more likely to adhere to the agreed treatment regimen, in particular if this takes their views, preferences and context into account [5, 9]. Thus, while concordance is related to adherence [10], it does not ensure it, as there may be numerous other issues interfering with adherence [11].

Nonadherence: is it relevant?

It has been well documented for a range of acute and chronic illnesses that nonadherence has a major impact on the efficacy of treatment and health outcomes for patients [12]. In childhood asthma, nonadherence has been known for decades to be related to increased symptoms, a higher risk of hospitalisation for asthma exacerbations and asthma deaths [13, 14]. Nonadherence has been recognised as an important cause of problematic severe asthma [15], although the extent to which problematic severe asthma can be explained by nonadherence is, as yet, unknown. Nevertheless, adherence is an issue worth exploring in every follow-up visit of children with a chronic (respiratory) disease, in particular when the disease remains poorly controlled despite the prescription of control medication.

Nonadherence: the scope of the problem

Contrary to what many people would instinctively think, nonadherence does not appear to be related to severity of disease [1, 2]. In a cohort of children with very problematic severe asthma referred to a national tertiary care referral centre in the UK, medication was either absent or out-of-date in 23% of home visits by an asthma nurse [16].

As a rule of thumb, patients take approximately half of the medication prescribed to them [1, 2]. In studies in children with asthma, where adherence was measured electronically by logging devices recording each actuation of the inhaler, median adherence rates usually vary between 30% and 70% of the total number of doses prescribed, with large variations between individual patients [17–21]. We recently reported a median adherence of 92% in 2–6-year-old children with asthma over a period of 3 months [22], illustrating that good adherence can be achieved in children with chronic respiratory disease [23].

Measuring adherence

As a rule, patients (or their parents) overestimate their own (or their child's) adherence to daily maintenance medication [24, 25]. Therefore, parental or patient reporting of adherence is unreliable [26]. Incidentally, the same applies for physicians' assessment of adherence in the patients to whom they provide medical care [24]. It has been suggested that enquiring about medication use, in an understanding and non-confrontational manner (e.g. “Many people find it difficult to take their medication daily. What’s this like for you?”), would provide a reasonable estimate of adherence [27], but very few studies have actually examined this. A recent study would suggest that asking parents about missed doses and about trials off medication at their own initiative could help in identifying nonadherence [24].

Pharmacy refill rates for prescriptions, although easily available, are similarly unreliable [26], because they only reflect whether the medication has been picked up from the pharmacy, not whether it has actually been taken. Pharmacokinetic and pharmacodynamic monitoring of drug effects, while highly accurate and reliable, is too invasive and costly to be used in large-scale clinical studies. Weighing used metered dose inhaler (MDI) canisters, or using the built-in counter devices of dry powder inhalers (DPIs) to calculate adherence is reasonably accurate, although “dumping” (emptying the device before returning it to the doctor to mimic good adherence) cannot be detected this way.

Electronic monitoring has emerged as the most accurate, valid and cost-effective method of monitoring adherence [28]. A number of electronic devices have been developed (fig. 1). The Doser device only counts the number of actuations and, therefore, does not detect dumping. The other devices record date and time of each actuation on a microchip which can store data for 4–6 weeks; these devices are able to detect dumping [28]. Three devices can monitor adherence of medication delivered by MDI (either with or without a spacer), and one is suitable for monitoring adherence to medication from the Diskus DPI, the SmartDisk (fig. 1). All electronic devices suffer to some extent from mechanical and electronic failures, but have otherwise been well validated [28]. There is now consensus in the literature that such electronic devices are the recommended method to measure adherence reliably, both in clinical practice and in research [26, 28].

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Electronic adherence monitoring devices suitable for use in clinical practice and research. a) Doser, b) MDILog, c) Smartinhaler and d) SmartDisk.

A useful model to understand medication taking behaviour

Recent research lends support to a theoretical model that explains differences between patients in the self-management of their chronic illness, including differences in their adherence to its treatment. This common sense model states that people try and make sense of a threat to their health in order to try and control the threat [29]. A patient with asthma who perceives the asthma to have an episodic nature will not perceive the necessity of taking preventive medication. This “no symptoms, no asthma” behaviour will lead to inadequate control of asthma [30]. By contrast, a patient who perceives the asthma to be a chronic condition that necessitates maintenance medication will be more likely adhere to medication use [31], and will have a higher chance of controlling their asthma. Patients create their own personal cognitive representation of their illness, which include beliefs about what may have caused the illness, the consequences the illness will have on their lives, how long the illness will last, and whether or not it is controllable or curable. In parallel, they also develop emotional responses to the threat. The cognitive and emotional representations of symptoms and illnesses are called illness perceptions [29]. These illness perceptions are shaped by early childhood experiences in which children learn how to respond to sickness and pain from caregivers and other influential people in their environment. In addition, public images (television, internet and the lay press) and the stories of parents, teachers and physicians play a role in shaping these experiences. Illness perceptions, therefore, are more strongly influenced by cultural, social and psychological factors than by the medical severity of the illness, or by intellectual capacity or socioeconomic class of the parents [3]. Cognitions about the benefits and harms of therapy are developed in a similar way. A schematic representation of the relationship between illness perceptions, medication beliefs and adherence is presented in figure 2 [3]. There is now good and consistent evidence from a range of chronic conditions [32, 33], including childhood asthma [22, 34, 35], that illness perceptions and, in particular, medication beliefs are strongly associated with (non)adherence. These cognitions of illness and treatment are stronger determinants of adherence than factors associated with the patient (such as age, sex or socioeconomic status), the disease (severity, chronicity) or the knowledge that the patient has about the disease. The latter explains why providing information (patient education) by itself is, therefore, not successful in improving adherence in paediatric chronic disease [36]. Only when the education is tailored to explore and modify illness perceptions and medication beliefs can it be expected to have a beneficial effect on adherence [29].

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Horne's extension of the common sense model, showing the relationship between illness perceptions, medication beliefs, and adherence. Reproduced from [3] with permission from the publisher.

Barriers to adherence

Although it is becoming increasingly clear that illness perceptions and medication beliefs are the strongest patient-related barriers to adherence, a number of other treatment-, physician- and patient-related barriers have been reported to be associated with (non)adherence (see barriers to adherence box) [37].

The more complex a treatment schedule is the less likely the patient is to adhere to it [38]. However, even with very complex maintenance treatment schedules, for example for patients with HIV or tuberculosis, excellent adherence to such treatment is possible when the importance of strict adherence has been discussed and agreed, including a balanced discussion of the pros and cons of treatment (i.e. when the medication beliefs of patients have been addressed). The difference in immediate effect between bronchodilators and ICS may help to explain why many patients with asthma over rely on bronchodilators to control their symptoms, at the expense of daily ICS use [39]. Adverse effects rank among the treatment-related barriers, while the (usually much stronger) perceived potential side-effects represent medication beliefs and patient-related barriers. Cost and reimbursement of medication may play a major role in adherence, in particular in patients with poor or absent health insurance [39].

Physicians who do not follow-up their patients regularly cannot be expected to build a strong therapeutic alliance, which is considered to be essential in the long-term management of paediatric chronic respiratory disease (www.ginasthma.org). Patients want their physician to be easily available for consultation and they prefer to see the same physician at every follow-up visit [37]. They also want their physician to show empathy, to be interested in their views, concerns and preferences [40]. Unidirectional, top-down information or advice (“you really should give up smoking now; how many times have I told you it’s bad for Jimmy's health?”) is largely ineffective [41]. What does help is exploring how patients feel about their disease and about taking medication on a daily basis (i.e. exploring their illness perceptions and medication beliefs), and tailoring the information provided on these cognitions [29]. Regular follow-up with the same empathic physician helps to allow this essential process to take place.

In addition to illness perceptions and medication beliefs, a number of other patient-related barriers to adherence may be encountered. Basic knowledge of the disease and its treatment is required to help patients to take the medication as prescribed, but we emphasise once again that such education per se does not improve adherence [36], but that it needs to be tailored to the information needs of patients. Families struggling with serious psychosocial issues, such as poverty, relationship problems or psychiatric illness, will face major problems in managing any chronic condition [42, 43].

What can the healthcare professional do to improve adherence?

The interventions physicians can employ to improve adherence should follow logically from the evidence reviewed above. Because adherence is a multidimensional construct, in which poor understanding of disease or treatment is only one of the many potential determinants (see box) it should come as no surprise that education per se does not improve adherence [36]. Although information on the disease and its treatment is an essential component of asthma self-management [44], the information should be tailored to the patient's information needs, views and preferences (see principles of improving adherence in paediatric chronic conditions box). To be able to build a therapeutic partnership with patients and their parents, repeated follow-up is needed. Scheduled follow-up helps to improve asthma control [7], probably through repeated tailored self-management education [45]. The beneficial effects of such self-management education persists for years after completion of the programme [46].

Principles of improving adherence in paediatric chronic conditions

In history taking, it is worthwhile to not only evaluate the patient's symptoms, but also pay specific attention to the family's context, their views, preferences and treatment goals, and their illness perceptions and medication beliefs. Such “interested enquiry” of the patient's perspective helps to identify major barriers to adherence (see box) and follows logically from the common sense model (fig. 2) [40]. Most patients and parents feel very comfortable in collaborating with healthcare professionals in this way [47], particularly if the conversation is performed empathically, in a warm, friendly and reassuring manner [48], and if active listening techniques are employed to ensure that the patient's perspective has been captured completely and accurately.

Consultations performed in this way work towards shared decision making, physicians and patients working together in developing a mutually agreed treatment plan [49]. This is a paradigm shift from the traditional doctor-centred approach of prescribing therapy. Based on the common sense model, patients cannot be expected to follow a prescription of medication when this does not take their own views (including illness perceptions and medication beliefs) and preferences into account. In shared decision making, the patient is explicitly asked to express his or her own treatment goals, and in weighing the pros and cons of therapy the patient's views, context and preferences are taken into account. A proof-of-concept study in adult asthmatics showed that shared decision making is associated with improved adherence [50].

The final step to ensure adherence is to evaluate the treatment and its success during further follow-up. Gentle probing of how the patient feels about taking medication and any problems the patient and parents may encounter in using the medication as agreed, may help to identify barriers to adherence, and opens up an exploration of the help required to overcome them.

Conclusions

Although nonadherence in paediatric chronic respiratory disease is common, empirical evidence strongly suggests that good adherence can be achieved. To understand a patient's medication-taking behaviour, the extended common sense model, comprising both illness perceptions and medication beliefs, is useful. In addition to these cognitions, a number of other barriers may hamper adherence. By exploring illness perceptions and medication beliefs, and by showing a genuine interest in the patient's context, views and preferences, these barriers can be identified, and tailored self-management education and help offered to overcome them. Rather than prescribing therapy, physicians can weigh the pros and cons of different treatment options with the patient and the parent, aiming at shared decision making of a mutually agreed treatment plan. These principles have been shown to be associated with improved adherence and can, therefore, be recommended for general use. Although this requires specific communication skills, these can be learned and mastered [51].