Development and validation of the Beliefs and Behaviour Questionnaire (BBQ)

Abstract

Objective

To develop and validate a questionnaire to screen for potential nonadherence in patients with chronic ailments.

Methods

Themes from qualitative interviews with chronic obstructive pulmonary disease patients were used in developing content of the questionnaire. The questionnaire was distributed to 525 ambulatory patients with chronic lung diseases. Principal components analysis was performed to identify the subscales in the questionnaire. Internal consistency, validity and stability of the subscales were also evaluated.

Results

The 30-item Beliefs and Behaviour Questionnaire (BBQ) measures beliefs, experiences and adherent behaviour on five-point Likert-type scales. Two hundred and eighty patients (53.3%) with a mean age of 71.1 years responded to the questionnaire. The ‘beliefs’ section had a two-factor solution—‘confidence’ and ‘concerns’ with internal consistencies of 0.82 and 0.45, respectively. The two domains identified from the section ‘experiences’—satisfaction’ and ‘disappointment’ had internal consistencies of 0.85 and 0.52, respectively. The ‘behaviour’ section, separately entitled the Tool for Adherence Behaviour Screening (TABS), had a two-factor solution—‘adherence’ and ‘nonadherence’, with internal consistencies of 0.80 and 0.59, respectively. All the domains demonstrated comparable reliabilities across two different patient populations. Their temporal stabilities ranged between 0.62 and 0.94.

Conclusion

The validity, reliability and utility of the BBQ and the TABS, a sub-scale of the BBQ that screens both intentional and unintentional nonadherence to pharmacological and non-pharmacological disease management, have been established.

Practice implications

The BBQ and the TABS have potential applications in screening adherence beliefs, experiences and behaviour in both clinical practice and research.

Introduction

Patient adherence to physicians’ recommendations is the key intermediary between medical practice and patient outcomes [1]. Adherence to recommended treatment for different disease conditions in various settings is known to be approximately 50% [2]. Adherence to non-pharmacological management (e.g. lifestyle changes such as diet, exercise, and smoking cessation) is known to be the most difficult for patients [3]. Nonadherence continues to be a major source of frustration for health professionals and an impediment to achievement of therapeutic goals [4].

The nature, extent and determinants of nonadherent behaviour are complex. There are no theories of adherence per se, but various models and theories such as Health Belief Model [5], Locus of Control Theory [6], and Self-regulatory Model [7] have been used to predict the variability that characterizes behavioural adherence. Despite extensive research, limited understanding of adherence phenomena and the absence of a standard theoretical framework suitable for all populations for empirically testing adherence outcomes against the determinants are common problems in adherence research [8]. Based on an extensive review of the literature on socio-psychological and related variables which have proven to be consistent predictors of compliance and patient acceptance of recommended health behaviours, Becker and Maiman [9] constructed a theoretically driven model for predicting and explaining compliance (adherence) behaviour with health and medical care recommendations.

Patient self-report using reliable and valid questionnaires is the most efficient and cost-effective method of assessing adherence and related beliefs and attitudes [10], [11], [12], [13]. From the clinician's point of view, self-reports are the most viable and useful measures of adherence [14], as they can identify the reasons behind nonadherence along with its detection, which could then help in rectifying or addressing those underlying issues. Self-report when combined with clinical observation has been shown to have better sensitivity than self-report alone [15]. Various self-report tools have been used for studying adherence behaviours and associated health beliefs and attitudes in both general and specific patient populations [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30]. The most commonly used measures of adherence are Morisky scale [16], the Medication Adherence Report Scale (MARS) [17] and the Brief Medication Questionnaire (BMQ) [18]. Weinman et al. [29] developed the Illness Perceptions Questionnaire to fit with Leventhal's description of illness representation for assessing the cognitive representation of illness. Horne et al. [30] developed the Beliefs about Medicines Questionnaire to quantify patients’ personal beliefs about the necessity of their prescribed medication and their concerns about taking it. The correlations between patient beliefs about prescribed medications and self-reported adherence were in the predicted direction in a cross-sectional study of patients with various chronic diseases [13].

Patients may exhibit different types of nonadherent behaviour. Unintentional nonadherence may be due to forgetfulness, or inability to follow treatment instructions due to poor understanding or physical problems such as poor eyesight or dexterity, whereas intentional nonadherence arises when the patient rejects either the doctor's diagnosis or the doctor's recommended treatment [31], [32]. The latter is often based on a patient's rational decision and has been called ‘intelligent nonadherence’ [33]. Previous studies have recognised the importance of screening for both intentional and unintentional nonadherence when adherence assessments are being made [21], [34]; however, it is critical that a single scale measures a single construct or characteristic [35]. Commonly used adherence screening instruments such as Morisky's scale compound the two facets of nonadherence – intentional and unintentional nonadherence – in one scale [34].

No single tool is available for measuring the different health beliefs and adherent behaviours with regard to both pharmacological and non-pharmacological management in patients with chronic ailments. Kravitz and Melnikow [1] have highlighted the need for new adherence metrics, especially those capable of screening patient's behaviours. Nonadherence is a multifactorial health behaviour, which can be best understood only within the patient's physical, economic, psychological and social considerations. In-depth interviews with subjects who are representative of the population of interest is an ideal way of generating credible and highly face-valid items for inclusion in health care scales [36]. The Becker and Maiman model offers a suitable conceptual framework for predicting and explaining adherence behaviour with health and medical care recommendations. The aim of this study was to develop and validate a health beliefs and behaviour questionnaire for screening adherent behaviour in patients with chronic ailments.