Patients’ trust in their physician—Psychometric properties of the Dutch version of the “Wake Forest Physician Trust Scale”

Abstract

Objective

Aim was to investigate the psychometric properties of a Dutch version of the “Wake Forest Physician Trust Scale”, which intends to measure patients’ trust in their physician.

Methods

A random sample of internal medicine patients visiting the outpatient clinic completed the questionnaire (N = 201). Dimensionality, reliability and validity of the instrument were examined.

Results

The structure of the questionnaire was best explained by a unidimensional construct. Reliability was confirmed: internal consistency was high (α = .88), and mean item-total correlations were all above .40. Construct validity was indicated by patients’ trust in their physician correlating significantly and as hypothesized with (1) satisfaction with their physician (r = .64), (2) with the length of the patient–physician relationship (r = .28), (3) with their willingness to recommend their physician (r = .71) and (4) their unwillingness to switch their physician (r = .61).

Conclusion

The results suggest the Dutch version of the Wake Forest Physician Trust Scale to be a psychometrically sound instrument to assess patients’ interpersonal trust.

Practice implications

Trust is a key feature of the patient–physician relationship, yet has been scarcely researched in other than Anglophone cultures. An adequate Dutch trust questionnaire forms the first step to gaining more knowledge about patient–physician trust in another culture and health care setting.

Introduction

Patients’ trust in their physician is an essential feature of the patient–physician relationship [1]. High trust in one's physician is associated with satisfaction with the physician, adherence to treatment, continuity of care, delayed care, unmet care needs and health outcomes [2], [3], [4], [5], [6]. Although these associations indicate the importance of trust in the patient–physician relation, little empirical research on patients’ trust in their physician has been done so far [1], [7].

In common language, ‘trust’ can be used interchangeably with terms such as ‘confidence’ or ‘faith’ [7]. In the medical setting, a differentiation between social or institutional and interpersonal trust has been made [1], [8]. Social trust refers to trust in the health care system as a whole. Interpersonal trust, which is the focus of this study, refers to patients’ trust in the individual health care provider. Such interpersonal trust has been defined as the optimistic acceptance of a vulnerable situation in which the patient beliefs that the physician will care for the patients’ interests [9]. Based on earlier research, Hall et al. proposed the following dimensions of patients’ trust in their physician: (1) ‘fidelity’, which means caring for the patients’ interest and avoiding conflicts, (2) ‘competency’, meaning good practice and good interpersonal skills and avoiding mistakes, (3) ‘honesty’, meaning telling the truth and avoiding intentional falsehoods, (4) ‘confidentiality’, which is the proper use of sensitive information, and (5) ‘global trust’, the so-called ‘soul of trust’ [9].

How can patients’ trust in their physician be measured? In 1990, the ‘Trust in Physician Scale’ [10] was published, and at that point in time, systematic research on patients’ trust in their physician started in the United States [2], [3], [12], [13]. In 2002, Hall et al. [13] argued that in existing questionnaires the conceptualization of trust was inconsistent, the initial item candidate pool was very small, the questionnaires were only tested on specialized populations, and no questionnaire had ideal psychometric properties. Therefore, they developed the ‘Wake Forest Physician Trust Scale’ (WF), using former questionnaires and research. Their ten-item questionnaire refers to the following four dimensions: ‘competence’ (three items), ‘honesty’ (one item), ‘fidelity’ (two items) and ‘global trust’ (four items). Items from the dimension ‘confidentiality’ performed too poorly to be included in the scale. A principal factor analysis with a varimax rotation revealed a two factor structure. However, due to the lack of a plausible interpretation of the factors, a high intercorrelation between the factors (r = .72) and the content of the factors (one factor contained only items worded in a positive direction, and the other factor contained only items worded in the negative direction), the authors concluded that the questionnaire was best explained by a unidimensional construct. This is congruent with findings of other studies on interpersonal trust questionnaires [2], [3], [12], [13], but contrary to theory [10] which would predict trust to be multidimensional. Hall et al. [13] compared their questionnaire with three other widely used questionnaires (‘Kao questionnaire’ [14]; ‘Trust in physician scale’ [11]; ‘Safran questionnaire’ [2]) and concluded that their instrument would be the first choice with regard to psychometric qualities. Its advantages are the good internal consistency (α = .93), the good test–retest reliability (r = .75) and the distribution: it is less skewed than other questionnaires. The WF now is frequently used for trust research by Hall and colleagues [15], [16], [17].

In earlier research on trust patient characteristics were inconsistently associated with patients’ trust in their physician. For example, patients with better health tended in some but not all studies to have more trust in their physician than patients with poorer health [6], [15]. Similarly, older patients tended in some but not all studies to have more trust in their physician than younger patients [17], [18], [19].

A strong association between patients’ trust in their physician and patients’ satisfaction with their physician has been found [10], [15]. Although an overlap between the concepts of trust and satisfaction is apparent, they differ with regard to the reference point of time: satisfaction is supposed to refer to past actions, while trust refers more to forward looking expectations [9]. Another variable that has shown a strong and stable association with trust in the physician is the duration of the patient–physician relationship. The longer the relationship, the greater the trust of the patient in his/her physician [3], [12], [15], [19]. Trust is developed through interaction and experience and continuity of care may provide patients with the time necessary for interpersonal trust to develop [12]. On the other hand, this association may reflect that only patients remain with their physician who trust in them. The willingness to recommend one's physician to others and to switch one's physician also showed a strong association with interpersonal trust. These intentions are often used to validate trust scales [13], [14], [16].

Surprisingly, in the Netherlands, no questionnaire on trust in a physician exists yet. Therefore, the aim of the present study was to investigate the psychometric properties of a Dutch version of the WF (WF-D). More specifically, to assess its dimensionality, reliability and validity.

It was hypothesized that the questionnaire shows similar psychometric qualities to the original version [13]. First, the structure of the questionnaire was supposed to be best explained by a unidimensional construct. Secondly, good homogeneity and internal consistency were expected. Thirdly, to test the instrument's validity, a high correlation between patients’ trust in their physician and (1) patients’ satisfaction with their physician, (2) willingness to recommend the physician to others, (3) willingness to switch physician and (4) the number of previous meetings with the physician was expected.

In addition, the dimensional structure of the instrument was examined with LISREL in order to find out whether the theoretically assumed four factors could be found in our data. Exploratively, differences in trust with regard to patients’ gender, age, educational level and self-reported health status were investigated.