Introduction

Recent United Kingdom healthcare policy has encouraged patients to ask questions of practitioners and to take an active part in decisions regarding their treatment.1,2,3 Such encouragement comes amidst a climate of increasing consumerism and questioning of professional dominance in healthcare encounters.4,5 The paternalistic model, underpinned by clinician-led decision-making and a presumption that the clinician knows what is best for the patient, is argued to have been largely supplanted.1,4,5 Alternative models of the practitioner-patient relationship include the 'professional as agent' model and the 'informed patient' model, the latter contending that good condition-related knowledge can empower patients to take control of the treatment decision making process and further that patients are desirous of an active role in this process.6 Research on patients' decisional role preferences, however, presents a complex picture, with some patients wishing to become actively involved in treatment decision-making, whilst others prefer to adopt a passive role in this process.7,8 Furthermore, preferences are individualistic – demographic variables have not, for example, been found to be a reliable predictor of patients' preferred level of participation.7,8 It has been suggested that matching patients' decisional preferences with their actual experience may enhance psychosocial outcomes.9 The need for clinicians to identify patients' preferred decisional roles – and for tools to facilitate this process – is therefore apparent.

Patients' treatment decision-making role preferences have been widely studied in the context of medical care, particularly the fields of cancer and surgery.6,7,8,9,10,11,12,13,14,15,16,17 Whilst there has been increased emphasis of late in dentistry on prevention, many patients will still require dental treatment, thus exploration of patients' preferred level of involvement in treatment decisions is of relevance in dental care. There have, however, been only a small number of studies exploring elements of the decision-making process in dental care18,19,20 and, to date, none has specifically addressed the extent to which patients would wish to participate in making treatment decisions. This is surprising, as patient participation is a field which has both ethical and legal implications in an increasingly user-focused, 'consumerist' health service. It is particularly pertinent in dentistry, given that most dental care is paid for, in part or whole, by the patient.

Some elements of the dentist-patient relationship, notably satisfaction, have been explored. Corah et al. studied patients' satisfaction,21,22 suggesting links between this variable and patients' service use.23 However, this work is now dated. More recently, satisfaction with the consultation has been linked with patients' ratings of treatment outcome.24 There is a dearth of literature regarding other aspects of the dentist-patient relationship, including the role that dental patients prefer concerning treatment decision-making. The few papers referring to the dental situation offer somewhat differing perspectives. Redford and Gift19 showed that dental patients are often excluded from treatment decision-making, whilst by contrast, Kay and Blinkhorn20 found that dentists are aware of the importance of patient preferences when making treatment decisions.

A number of methods have been used to examine patients' decisional role preferences. One, the Autonomy Preference Index,10 involves giving patients three examples of increasing ill health and treatment decisions associated with them. Patients are asked to assign control for the management of these three conditions to 'physician alone', 'mostly the physician', 'the physician and patient equally', 'mostly the patient' or the 'patient alone'. A later development was the Control Preferences Scale.7 This methodology involves presenting individuals with five cards, each with a written statement (see Box 1)boxed-text. These cards describe increasing levels of patient involvement in treatment decision-making, from the patient completely relinquishing control to clinicians, through to the patient maintaining complete control of treatment decision-making. Patients are asked to choose their preferred role from all possible pairings of cards so that an hierarchy, or preference order, can be constructed. From these hierarchies, it can be determined whether the patient prefers a passive, collaborative or active role. The Control Preferences Scale7 was developed for use in cancer care, and has been used with several cancer groups.11,12,13,14,15 It has also been successfully transferred outside of the cancer field,7,16 but has not as yet been applied in dental settings.

Study aims, design and methods

Aims of the study

The aims of this study were:

  • To assess the transferability of the Control Preferences Scale7 to a dental setting.

  • To explore patients' preferred and perceived roles in dental treatment decision-making

Study site and sample

The study was conducted at two sites – the University Dental Hospital of Manchester (UDHM) and a general dental practice in Cheshire. A convenience sample of twenty patients was recruited to the study from each site, giving a total sample size of N=40. Data were collected over a two-week period at each site. Patients were all under the care of a single dentist in the general dental practice and were recruited from clinics at the UDHM. All patients were asked to provide written consent before being involved in the study and each patient was given an information sheet outlining the nature and purpose of the study.

Study design and methods

The study employed a cross-sectional survey design, with face-to-face data collection. The instrument used was the Control Preferences Scale, CPS7 (see Box 1).

The CPS enables identification of a role preference hierarchy for each respondent ie an order of preference from most preferred to least preferred role. The hierarchy of role preferences was obtained using a paired comparisons approach. This involved respondents making choices as to the more preferred of pairs of cards. The cards can be presented in a number of ways (eg randomly, fixed order, all possible pairings presented sequentially). In keeping with other UK-based studies12,13,16 random presentation was employed. This is less repetitious than sequential presentation of all pairs and eliminates the possible introduction of bias which exists if the fixed order approach is used. By the end of the procedure, the five cards were ordered in a pile from most to least preferred and this preference order was recorded. Individuals were then classed as having an active, collaborative or passive decisional role preference, according to their most preferred card (card A or B = active, card C= collaborative, card D or E = passive). Once the role preference hierarchy had been obtained, respondents were asked to give a rationale for their role preference (after Caress et al.16) and this was recorded verbatim. The respondent's perceived role was also identified (see Box 2). This was done by using a 'pick one' approach, respondents being required to select the single card which most closely approximated their perceived role in treatment decision-making; the rationale for their choice was again explored. Supplementary questions regarding rationale for selection of most preferred and perceived role were asked in a standardised format. The procedure took approximately 10 minutes per patient. These procedures have been applied successfully in previous studies11,12,13,14,15,16 and the researcher is present throughout, which ensures that respondents can obtain clarification of any queries.boxed-text

Data analysis

Data from the Control Preferences Scale can be analysed in a number of ways. Degner et al7 describe these in detail. The simplest approach, adopted here, is to extract the most and least preferred roles from each patient's preference order, then to do a frequency count for each role. A simple frequency count was employed for perceived role. In keeping with other similar work,16 non-numerical data relating to patients' rationales for choice of role preference were content analysed to enable identification of themes.

Coombs25 demonstrated that a number of 'dimensions', which are best thought of as a 'shared world view' amongst a sample of respondents, could be identified within a body of preferential choice data. With five options (represented here by the five sort cards), there are sixty possible 'dimensions' which could arise. Previous theoretical work (described in detail by Degner et al.7) had led to the hypothesis that a single 'dimension', in which the distribution of role preferences within a sample lies along a continuum from active through collaborative to passive role preference, would predominate. Other competing 'dimensions' which could underpin the data include, for example, one where role preferences were not seen as a continuum, but rather are viewed in terms of 'collaborative roles (C,D,E) versus extremes (A and E)'. Alternatively, it is possible that no single underlying dimension underpins the data ie that respondents' role preference orders are random, rather than systematic.

With five sort cards, there are 120 possible ways in which individual role preferences can be ordered from most to least preferred role (eg ABCDE, CDEBA, EDCBA etc). Coombs25 established that only 11 of these combinations will be consistent with a particular 'dimension'. This is described as the preference order being 'transitive' for that 'dimension'. All other orders are 'intransitive' for (ie inconsistent with) that particular dimension (but may, however, be transitive for another dimension). The eleven 'transitive' orders for the hypothesised 'active-collaborative-passive dimension' are ABCDE, BACDE, BCADE, BCDAE, CBDAE, CDBAE, CDBEA, CDEBA, DCEBA, DECBA and EDCBA.

Coombs25 has very stringent 'goodness of fit' criteria which must all be met in order for the existence of a single underlying dimension to be inferred. These are that 50%+1 of the preference orders must be transitive; that all the transitive orders must be represented within the sample; and that the 'mirror image' orders which are found at the start and the end of the dimension (in this case ABCDE and EDCBA) must be present.

Analysis using Coombs' Unfolding Theory25,26 was therefore undertaken to establish whether a single underlying dimension of decisional role preference (ie 'shared world view') existed within the role preference data and, if so, whether it was consistent with Degner et al's7 hypothesised 'active-collaborative-passive' dimension.

Results

Characteristics of the sample

A convenience sample of 40 patients (20 from each site) was obtained. No individuals declined to take part in the study and no individuals were excluded. Table 1 presents the characteristics of the sample. There was a broad age range at both sites. The gender distribution was fairly even, as would be expected in a sample of dental patients. Both samples were relatively highly educated. This is likely a reflection of their locations – the UDHM is situated on a university campus (but – as is reflected in the sample – also serves the local community, which is socially deprived) and the general dental practice is within a largely affluent suburban area. The samples at the two sites were generally comparable.

Table 1 Table 1
Full size table

Acceptability of the CPS

All 40 patients appeared to understand the cards and their applicability to dental situations was confirmed. No further revision of the cards was necessary. The concept of decisional role preference was understood by all participants and each was able to identify their role preferences. The paired comparisons approach was acceptable, though some respondents required clarification – the advantage of the face-to-face data collection approach adopted for the study was apparent at such times.

Most preferred roles in treatment decision-making

Figure 1 presents the most preferred roles for the sample. At both sites, more respondents selected the collaborative role (card C) as most preferred than any other role. Only one individual at each site selected Card A, the fully active role, as their most preferred. In the general dental practice, most preferred roles were distributed fairly evenly across the three types of role (ie active, collaborative and passive), whereas at the hospital site the active role type was less commonly preferred than the collaborative or passive types.

Figure 1
figure 1

Distribution of most preferred roles (N=40)

Full size image

Least preferred roles

Figure 2 shows the distribution of least preferred roles in the sample. The two extreme choices 'A' and 'E' were overwhelmingly the least preferred at both sites, with the fully passive role (card E) being particularly unpopular.

Figure 2
figure 2

Distribution of least preferred roles (N=40)

Full size image

Preference orders for the sample

Table 2 shows the preference orders for the total sample. There was a range of preference orders, from the most active (ABCDE) to the most passive (EDCBA). In the hospital sample, the most common preference order was CBADE, a 'collaborative-active' preference order. By contrast, the most passive order (EDCBA) was the one most commonly selected at the general dental practice.

Table 2 Table 2
Full size table

Table 3 outlines the extent to which Coombs'25 goodness of fit criteria were met for the total sample and at each of the two sites. Independently, neither of the two sites' data met Coombs'25 criteria, hence it was not possible to infer the existence of the underlying dimension hypothesised by Degner et al.7 However, the data for the sample as a whole did meet Coombs'25 criteria for the hypothesised 'active-collaborative-passive' dimension.

Table 3 Table 3
Full size table

Perceived roles in treatment decision-making

Figure 3 shows the distribution of perceived roles for the sample. The majority of patients at both sites perceived themselves as attaining a passive role in treatment decisions – this was particularly marked in the hospital sample, with 17 out of 20 of these patients considering that they had little or no involvement in treatment decisions. No one considered that they had attained the most active role ('A') in treatment decision-making.

Figure 3
figure 3

Distribution of perceived roles (N=40)

Full size image

Discrepancy values

It is possible to calculate the extent of the discrepancy between an individual's most preferred and perceived decisional roles. This is done by assigning each role a number ('A' = 1, 'B' = 2, etc), then subtracting the perceived role from the most preferred role. Thus, if the patient's preferred role was 'A' and their perceived role was 'E', the discrepancy value would be 1 – 5 = -4.

Table 4 presents the discrepancy values for each site. As can be seen, only half the patients from the general dental practice and slightly less than half the patients at the hospital site reported complete congruence between their most preferred and perceived roles. At the hospital site, all instances of a discrepancy involved patients feeling less active in treatment decisions than they would have preferred. At the general dental practice, 6 patients reported this type of discrepancy, but 4 indicated that their level of participation was slightly more active than they would ideally have preferred.

Table 4 Table 4
Full size table

Only one individual, from the hospital site, had the maximum discrepancy value of -4. This patient preferred to be totally active ('A'), but her perceived role to be completely passive ('E'). This patient was interviewed through a translator and she felt that her extreme choices were due to a number of factors. She wanted to be totally in control due to bad past experiences, but felt that due to her language difficulties and feeling intimidated, she actually assumed a very passive role with her dentist.

Analysis of verbatim data

Content analysis of the verbatim data regarding patients' rationales for their role preference revealed the following themes:

  1. 1

    Knowledge of subject

  2. 2

    Trust

  3. 3

    Consumerist stances

  4. 4

    Time constraints

  5. 5

    Factors that would alter the individual's role preference

Knowledge of the subject

Eighteen of the 20 patients interviewed at the hospital site mentioned lack of knowledge of the subject of dentistry as influencing their ability to participate in treatment decisions, including the one person who chose 'A' as their most preferred role. All of the seven hospital site patients who chose card 'A' (most active) as their least preferred role cited lack of knowledge about dentistry as a reason for their choice.

At the general dental practice, lack of knowledge was mentioned by those patients who most preferred a passive role (N=7) or a collaborative role (N=7), in all instances as something which disempowered them from participating in the decision-making process.

Typical comments were:

'I don't know the science behind dentistry, so I'll leave the decision to someone who does.' (Hospital patient)

'I don't know as much as the dentist about dentistry!' (Hospital patient)

'I am paying to see the expert/professional, therefore should consider his opinion, since I don't know much about dentistry.' (General dental practice patient)

'[It's] up to him to tell me what has to be done because I don't know what is wrong.' (General dental practice patient)

Trust

There was an interesting contrast between the hospital and general dental practice with respect to patients' views on trust. Patients from the hospital site typically mentioned trust in a positive context as a contributor to a passive role preference. Patients from the general dental practice, however, typically cited trust as a reason for their disliking card E (the most passive role option).

Trust in the dentist was specifically referred to by nine of the 20 hospital patients. Of these, three chose 'C' as their most preferred role, four chose 'D' and two chose 'E'. This suggests that there may have been a relationship between trust in the dentist and passivity in the decision-making process.

Typical comments were from the hospital patients were:

'The dentist is a professional, therefore you should trust him.'

'I trust his judgement because he knows best.'

'If you can't trust the dentist, there's something wrong!'

Lack of knowledge about dental care appeared to be closely linked with trust for these patients – they typically considered that the dentist's professional status resulted in their being highly knowledgeable, which engendered trust.

By contrast, lack of trust in the dentist was the typically cited rationale amongst those patients from the general dental practice (N=12) who least preferred the fully passive role (card E). Typical comments made by these patients were:

'I think he can make the wrong decision.'

'I don't want him to tell me anything, I don't trust him enough.'

'I wouldn't like to leave everything up to him, he could mess everything up.'

Bad experiences with previous dentists and perceptions based on media reports of dental malpractice or deaths during dental treatment appeared to have contributed to these patients' lack of trust.

Consumerist stances

Comments suggestive of a consumerist stance were made by 11 patients from the hospital site and eight from the general dental practice. In all instances, these were patients who preferred an active or collaborative decisional role.

Typical comments relating to this theme were:

'I have a right to be involved in decisions about treatment for me.' (Hospital patient)

'This is my mouth so I have a right to say what's done with it!' (Hospital patient)

'The dentist cannot ethically make a decision without my informed consent.' (Hospital patient)

'It's the day of informed consent; I have to be involved in making decisions.' (General dental practice patient)

'The dentist has to consider my opinion, because I know what is right for me.' (General dental practice patient)

'I am responsible for my own dental care.' (General dental practice patient)

Time constraints

A further common theme amongst the hospital patients was lack of time for discussion. Ten patients of the 20 hospital patients cited lack of time as a reason why, with their 'usual' (non-hospital) dental practitioner, their perceived role was more passive than the role they most preferred. Typical comments were:

'The dentist has more time constraints so there isn't time for more dialogue.'

'There isn't enough time for the dentist to really consider my opinions.'

'The dentist just needs to get my treatment done as quickly as possible.'

'There's never enough time to sit and discuss everything.'

However, hospital patients typically did not feel constrained by time during their consultations at UDHM:

'These are specialists so you feel that you can ask them more questions.' 'This is a teaching hospital so more time is given to you explaining all the options – the students need to learn all the options available.'

'Because they are students the appointments are longer and there's more time to sit and discuss things.'

By contrast, patients from the general dental practice seldom raised time constraints as a barrier to participation in treatment decisions. This may be because they were satisfied with the time available for consultations. A possible alternative explanation is that, unlike the hospital patients, they did not have a comparator against which to assess the adequacy of consultation length.

Factors that would alter the individual's role preference

Although patients were quite clear about their preferred and perceived roles, some did mention specific situations in which their role preference would change. At the general dental practice, eight patients identified situations where their role preference would change – in all instances to a more active role. Loss of confidence in the dentist was the typical factor cited by these patients, as the following data extracts illustrate:

'If I thought his treatment was not to the standard I expect I would want more of a say.'

'If I lost confidence in my dentist because I thought he had made a mistake I would not rely on his opinion as much [or] if I thought he saw his patients too quickly and talked to his dental nurse more than he talked to me.'

An increase in the patient's own confidence was also cited by one general dental practice patient:

'If I became more confident, less scared of the dentist I would like to have more of a say.'

Four of the hospital patients gave examples of situations in which they would wish to adopt a more active decisional role than usual. Factors cited as leading to this were pain and the nature of the proposed treatment, as the quotes below illustrate:

'If I had to have dentures I'd actively refuse.'

'If I was in a lot of pain I'd tell the dentist what I wanted done – even if he could save the tooth I'd have it taken out.'

Two hospital-based patients, conversely, highlighted situations in which they would prefer a more passive role than usual, in both cases due to severity of their condition:

'If I was in severe pain I would just let the dentist decide so that I could be out of pain as soon as possible.'

'I'm a haemophiliac so the more serious my condition the more I have to let the dentist decide for me.'

Discussion

This study aimed to assess whether dental patients had identifiable decisional role preferences and to explore the transferability of Control Preferences Scale.7 Previous studies using the Control Preferences Scale7 have mainly been with patients suffering from cancer or other serious illness.11,12,13,14,15,16 However, the methodology worked well in the dental context and gave rise to interesting insights which are of great relevance to dental practitioners. This would suggest that this extremely useful methodology is transferable to dental settings.

Coombs'25 goodness of fit criteria for inference of a single underlying dimension in a body of preferential choice data are very stringent.7,27 It has consequently been argued that these criteria are difficult to meet in small samples.27 The data from the total sample (N=40) did, however, meet Coombs'25 criteria and the existence of the 'active-collaborative-passive' dimension hypothesised by Degner et al.7 was confirmed. Establishment of unidimensionality as hypothesised by Degner et al.7 supports the transferability of the Control Preferences Scale to a dental setting.

This group of dental patients was comparable with other types of patients13,16 in that they predominantly opted for collaborative and passive decisional roles. Some of the contributors to passive role preferences appeared to be 'disempowering' factors, eg lack of time and lack of knowledge. However, in other patients, there appeared to be more of a positive choice to defer to professionals, eg due to trust. This is an important distinction. The disempowering factors may be amenable to intervention. For example, Neufeld et al.17 found that increasing patients' knowledge of available options increased their willingness to participate in treatment decisions. Likewise, it has been suggested that clinicians who offer patients longer consultations should be rewarded, since this may contribute to enhanced patient participation.28 It would be very interesting to see if this model of service provision was suitable in dentistry, where most practitioners operate on a 'fee for service' basis. Awareness of this distinction between 'positive choice' and 'disempowerment' is also important in the context of current policy initiatives. If some individuals are making a positive choice to defer to professional expertise, is the uniform promotion of patient participation an appropriate model? This aspect merits further attention in dental care, particularly since studies in this area suggest that the nature of the dentist-patient relationship may impact on treatment decisions.19,20

Some individuals, however, are desirous of higher levels of participation in decision-making. Analysis of patients' rationales for their role preference identified a theme related to consumerist stances. Patients are becoming increasingly aware of such matters as patients' rights. The imbalance of power that has previously favoured healthcare professionals appears to have changed.1,2,3,4,5 The National Health Service, particularly, has been striving to become more user-friendly and 'consumerist'.1,2,3 Ethical, legal and social reasons for this change have been described4,5,6 and recent United Kingdom healthcare policy1,2,3 has effectively aimed to encourage 'consumerism' in the health service and has drawn patients' rights to the public's attention.5 This is particularly relevant to dental practice where every patient is a paying customer and perhaps therefore more aware of their rights as a consumer than in other healthcare settings. Exploration of the influence of payment on patients' decisional role preferences offers exciting research potential.

For some patients, there was a discrepancy between the individual's preferred and perceived roles. Typically, where discrepancies occurred, patients felt that their perceived role was less active than their preferred role. Other similar work12,13,16 has suggested that professionals appear to be offering a uniform passive decisional role to patients. It would appear that this was also the case in the present study. Such a situation would be consistent with a 'paternalistic' model of provider-recipient relationship, rather than the more collaborative role advocated in recent literature and healthcare policy.1,2,3,4,5

For some patients, there were circumstances under which role preference would change. Such factors as pain and severity of the condition appeared to increase the likelihood of patients preferring a passive role, whilst loss of confidence in the dentist appeared to increase patients' desire for participation in treatment decisions. These data would suggest that role preference is not a static entity, but rather develops and changes according to circumstances and experiences. Longitudinal work exploring the evolution of role preferences over time, particularly in patients requiring long-term treatments, would be worthwhile.

Conclusions

This study introduces the concept of patients' role preferences in relation to dental treatment decision-making. In this study, the collaborative role was most commonly preferred, but a passive role was more commonly perceived as the one attained. The study suggests that a well-established methodology for exploring patients' decisional role preferences – the CPS – is transferable to a dental setting and has the potential to facilitate identification of patients' role preferences in this field. Identifying and accommodating patients' role preferences has great relevance to dental practice. Further application of this approach with larger samples would be beneficial, as would further exploration of the factors which influence dental patients' decisional role preferences.