Considerations of ‘fit’ and patient involvement in decision making

Several papers in this issue of Health Expectations consider and raise questions about the ‘fit’ between the health care that people need or prefer and the health care that services provide. Most obviously, Gould et al. highlight a lack of fit between what Canadian health services provide to women with breast cancer and the concerns of women who are diagnosed with breast cancer before the age of 45.1Current breast cancer services, for example, are not geared to answer the questions being asked by women who are pre-menopausal and want to know about the likely effects of treatments on their fertility and sexuality. They also fail to provide appropriate information and support to women who are at busy stages of their working lives and possibly financially insecure, and who may be struggling to care for young or adolescent children and to communicate with them about their disease, treatment and prognosis. It is encouraging to hear that the Canadian Breast Cancer Network is developing a national strategy to address the gaps in services that were highlighted by this study. However, questions do need to be asked about how far publicly funded health services should go to achieve a good fit with the preferences that different groups and individuals hold in relation to different aspects of care. Two other papers in this issue report on a domain in which the importance (or even appropriateness) of tailoring health care to fit patients’ expressed preferences is a matter of some dispute: patients’ involvement in treatment decision making. Carlsen et al. examined the fit between patients’ and their general practitioners’ (GPs) orientations towards sharing information and decision making.2They found no relationship between this fit and patients’ satisfaction with their GPs. However, those GPs who were more oriented to share information and decision making had more satisfied patients. The findings from this Norwegian study are broadly consistent with those of an earlier American study on which it was based.3 Krupat et al. had found that patients were more likely to report maximum satisfaction with their physicians if their physicians were equally or more inclined than they were to deem it appropriate for patients to access information and be involved in treatment decision making. Krupat et al. identified two plausible explanations for these findings: either physicians who endorse statements associated with sharing information and decision making ‘exhibit a style that is satisfying because it is open and sharing’ and ‘win over’ even those patients who report differing beliefs about these issues, or they are ‘more sensitive to the perceived needs of their patients, … are good at identifying the extent to which their patients want to receive information and be involved in decision-making… and use a flexible style in which they adapt to the needs of their patients’. Carlsen et al. favour the second explanation, and another study published in this issue of Health Expectations perhaps lends it some support. Ford et al. asked the patients of 13 GPs in England to indicate their preferred level of involvement in decision making and to report the level of involvement they had during a particular consultation.4 They were able to divide the GPs into two distinct groups according to the proportions of their patients who reported a level of involvement that matched their preferred level. When they analysed video recordings of the consultations, they found that the GPs in the group with higher proportions of patients reporting levels of involvement congruent with their preferences had not only higher mean scores for communication skills and effectiveness, but also a greater dispersion of scores, suggestive of a greater flexibility in consulting style. However, evidence about the implications of a match between patients’ preferred and achieved levels of involvement in decision making is somewhat mixed. For example: Ford et al. found that those patients whose self-reported level of involvement in their GP consultation matched the level for which they expressed a general preference (at the same time point) were no more ‘enabled’ than those for whom there was not a fit.5 Gattellari et al. found that a lack of fit between a preferred decision-making role expressed prior to a cancer consultation and role played in decision making as self-reported 2 weeks later predicted changes in anxiety levels between the two time points but not satisfaction with the consultation.6 Lantz et al. found that women for whom there was a match between their preferred decision-making role and the role they played in a breast cancer treatment decision reported greater satisfaction with the decision than those who were classified as having played more active or more passive roles than they had expressed a preference for.7 Hack et al. found that the congruence or otherwise of women's self-reported roles in a breast cancer treatment decision and their retrospectively expressed preferences for involvement in that decision was not associated with quality of life 3 years post-treatment.8 The story so far is that doctors with more positive attitudes to sharing information and decision making have more satisfied patients, and it seems plausible that they achieve this by adopting a flexible approach and tailoring their communication to suit individual patients’ preferences for involvement. However, it is not clear that patients whose self-reported roles in decision making match their self-reported role preferences are more satisfied or otherwise better off than those whose reported roles and role preferences do not match. It is worth taking a careful look at the evidence about the implications of a match or otherwise between patients’ roles and role preferences. Obviously, the studies outlined above explore relationships between this match and some quite different variables, and they vary in terms of the timings of their assessment of role preferences and roles played. These factors may account for some of the variations in the strength and direction of the associations reported. But a feature that the studies have in common might also warrant consideration. All of the studies assessed people's roles and role preferences by asking them to select from (or sort into preference order) similar sets of five descriptions of doctor and patient roles. These descriptions differ primarily in terms of who makes the decision (doctor, patient or both) and the extent to which they take the other's opinion into account. It is possible that patients’ preferences relating to the way doctors discuss treatment decisions with them may be more nuanced than these descriptions allow. Also, the considerations that lead patients to select a particular descriptor as best reflecting the role they actually play in decision making have been shown to vary,9–11 and the role descriptions might not adequately capture all the aspects of involvement that matter to patients. It remains quite possible that some kind of fit between the forms of involvement in treatment decision making that a patient experiences in a given situation and the forms that s/he prefers in such a situation is an important mediating variable between the doctor's communication behaviours and patient outcomes, but that the approaches that have been used to date to assess this fit have made the relationship difficult to see. So should health professionals be encouraged to tailor their communication in consultations to fit individual patients’ preferences for involvement in decision making? None of the evidence outlined above suggests that such tailoring would do any harm, and some would argue that it is required by the widely accepted principle of respect for people's (informed) choices about how they shape their lives. However, others would argue that this same principle supports the view that health professionals should be communicating with a view to helping (or perhaps even persuading) patients to play shared or otherwise active roles in decisions about their own health care – and/or that such communication should be encouraged because patients who play shared or otherwise active roles in these decisions have better outcomes. The evidence that is sometimes cited in support of this, however, also needs to be looked at carefully. For example, in addition to the findings outlined above, Gattellari et al. reported that patients who indicated they had played some form of shared role in decision making were more satisfied irrespective of their preferred role,6 and Hack et al. found that women who indicated they were actively involved in the treatment decision reported higher overall quality of life 3 years, again irrespective of their preferred role.8As the authors point out, however, neither study was designed to examine causal relationships between these variables. There is clearly significant scope for further conceptual, methodological and empirical work to shed new light on questions relating to patient involvement in decision making. Health Expectations will be pleased to receive articles that contribute to this debate – and to debates about the ways in which health services do or should strive to satisfy their users’ preferences relating to other aspects of care.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.