| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |
|
|
|
|||||||
|
|||||||||
TRACK to:
|
|
Electronic letters published:
![[Read Comment]](/icons/misc/sectionDOWN.gif){kind=icon}
Thoughts on taxonomy and theory
Where do we go next...?
|
|
|||
|
Paul S Little, Southampton Professor of Primary Care Research
Send response to journal:
|
I am delighted to be invited to provide a commentary on this paper. This is a very valuable paper, and the authors are to be applauded not only for their exhaustive search of the literature, but in so doing for providing reflection and a very interesting framework for discussion. I don’t think there is a great deal to add to the careful and thoughtful approach the authors have taken. However, a number of relatively minor points occurred to me from this paper. 1) Taxonomy of outcomes and which outcomes are important. The authors have provided valuable obervations in highlighting the paucity of ‘disease process’ outcomes. The authors also commendably highlight the importance for patients of their health experiences – and it is arguably these that are a central issue to address in the consultation, particularly in the context of the patient centred model? We certainly do have a long way to go – and the authors give cogent arguments for the next steps. However, I was very encouraged that across the range of studies, for patients experience of health outcomes nearly 80% of outcomes reported were positive, which makes it very likely that improving communication improves outcomes for patients. Secondly, the area of trials in communication makes for particular methodological problems in using ‘objective’ measures: I wonder how objective can ‘objective’ measures be in trials of patient communication given the difficulty of blinding patients and likely placebo and white coat effects on outcomes such as blood pressure, cholesterol and blood glucose? We also perhaps need to be aware that in making a dichotomy we remain alert to the interrelation between ‘objective’ and ‘subjective’variables rather better: for example psychosocial variables may well predict mortality as well as biological variables (e.g. depression after MI). Perhaps we might describe the outcomes as biomedical (because ‘disease process’ assumes we understand the disease process whereas often we do not very fully?) vs health experience? 2) Valuing process. Related to the last point, the authors commendably highlight the ‘process outcomes’ and their overlap with satisfaction. There is a key overlap between satisfaction and process variables: patients’ perceptions of a whole range of aspects of communication – the extent to which patients felt listened to, their concerns taken on board, how their lives have been affected - are closely related to satisfaction with the consultation, and closely relate to intention to comply with treatment. As the authors comment these are valuable in their own right and perhaps should have equal status with ‘biomedical’ variables such as blood pressure? 3) The use of theory. I agree with the authors that we need to explicit about theory, and use theory to inform interventions but also try and understand and measure both process and outcome. However, we must perhaps remain sensitive to the limitations of theory. We have the problem about which theory to use – e.g. should the Theory of Planned Behaviour or the Self Regulatory Model be used when thinking about patient behaviour? This is not an argument for jettisoning theory, but for both designing studies which do not stay narrowly defined to one theory, and/or designing studies to test alternative theories. Also what constitutes a theory? - at the moment we have a framework of ideals and descriptors for the consultation in the patient centred model, but this is not the less valuable for not being a theory. Our theories and their measurement also only do so well - even the best psychological theories only explain a minority of the variance of behavioural outcomes. This means in addition to trying to use and improve theory, and the measurement of the theoretical constructs, we need to be aware of the potential limitations of theory, and value both empirical evidence and our aspirations for the consultation. 4) Negative outcomes. The authors highlight cost and show negative findings in slightly more studies than there are positive findings which is sobering. A recent study from our group also highlighted that empowering patients may lead to more unnecessary investigation1. This is not an argument for avoiding better communication nor the empowerment of patients, but that as supporters of the patient centred model we need to remain open to possible negative effects of improving communication, and how to avoid the pitfalls. These are all minor points and I would not wish to detract from a very useful, interesting and informative paper – thank you! Reference List 1. Little P, Dorward M, Warner G, Moore M, Stephens K, Senior J et al. A Randomised controlled trial of effect of leaflets to empower patients in consultations in primary care. BMJ 2004;328:441-4. Competing interests: None declared |
|||
|
|
|||
|
John G.R. Howie, Edinburgh, UK Professor of General Practice. emeritus, University of Edinburgh
Send response to journal:
|
It is a pleasure to be asked to comment on this predictably admirable review by Griffin and his team from Cambridge UK and London Ontario. Given the intellectual rigour displayed, it is tempting to say that there isn’t anything more to be said. But the paper ends with a challenge to advance the field (… of patient-centred care) ‘beyond a series of tantalising but disconnected and unconfirmed results’. The implication is that this should be through more and better theory-based clinical trials. Is the best or only way ahead? My first comment is that Griffin’s review starts with some 20,00 hits on key-words, but ends with only 35 ‘adequate’ trials from a sub-selection of 145. Of the 35, only a handful influenced health outcomes significantly. Maybe the clinical trial route is, for whatever reason, not - or not yet - a sufficiently productive way to go. My second observation is that this may because of the overwhelming influence of other confounding issues, and the complexity of their interactions. ‘Patient-centredness’ includes negotiating ‘patient preferences’, and demonstrating suitable clinical behaviours as well as increasing ‘patient participation’, and it has been documented that not all patients want to participate in decision-making either at any time or at all times. Finding suitable patients to randomise becomes so problematic that results become in danger of not being generalisable to the larger populations of interest to researchers, clinicians or patients. Perhaps the field is as yet too difficult to fit into a single theory, and there may still be a place for more use of descriptive work before returning to the trials’ arena.(1) My third comment is that perhaps other determinants of good inter- personal outcomes dwarf the variables presently being randomised. Candidates include the widely prevalent perverse incentive structures that militate against good inter-personal consulting practice, and the problem of diminished continuity of personal doctoring in modern multi- disciplinary primary care teams. It is highly possible that these influences will combine to defeat the balancing effects of randomisation in what are often fairly small numbers of patients available for study. It's a bit Kuhn-like I feel. Griffin and his team are absolutely right to emphasise the need for a good theory to work forward from. If we don’t yet have it, we first need to develop a model which is sophisticated enough to be meaningful, but simple enough to be useful. John Howie Professor emeritus, University of Edinburgh John.Howie2@btopenworld.com Reference: Howie JGR, Heaney DJ, Maxwell M. Quality, core values and the general practice consultation: issues of definition, management and delivery. Family Practice 2004;21:458-468. Competing interests: None declared |
|||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |
TRACK: