Letters

Patient centred care of diabetes in general practice

Doctors and nurses must understand meaning of "communication"

Study failed to measure patient centredness of GPs' consulting behaviour

Might difference in prescribing explain some of the findings?

Authors' reply

Doctors and nurses must understand meaning of "communication"

EDITORIn Kinmonth et al's randomised controlled trial of patient centred care of diabetes one group of health professionals (the intervention group) received training in patient centred care while the other (the comparison group) did not.¹ Subsequently, scores on quality of communication with patients favoured the intervention group, but scores on patients' knowledge about their disease favoured the comparison groupto put it crudely, patients assigned to the intervention group were relatively happy but ignorant.

Sadly, the straightforward paradox of saying, in effect, "he communicates well but tells you nothing" is all but lost in the way that the word communication is used in medical education these days. Indeed, in a manner that many medical educators would find unexceptionable, Kinmonth et al gloss this term as meaning "the ability to tell the doctor or nurse personal or troubling things and feeling understood"as if the transfer of information from professional to patient was something else. Where does this content free definition come from?

Perhaps the problem is that the whole communication skills movement arose from a perception that medicine (if not nursing) had insufficient empathy; the resultthe common, lazy identification of empathy with patient centredness, and of both with communicationtherefore has its own historical logic. But we cannot propagate a view of communication that suggests that "just to be there is enough."

We need a more grown up and integrated view of what is at stake. Put simply, health professionals have skill: if they didn't, patients wouldn't visit them. It is hardly patient centred to act as if things were otherwise, or rational to deny that it is knowledge that people seek to communicate.

John Skelton, Senior lecturer in communication skills. 
Department of Primary Care and General Practice, Medical School, University of Birmingham, Birmingham B15 2TT j.r.skelton{at}bham.ac.uk

Study failed to measure patient centredness of GPs' consulting behaviour

EDITORKinmonth et al should not claim that their study "shows the power of a consultation to affect patients' health and wellbeing."¹ They evaluated some of the outcomes of training health professionals in patient centred clinical method but failed to measure the patient centredness of general practitioners' or nurses' consulting behaviour. Consequently, we have no idea whether the training actually resulted in clinicians providing more patient centred care; the study provides circumstantial evidence about the influence of patient centred clinical method on clinical outcomes.

Before we can accurately investigate the effects of patient centred care we need reliable instruments for measuring the relevant aspects of doctor-patient communication.² Although several communication coding systems exist, few have been validated adequately for their intended uses.³ One of these research tools was designed to measure clinicians' patient centredness,⁴ and there is some evidence of its reliability⁴ and validity.⁵

Further development of this instrument will clarify its strengths and weaknesses for measuring clinicians' patient centredness in routine consultations in primary care. This would enable those investigating the effects of patient centred care to use it to monitor doctor-patient communication. Although this is methodologically challenging and likely to be costly, it could ultimately lead to trials providing more relevant information. If Kinmonth et al had used this approach they would perhaps have produced hard evidence to relate their unexpected findings to clinicians' consulting behaviours.

Tim Coleman, Clinical lecturer. 
Department of General Practice and Primary Health Care, University of Leicester, Leicester General Infirmary, Leicester LE5 4PW

Might difference in prescribing explain some of the findings?

EDITORKinmonth et al's study has important implications for general practice as the authors suggest that practitioners who concentrate on being patient centred might be at risk of losing the focus on disease management.¹ One of the findings that supports this conclusion is that patients in the intervention group gained weight.

In the discussion, the authors suggest that this difference between the two groups could not be accounted for by small differences in prescribing rates of hypoglycaemic agents (50% in the intervention group and 46% in the comparison group). There were, however, larger differences between the groups in the type of hypoglycaemic agent prescribed. In the intervention group 47% of the patients were prescribed a sulphonylurea and 11% a biguanide. In the comparison group the figures were 40% and 15% respectively. Given that sulphonylureas tend to cause weight gain and biguanides promote weight loss, might not some of the difference in weight be accounted for by this difference in prescribing?

Christopher Hand, Adviser in postgraduate general practice, Anglia and Oxford Region. 
School of Health Policy and Practice, University of East Anglia, Norwich NR4 7TJ c.hand{at}uea.ac.uk

Authors' reply

EDITORDefining and measuring patient centred consulting and its outcomes are problematic, as our study shows. We defined patient centred consulting as an integration of the individual patient's perspective with the practitioner's general knowledge of disease, recognising that doctors and nurses do not deal with diseases alone; they deal also with individuals who are ill or concerned about their health.

Skelton argues for the primacy of communication of medical knowledge in the consultation, but our assertion is that empathy is the means by which the knowledge possessed by the practitioner becomes applicable by the patient. The prevailing approach to the management of patients with newly diagnosed type 2 diabetes remains one of "downloading information," despite evidence of its relative ineffectiveness in enabling self care.¹ Approaches taking account of the particular situation and feelings of the individual patient seem more likely to support the integration of information into perception and action.² But teaching this integrated approach did not prove easy and deserves further effort.

Coleman suggests that we should have directly observed the patient centredness of practitioners' consulting behaviour. Evidence is emerging, however, that it is patients' views of the process of the consultation, rather than researchers' perceptions of it, that are key predictors of health outcomes. ^{2 3} We observed consultations during pilot work⁴ and subsequently measured the effect of training on practitioners' attitudes, behaviours, and style of care through patients' and practitioners' reports.⁵

Weight was higher in the intervention than the comparison group, but only among the 112 (48%) patients prescribed hypoglycaemic drugs. Hand wonders if this is because of a 7% excess of sulphonylurea prescribing (which might promote weight gain) in the intervention group and a 4% excess of metformin prescribing (which might promote weight loss) in the comparison group. We also wondered about this, but adjustment for category of drug in a regression analysis made no impact: the difference between groups in body mass index at one year was 2.00 kg/m² after adjustment for sulphonylurea and 2.12 kg/m² after adjustment for metformin compared with 2.04 kg/m² before adjustment. No measure was made of adherence to the prescribed drug in the study. Without such a measure we cannot tell whether the training given to practitioners produced the reported difference in outcomes through greater adherence to the drug in the intervention group.

Ann Louise Kinmonth, Professor of general practice. ^a
Simon Griffin, University lecturer in general practice. 
General Practice and Primary Care Research Unit, Institute of Public Health, Cambridge CB2 2SR

Alison Woodcock, Lecturer. 
Department of Psychology, Royal Holloway, University of London, Egham, Surrey TW20 0EX

Mike Campbell, Professor of medical statistics. 
School of Health and Related Research, Northern General Hospital, Sheffield S5 7AU On behalf of the Diabetes Care from Diagnosis Research Team

^a alk25{at}medschl.cam.ac.uk