Family Practice Advance Access originally published online on January 31, 2006
Family Practice 2006 23(2):192-197; doi:10.1093/fampra/cmi123
General practitioners may diagnose type 2 diabetes mellitus at an early disease stage in patients they know well
The Research Unit and Department of General Practice, University of Copenhagen, Denmark
Correspondence to Thomas Drivsholm, The Research Unit of General Practice, University of Copenhagen, Center for Sundhed og Samfund, 5, Øster Farimagsgade, PO Box 2099, DK-1014 Copenhagen K, Denmark; Email: thomas.drivsholm{at}gpract.ku.dk
Received 6 July 2005; Accepted 28 December 2005.
Drivsholm T and Olivarius NF. General practitioners may diagnose type 2 diabetes mellitus at an early disease stage in patients they know well. Family Practice 2006; 23: 192–197.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Background. Continuity of care may not only save time and money for the health care system, but may also be beneficial for the individual patient. How well GPs know their patients may potentially lead to an early diagnosis of a slowly progressive chronic disease like type 2 diabetes mellitus (T2DM). The aim of this paper is to investigate this hypothesis.
Methods. A cross-sectional, population-based study of 1136 patients newly diagnosed with T2DM by their GP. Our main outcomes were how well the GPs' knew their patients (questionnaire) and centralized analysis of glycosylated haemoglobin A1c. Multivariate linear regression models were used to allow adjustment for confounding variables.
Results. GPs classified how well they knew their patients as being not well for 13.5% (153/1136) of their patients, as fairly well for 38.6% (438/1136) and as very well for 48.0% (545/1136). Patients whom the GPs classified as not knowing well had relatively high glycaemic levels compared with levels among other patients, a finding that was confirmed in multivariate linear regression models.
Conclusions. Our data show that the glycaemic level among patients whom the GP characterize as knowing well or fairly well is relatively low compared with among patients whom the GP characterize as not knowing well. We suggest that this reflects a late diagnosis in these patients, and that GPs should be especially aware of undiagnosed T2DM among patients whom they do not know well.
Keywords. Continuity of patient care, epidemiology, family practice, glycaemia, type 2 diabetes mellitus.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Continuity in the patient–doctor relationship has been considered to be the basis of GPs' work for three decades.1,2 According to Gray et al.3 the essence of a theory of continuity of care in primary care is that ‘a personal doctor with accumulated knowledge of the patient's history, values, hopes and fears will provide better care than a similarly qualified doctor who lacks such knowledge, and that the benefits of such continuity will include not only greater satisfaction for the patient but also more efficient consultations, better preventive care and lower costs’.
The concept of continuity of care has been reviewed recently by an international working group aiming at developing a common understanding ‘as a basis for valid and reliable measurements’.4 Three types of continuity were defined: informational, management and relational continuity of care. While the dimensions informational and management continuity of care relate to structural aspects (e.g. availability of information, visit patterns and a consistent and coherent approach in the organization), the relational (interpersonal) dimension is associated with more qualitative aspects of the patient–doctor relationship.4 Most measurement techniques and evidence of effects of continuity of care reflect structural characteristics rather than the relational dimension,5 although degree of relational continuity has shown stronger associations to desirable outcomes than measures of the two structural dimensions.6
As relational continuity and quality of care seem to be associated, it is plausible that good relational continuity will lead to an early diagnosis of a slowly progressive chronic disease like type 2 diabetes mellitus (T2DM). The aim of this paper is to investigate this hypothesis by analysing clinical and socio-economic characteristics of newly diagnosed diabetic patients according to how well their GPs knew them.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Study population
From March 1989 through February 1992, 474 GPs recruited all enlisted patients aged
40 years with newly diagnosed diabetes (n = 1381) for the study Diabetes Care in General Practice, a randomized controlled trial of structured diabetes care.7 More than 97% of the population in Denmark are listed with a GP in a tax-based, social insurance system.8,9 The diagnosis was suspected from raised blood glucose values and/or typical diabetic symptoms and confirmed by a single fasting whole blood (plasma) glucose sample 7.0 (8.0) mmol/l. Besides the 1381 patients enrolled into the study, 162 patients were excluded on the basis of protocol-based exclusion criteria.7 For the present analyses we excluded 47 patients treated with steroids, 197 patients diagnosed outside general practice and one patient with missing data regarding how well the GP knew him. Thus, the present analyses include 1136 patients, almost exclusively with T2DM.7
Methodology and definitions
At the time of the diabetes diagnosis the GPs completed a questionnaire including the following two questions: (1) How well do you know the patient? Possible answers: very well, fairly well or not well; and (2) How long (number of years) has the patient been listed at your practice?
Samples for diagnostic plasma or whole blood glucose tests were drawn after a minimum of 8 hours fasting and analysed at (n = 72) local laboratories. A factor of 1.15 was used for converting (n = 678) fasting whole blood measurements to plasma glucose (FPG). Centralized measurements of glycosylated haemoglobin A1c (HbA1c) were made at Odense University Hospital by ion exchange, high performance liquid chromatography (reference interval: 5.4–7.4%). Only samples taken up to 45 days after the diagnosis of diabetes were accepted (n = 1002).
Other methods, i.e. recordings by GPs, questionnaires filled in by patients, and definitions of antihypertensive drug treatment, complications, and micro- and macrovascular conditions, all recorded at the diabetes diagnosis, have been described in detail in prior publications.10,11
Statistical analysis
Univariate analyses were done by routine statistical methods. Two multivariate linear regression models were used to allow testing of the association between FPG and HbA1c at diagnosis (dependent variable), respectively, and how well the patient was known by the GP (independent variable) while adjusting for confounding variables, i.e. sex, age, BMI, educational level, macrovascular conditions, retinopathy, peripheral neuropathy, renal involvement and use of antihypertensive medication. FPG and HbA1c were logarithm transformed prior to multivariate analyses. To test whether how well the GPs knew their patients was a proxy for type of doctor, information of GP-identification was included as a random effect in the multivariate models. Statistical significance was defined as two-sided P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
GPs responded that they knew 48.0% (545/1136) of their patients very well, 38.6% (438/1136) fairly well and 13.5% (153/1136) not well. HbA1c was measured 7 days (median, inter quartile range 3–15) after diagnosis, while the questionnaire was completed after 8 days (3–19).
Among patients whom the GPs did not know well, levels of FPG and HbA1c were relatively high compared with levels among other patients (Table 1 and Fig. 1).
|
|
Patients whom the GPs did not know well had glycaemic symptoms more frequently (Table 1) and had had these symptoms for a longer period than patients known well or fairly well by their GP (Table 2).
|
The association between glycaemia and how well the GPs knew their patients was confirmed in multivariate linear regression models for both FPG (P < 0.05) and HbA1c (P < 0.01) (Table 3).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
In this cross-sectional study we found that patients whom their GP classified as not knowing well on average presented with relatively high glycaemic levels at the diabetes diagnosis. As T2DM is well documented as being a slowly progressive disease,12 this observation may indicate a relatively late diagnosis in these patients.13
Major strengths of this study are the large population-based patient sample, the centralized analysis of blood and urine samples, and the short time interval between diabetes diagnosis and completion of the questionnaire.
The observed association between how well the GPs knew patients and glycaemic level however could be explained by several models, and in the interpretation of the results, several limitations of this study should be taken into consideration.
Firstly, we used two simple and non-validated questions about the GP–patient relationship. Despite the simplicity, our question regarding the GPs' subjective evaluation of how well they knew their patients, may however be a reasonable indicator of the relational dimension of continuity of care. The knowledge accumulated in the GP–patient relationship incorporates complex knowledge used both consciously and subconsciously by the GP, leading Hjortdahl to conclude that ‘it is reasonable to accept the doctor's own evaluation as a valid indication of his or her general knowledge of each patient’.14 We had no information about the patient's perspective on the GP–patient relationship, however.
Secondly, it could be argued that our finding may simply mirror patients' help-seeking behaviour. Indeed, we do not have data on patients' actual visit patterns. A better knowledge of the patient may be expected in patients with whom the GP has frequent contact, e.g. patients with chronic diseases. Our observation of a higher frequency of cardiovascular disease and antihypertensive treatment among patients whom the GPs characterized as knowing well could indicate opportunistic case finding among these patients. The association between how well a GP knows a patient and hyperglycaemia, however, is not explained by including information on these possible confounders in multivariate models.
Patients whom the GPs characterized as not knowing well more frequently had glycaemic symptoms and had had these longer than patients known well or fairly well by their GP. Both the higher frequency and the longer duration would be expected to lead to more frequent GP contact, but could reflect the reverse, i.e. that this group of patients were less likely to seek help.
Patients' help-seeking behaviour may be influenced by social or psychological patient characteristics, and ‘not being known well by the GP’ could be a proxy measure for a subgroup of patients less prone to visit their GP, thus delaying the diagnostic process. Accordingly, we found a lower educational level among the subgroup of patients not known well by the GP. Indeed, both the availability and the use of care tend to vary inversely with social gradient in most health care systems.15 We did not, however, find differences between groups in other demographic variables typically associated with an adverse health profile, e.g. age, housing, (former) occupation, self-rated general health and physical activity. Furthermore, taking socio-demographic information into account in multivariate models only slightly weakened the association between how well the GP knew a patient and the hyperglycaemic level. Other patient- and/or doctor-related factors not measured in our study however could also have an effect on the patients' propensity to seek care when needed and to establish a personal GP–patient relationship, e.g. the physical distance between patient and GP and patients' attitudes about when a GP is contacted.
It should be noted that we did not find a trend in glycaemic level with increasing knowledge of patients, but rather that a subgroup of patients whom the GPs did not know well differed from other patients, a finding also reported from a similar Norwegian study setting.14 Altogether, we cannot fully rule out that GPs classifying a group of patients as not being known well is not simply a proxy for social or psychological patient characteristics not measured in detail in our study.
Including information of GP identification in the multivariate models had no effect on the results, thus ruling out that how well the GPs knew their patients was a proxy for type of doctor.
Finally, it is possible that the observed association does indeed reflect a relatively early diagnosis of patients whom the GP knew very well or fairly well. If compared with the average progression of hyperglycaemia among newly diagnosed diabetic patients from the UK Prospective Diabetes Study,13 the observed difference of 1 mmol/l between groups corresponds to a time period of 
4 years. Even though persuasive, this interpretation is somewhat hypothetical, as it is based on the assumption that all patients with T2DM have the same rate of metabolic deterioration. An early diagnosis of type 2 diabetes would be expected to be associated with other favourable risk factor levels than the glycaemic level, especially risk factors and complications not perceptible to the patients, e.g. dyslipidaemia, hypertension and renal involvement. We did not, however, find any differences between groups with regard to renal involvement and blood pressure, although the latter finding may be explained by the differences in the use of antihypertensive medication. Specifically, we have no explanation for our finding of elevated triglyceride levels among patients whom the GPs knew well, other than the possibility of a type one error.
The literature on this research area includes inconsistent observations. In data from NHANES (n = 18 162) Koopman et al.16 found a lower fraction of undiagnosed diabetes in individuals having a usual health care provider as compared with individuals not having such. O'Connor et al.17 compared diabetic patients, 84% with type 1 DM, who had a regular health care provider to patients who denied having such, and found a lower fraction of patients with high levels of glycosylated haemoglobin (HbA1c > 10%) in the first group, partly mediated by more clinic visits. Also among 256 patients with T2DM, who had an established relationship with a community health centre, Parchman et al.18 found glycaemic deterioration to be inversely associated with the intensity of continuity of care. In contrast to these established associations, Gill et al.19 found no association between provider continuity and completion of diabetes monitoring tests, and Hänninen et al.20 even found less satisfactory glucose control in individuals with good continuity of care. Overall, the studies in this research area are very difficult to compare due to great differences in health care systems, study populations and methodology used.
In conclusion, our data show that the glycaemic level among patients whom the GP characterizes as knowing well or fairly well is relatively low compared with patients whom the GP characterizes as not knowing well, independently of confounding variables. Consequently, GPs should be especially aware of undiagnosed T2DM among patients whom they do not know well. We suggest that the observed association reflects a late diagnosis in these patients, even though it is not possible to draw clear-cut conclusions on the basis of this cross sectional study nor can the possible role of residual confounding be ruled out. Prospective studies may help to unravel the possible connection between continuity of care and quality of care in family practice.
|
Declaration |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Funding: The Danish Medical Research Council; The Danish Research Foundation for General Practice, The Health Insurance Foundation, The Danish Ministry of Health; Novo Nordisk Farmaka Denmark Ltd; The Pharmacy Foundation; The Foundation for General Practice in Copenhagen, Frederiksberg, Tårnby og Dragør; Doctor Sofus Carl Emil Friis and his Wife Olga Doris Friis' Trust; The Danish Medical Association Research Fund; The Velux Foundation; The Rockwool Foundation; Novo Nordisk Ltd; The Danish Diabetes Association; The Oda og Hans Svenningsens Foundation; The A P Møller Foundation for the Advancement of Medical Science; The Novo Nordisk Foundation; Captain Axel Viggo Mørch and his Wife's Trust; The Danish Eye Health Society; Mogens and Jenny Vissing's Trust; and Bernhard and Marie Kleins Trust.
Ethical approval: informed consent was obtained from all study participants. The protocol was in accordance with the Helsinki declaration and was approved by the Ethical Committee of Copenhagen and Frederiksberg.
Conflicts of interest: none.
|
Acknowledgments |
|---|
We thank the patients and GPs who took part in the study, Lise Bergsoe for secretarial assistance, and Maeve Drewsen for a linguistic revision of the paper.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
1 McWhinney I. A Textbook of Family Medicine. Oxford University Press New York; 1989.
2 Leeuwenhorst Working Party (1977), The work of the general practitioner. Report of a statement from the second European Conference 1974. J R Coll Gen Pract 1977; 27: 117.
3 Gray DP, Evans P, Sweeney K et al. Towards a theory of continuity of care. J R Soc Med 2003; 96: 160–166.
4 Haggerty JL, Reid RJ, Freeman GK, Starfield BH, Adair CE, McKendry R. Continuity of care: a multidisciplinary review. BMJ 2003; 327: 1219–1221.
5 Saultz JW. Defining and measuring interpersonal continuity of care. Ann fam med 2003; 1: 134–143.
6 Freeman G, Hjortdahl P. What future for continuity of care in general practice? BMJ 1997; 314: 1870–1873.
7 Olivarius NF, Beck-Nielsen H, Andreasen AH, Horder M, Pedersen PA. Randomised controlled trial of structured personal care of type 2 diabetes mellitus. BMJ 2001; 323: 970–975.
8 Krasnik A, Groenewegen PP, Pedersen PA et al. Changing remuneration systems: effects on activity in general practice. BMJ 1990; 300: 1698–1701.
9 Olivarius NF, Jensen FI, Gannik D, Pedersen PA. Self-referral and self-payment in Danish primary care. Health Policy 1994; 28: 15–22.[Medline]
10 Drivsholm T, Olivarius NF. Routine diagnosis of Type 2 diabetes mellitus in general practice and hospitals: how do patients differ? Diabet Med 2005; 22: 336–339.[Medline]
11 Drivsholm T, Olivarius NF, Nielsen AB, Siersma V. Symptoms, signs and complications in newly diagnosed type 2 diabetic patients, and their relationship to glycaemia, blood pressure and weight. Diabetologia 2005; 48: 210–214.[CrossRef][Web of Science][Medline]
12 Beck-Nielsen H, Groop LC. Metabolic and genetic characterization of prediabetic states. Sequence of events leading to non-insulin-dependent diabetes mellitus. J Clin Invest 1994; 94: 1714–1721.[Web of Science][Medline]
13 Colagiuri S, Cull CA, Holman RR. Are lower fasting plasma glucose levels at diagnosis of type 2 diabetes associated with improved outcomes?: U.K. prospective diabetes study 61. Diabetes Care 2002; 25: 1410–1417.
14 Hjortdahl P, Borchgrevink CF. Continuity of care: influence of general practitioners' knowledge about their patients on use of resources in consultations. BMJ 1991; 303: 1181–1184.
15 Hart JT. The inverse care law. Lancet 1971; 1: 405–412.[CrossRef][Web of Science][Medline]
16 Koopman RJ, Mainous AG, III, Baker R, Gill JM, Gilbert GE. Continuity of care and recognition of diabetes, hypertension, and hypercholesterolemia. Arch Intern Med 2003; 163: 1357–1361.
17 O'Connor PJ, Desai J, Rush WA, Cherney LM, Solberg LI, Bishop DB. Is having a regular provider of diabetes care related to intensity of care and glycemic control? J Fam Pract 1998; 47: 290–297.[Web of Science][Medline]
18 Parchman ML, Pugh JA, Noel PH, Larme AC. Continuity of care, self-management behaviors, and glucose control in patients with type 2 diabetes. Med Care 2002; 40: 137–144.[CrossRef][Web of Science][Medline]
19 Gill JM, Mainous AG, III, Diamond JJ, Lenhard MJ. Impact of provider continuity on quality of care for persons with diabetes mellitus. Ann Fam Med 2003; 1: 162–170.
20 Hanninen J, Takala J, Keinanen-Kiukaanniemi S. Good continuity of care may improve quality of life in type 2 diabetes. Diabetes Res Clin Pract 2001; 51: 21–27.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||