Treatment satisfaction of diabetic patients: what are the contributing factors?

Transcription

1 Ó The Author Published by Oxford University Press. All rights reserved. For permissions, please doi: /fampra/cmp007 Family Practice Advance Access published on 2 March 2009 Treatment satisfaction of diabetic patients: what are the contributing factors? Aya Biderman a, Enav Noff b, Stewart B Harris c, Nurit Friedman d and Amalia Levy e Biderman A, Noff E, Harris SB, Friedman N and Levy A. Treatment satisfaction of diabetic patients: what are the contributing factors? Family Practice 2009; 26: Background. Treatment satisfaction is an important factor of quality of care, especially in treating chronic diseases such as diabetes mellitus. Identifying factors that independently influence treatment satisfaction may help in improving clinical outcomes. Objective. To find the relationship between treatment satisfaction of diabetic patients and socio-demographic, clinical, adherence, treatment and health perception factors. Methods. Patients were interviewed by telephone about their socio-demographic parameters, health status, clinical data and treatment factors. The Diabetes Treatment Satisfaction Questionnaire (DTSQ) was used to measure satisfaction and adherence. This is a cross-sectional study, as part of a larger study of chronic patients in Israel. Subjects were randomly selected diabetes patients. The main outcome measures were DTSQ levels. A multivariate linear regression model was constructed to identify factors independently associated with patients satisfaction. Results. In all, 630 patients were included in the study. Multivariate analysis indicated that demographic parameters (e.g. female gender, P = 0.036), treatment factors (e.g. type of medication, P < 0.001), adherence factors (e.g. difficulty attending follow-up or taking medications, P < 0.001) and clinical factors (e.g. diabetes complications, P < 0.01) were independently associated with lower treatment satisfaction. Conclusions. Treatment satisfaction is lower among diabetic patients who have a lower educational level, who are insulin treated or have a diabetic complication and is related to difficulties in taking medications and coming to follow-up visits. Addressing the specific needs of these patients might be effective in improving their satisfaction, thus having a positive influence on other clinical outcomes. Keywords. Diabetes mellitus, health status, patient satisfaction, primary health care, quality of health care. Introduction Studies that examined the clinical efficacy of different types of diabetes treatment models have addressed issues such as quality of life among patients, 1 quality of treatment, 2 clinical outcomes in patients followed by different groups of physicians 3 and satisfaction with treatment. 3,4 The importance of studying treatment satisfaction is well documented: numerous health organizations have implemented a measurement of patient satisfaction in projects designed to improve quality of care. 5,6 Quality of care is evaluated by three equally important measures: structure, process and outcomes. Treatment satisfaction is included in the process component and is used as an important indicator of quality of care. 6 Moreover, greater satisfaction has been found to be correlated to higher rates of compliance, 6 lower haemoglobin A1c (HbA1c) values and lower body weight, 4 suggesting that a higher satisfaction is related to better clinical outcomes. Received 31 March 2008; Revised 18 December 2008; Accepted 1 February a Department of Family Medicine, Siaal Research Center for Family Medicine and Primary Care, Faculty of Health Sciences, b Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer-Sheva, Israel, c Center for Studies in Family Medicine, University of Western Ontario, London, Ontario, Canada, d Maccabi Healthcare Services, Tel Aviv, and e Department of Epidemiology and Health Services Evaluation, Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer-Sheva, Israel. Correspondence to Aya Biderman, Department of Family Medicine, Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer- Sheva, Israel; sbider@netvision.net.il. 102

2 Satisfaction of diabetic patients 103 Factors that influence treatment satisfaction are difficult to assess because of the inter relations among them. Former studies have measured only one or a few parameters. 1,4 The objective of this study was to determine various parameters that are independently related to treatment satisfaction, in a community-based sample of diabetic patients. Identification of those parameters may help the health care system and the primary care physician to focus and address them directly, in order to improve patient satisfaction and possibly influence clinical outcomes. Methods Study population Universal health care in Israel is provided through one of four health care organizations. All patient information is centrally computerized. For this study, patients were randomly chosen from the computer database of the two largest health care organizations in Israel, providing health care to approximately 70% of the population of Israel. Confirmation of a diabetes diagnosis was received by telephone interviews (by EN). This study was approved by the local institutional ethics committee. Study design In this cross-sectional study, patients with one of three chronic conditions (hypertension, diabetes mellitus or ischaemic heart disease) were interviewed by telephone during as part of a large, prospective study titled Cost-Sharing, Health Services Utilization and Patients Health Status. Patients with diabetes mellitus were interviewed again from September 2003 to April 2004 as part of this study, in order to complete the data that were not collected during the first interview. Socio-demographic and general health status data were gathered during the first interview. Data regarding treatment satisfaction, adherence, treatment regimen and clinical information were gathered during the second interview. Study tools The Diabetes Treatment Satisfaction Questionnaire (DTSQ) 7 was used to evaluate patient satisfaction with treatment. This tool has been identified by the World Health Organization and the International Diabetes Foundation as useful in assessing outcomes of diabetes care. 8 The questionnaire, as was used in other studies, 1 consists of six items assessing treatment satisfaction, each is scored on a scale of 0 6, with six representing the greatest satisfaction. Seven questions related to adherence included difficulty in changing dietary habits, in changing physical habits, in taking medication, in attending follow-up visits, owning a blood glucose meter, using the meter and frequency of self-monitoring of blood glucose. SF-12 questionnaire, which was validated to Hebrew, was used to evaluate patients self-rated health status. Treatment factors and clinical data Data regarding the treatment factors included: type of medical care provided (family physician only or family physician plus diabetes clinic); proximity of clinic to patient s home; whether patient was under routine surveillance of a diabetes nurse and, if so, where; date of last fasting venous glucose blood test and HbA1C performed and who ordered this test; type of treatment [diet, oral hypoglycaemic agent (OHA), insulin or both] and whom the patient noted as the one who helped them most in managing their diabetes (family physician, diabetes specialist or nurse). Self-reported clinical data collected included: presence of co-morbidities and diabetes-related complications, remembering the most recent HbA1c test, and, if so, the HbA1c value. Statistical analyses Statistical analysis was performed using SPSS package (SPSS 12.0, Chicago, IL). Relationship between treatment satisfaction (mean scores of DTSQ) and sociodemographic parameters, diabetes complications and difficulties in adherence was examined. Statistical significance was ascertained using the one-way analysis of variance (ANOVA). The correlations between continuous variables and treatment satisfaction as well as mental and physical status (SF-12 questionnaire) were tested by Pearson analysis. A multivariate linear regression model with stepwise elimination was constructed in order to identify independent variables associated with treatment satisfaction. All statistically significant (P < 0.05) variables in the univariate analyses were included in the initial regression model. Results Patients Of the 1223 diabetic patients from the Cost-Sharing, Health Services Utilization and Patients Health Status study, a total of 701 patients completed the second telephone interview. Of the 701 study participants, 43% (300) were from one of the health care organization (Group A) and 57% (401) from the other (Group B). Of the 522 patients who did not complete the second interview, 100 indicated that they did not have diabetes, 71 refused to participate, 32 had died and 319 could not be reached by telephone. There were no statistically significant differences in baseline socio-demographic characteristics (gender, age, marital status, country of birth, education, income and occupation) between the 522 nonparticipants and the 701 participants. Of these 701

3 104 Family Practice an international journal diabetic patients, 71 were treated with insulin only and could not be accurately diagnosed as type 2 patients; therefore, these patients were not included in the analysis. The socio-demographic and clinical characteristics of the study participants are shown in Table 1. The mean age was ± 9.29 years, 53.6% were female, 75% were living alone, 22% had more than 12 years of education, 50% had below average income and 25% were employed. Table 1 also shows the clinical data and treatment factors of the study sample. Approximately 75% were treated with OHAs, 63% had hypertension and more than 54% had diabetes-related complications, with neuropathy being the most frequent (33%). Only 20% remembered their last HbA1c value, and the average value among those patients who remembered was 8.29% ± 2.17%. About 50% of those who remembered their last HbA1c had values higher than the guideline-recommended target of 7.0%. 9 Treatment satisfaction The mean score of the DTSQ was 4.54 ± 1.12 out of 6. Analysis of the mean DTSQ score to sociodemographic parameters, diabetes complications and difficulties in adherence are described below. Socio-demographic parameters and satisfaction The univariate analysis revealed that female patients had lower satisfaction with treatment than male patients (P < 0.001). We found a strong association between greater treatment satisfaction and higher socio-economic status (Table 2). Patients who had an income below average (P = 0.001) and had up to 6 years of education (P < 0.001) were less satisfied with treatment. No correlation was found between age and treatment satisfaction (Pearson correlation (r = 0.06, P = 0.12). Clinical parameters and satisfaction There was no statistically significant association between presence of co-morbidities and treatment satisfaction, although mean DTSQ scores of patients with more than 2 co-morbidities were lower than patients with up to 1 co-morbidity (4.59 ± 1.1, 4.44 ± 1.17, respectively, P = 0.096). Presence of even one diabetes-related complication (regardless of the type) was significantly related with lower treatment satisfaction scores (mean = 4.79 ± 1.06 without complications, mean = 4.38 ± 1.10 with complications, P < 0.001). Table 3 shows that having any diabetes-related complications was associated with lower satisfaction with treatment. Patients with foot ulcers were the least satisfied with treatment (mean DTSQ = 3.93 ± 1.15). When the number of complications per patient were combined (from 0 to 3) and the scores of each group (e.g. groups of patients with 0, 1, 2 or 3 complications) TABLE 1 Socio-demographic, clinical characteristics and treatment factors of patients (N = 630) Variable N % Age (mean ± SD) ± 9.3 Gender Male Female Family status Alone Married Country of birth Israel Other Education (years) Income Below average Average Above average Work status Working Not working Co-morbidity Hypertension Heart diseases Other chronic disease Diabetes complications Total Retinopathy Nephropathy Neuropathy Foot ulcer Treatment Diet OHA Insulin and OHA HbA1c Remember value Average value (SD) 8.29 (2.17) HbA1c < Data are expressed as numbers and percentages or means ± standard deviation (SD). were compared with ANOVA analysis, we found decreasing treatment satisfaction with increasing number of complications (mean = 4.79 ± 1.06 without complications descending to mean of 3.87 ± 1.12 with three complications, P < 0.001). We found no relation between treatment satisfaction and remembering the last HbA1c or the HbA1c value. DTSQ scores were higher in patients with HbA1c values lower than 7.0%, compared to those with HbA1c values above 7.0%, but the difference was not statistically significant (mean = 4.62 ± 1.01, mean = 4.32 ± 1.23, respectively, P = 0.14). Adherence and satisfaction Of the 630 patients, 51.9% reported difficulties in changing their dietary habits, 50.3% in changing their physical activities, 22.7% in attending follow-up visits

4 Satisfaction of diabetic patients 105 TABLE 2 Relationship between treatment satisfaction and sociodemographic parameters Relationship between treatment satisfaction and diabetes treatment modalities, complications and difficulties in adherence TABLE 3 Socio-demographic parameters Mean DTSQ a score ± SD P-value DTSQ a P-value score mean (SD) Gender Male 4.73 ± 1.03 <0.001 Female 4.38 ± 1.17 Family status Alone 4.61 ± 1.06 <0.001 Married 4.32 ± 1.26 Country of birth Israel 4.44 ± Other 4.63 ± 1.07 Education Up to 6 years 4.25 ± years 4.65 ± 1.08 >12 years 4.58 ± 1.20 Income Below average 4.44 ± Average 4.65 ± 1.07 Above average 4.67 ± 1.08 Work status Working 4.64 ± Not working 4.5 ± 1.14 a DTSQ (scale ranging from 0 = lowest to 6 = highest). and 9.8% in taking medications. The strongest association was found between difficulties in changing dietary habits and physical activities, with an overlap of 60% between the two parameters (P < 0.001). Patients with any one of the difficulties in adherence scored significantly lower in the DTSQ than patients with no difficulties (Table 3). A total of 445 patients (70.6%) owned a blood glucose meter and 92% reported using it. We found association neither between treatment satisfaction and owning a glucose meter, using it, nor between different frequencies of reported self-monitoring of blood glucose. Treatment factors and satisfaction Patients who were treated with diet alone were the most satisfied with treatment (mean = 5.27 ± 0.87). Patients taking OHAs were more satisfied than those taking OHA and insulin (mean = 4.56 ± 1.09, mean = 3.77 ± 1.13, respectively, P < 0.001). Of all patients, 425 (67.5%) were treated merely in their primary care clinic, 204 (32.4%) were also treated in a diabetes clinic, 470 patients (74.6%) were treated in clinics located near their home and 151 (24%) were under routine surveillance of a diabetes nurse (66% of them within their primary care clinic). Of all 630 study participants, 463 patients (73.5%) were assigned to perform blood tests, the majority (74.8%) by their primary care clinic. Of all these treatment parameters, only two were found significantly related to treatment satisfaction: patients treated only in their primary care clinic were more satisfied than Treatment modality Diet 5.27 ± 0.87 <0.001 OHA 4.56 ± 1.09 OHA + insulin 3.77 ± 1.13 Easy accessibility to clinic Yes 4.67 ± 1.04 <0.001 No 4.15 ± 1.27 Diabetes complications Retinopathy Yes 4.29 ± No 4.61 ± 1.12 Nephropathy Yes 4.16 ± No 4.59 ± 1.09 Neuropathy Yes 4.26 ± 1.16 <0.001 No 4.66 ± 1.09 Foot ulcer Yes 3.93 ± 1.15 <0.001 No 4.60 ± 1.11 Difficulties in adherence Difficulty in changing dietary habits Yes 4.46 ± No 4.63 ± 1.10 Difficulty in changing physical activities Yes 4.44 ± No 4.64 ± 1.05 Difficulty in taking medications Yes 3.35 ± 1.44 <0.001 No 4.64 ± 1.01 Not taking medication 5.17 ± 0.81 Difficulty in arriving to follow-up Yes 3.97 ± 1.31 <0.001 No 4.71 ± 1.00 a DTSQ (scale ranging from 0 = lowest to 6 = highest). those treated in a diabetes clinic as well (mean = 4.63 ± 1.08, mean = 4.39 ± 1.15, respectively, P = 0.012), and patients who were treated near their home were more satisfied than those who were not (mean = 4.67 ± 1.04, mean = 4.15 ± 1.27, respectively, P < 0.001). Regarding the question who helps you most with the treatment of your diabetes, 88% of the patients who were treated only in their primary care clinic chose family physician. Among patients who were treated in a diabetic clinic as well, 24% chose family physician and 54% chose diabetes specialist (P < 0.001). Self-rated health status and satisfaction General health perceptions questionnaire (SF-12) measures patients self-rated view about their mental and physical health status. Using Pearson analysis, we found a positive correlation between higher mental status, higher physical status and higher treatment satisfaction (r =+0.23,P <0.001andr = +0.24, P <0.001).

5 106 Family Practice an international journal Multivariate analysis and satisfaction Table 4 summarizes the results of the multiple linear regression analysis examining the relation between all statistically significant parameters discussed above and DTSQ scores. This provides a model of the parameters that are independently associated with treatment satisfaction. For example, according to the model, OHA therapy was associated with lower treatment satisfaction than diet treatment only. Adherence difficulties, having a diabetes complication and up to 6 years of education, were associated with lower treatment satisfaction. Discussion The main findings of this study, according to the multivariate analysis, were that lower treatment satisfaction is related to difficulties in adherence to taking medications and attending follow-up clinic visits, treatment with insulin or oral medications, any diabetes complication, being female and less than 6 years of education. The relation between better adherence and higher treatment satisfaction is well documented. 2,6,10 Adherence to taking medications as prescribed is associated with higher treatment satisfaction. 10 Our findings, that difficulty in taking medications is related to lower satisfaction, support this result. We found, as others, 3,11 that insulin-treated patients are least satisfied with treatment, followed by patients treated with oral agents. Aside from the obvious fact that injecting insulin is less comfortable than taking a pill, this outcome may also reflect patients perceptions that insulin treatment means that their health status has deteriorated. Another possible explanation is that type 2 diabetic patients, who need insulin, have longer disease duration, with more complications. We suggest that by addressing this misperception, health care providers may be able to increase treatment satisfaction among insulin-treated patients. TABLE 4 Multiple linear regression for the association between treatment satisfaction and clinical parameters, treatment factors, difficulties in adherence and socio-demographic parameters Parameter b P-value Difficulty in taking medications 0.24 <0.001 (yes versus no) Insulin + OHA treatment (versus diet) Difficulty in arriving to follow-up 0.18 <0.001 (yes versus no) OHA treatment (versus diet) 0.25 <0.001 Diabetes complication (any versus none) Education (more than 6 years versus up to 6 years) Gender (female versus male) Similar to other studies, 1,3,11 we found lower treatment satisfaction in the presence of diabetes-related complications. Furthermore, we found less satisfaction with having any complication at all and a constant decline in treatment satisfaction with increased number of complications. Of all the complications, foot ulcers were associated with the lowest treatment satisfaction. We assume that fear of amputation, mobility difficulties and the long duration of treatment needed for foot ulcers are the main causes for this finding. Our finding that people with lower education levels are less satisfied with treatment was also found by Nicolucci et al. 11 This might reflect a general lower satisfaction with life of people with lower educational levels. Women were less satisfied than men, consistent with findings in other countries, such as in Sweden 12 and Italy. 11 In the univariable analysis, some other parameters were found to be related to lower treatment satisfaction. Narayan et al. 13 found that satisfaction was positively associated with higher income and employment. Nicolucci et al. 11 also found that satisfaction was lower among unemployed diabetic patients. Our study shows the same trend regarding income, but no association between satisfaction and employment, which may be explained by the high number of unemployed patients in our study, and their older age. Trief et al. 14 found that the better the quality of the marital relationship, the higher the treatment satisfaction. We did not address the quality of marital relationship, but we did find that married patients were less satisfied with treatment than non-married patients. However, we note that this finding was based on a sample of only 161 married patients out of the 630 patients in our sample. Some authors found that older patients are more satisfied with treatment than younger patients; 1 however, in our study, as well as others, 11,13 we found no association between treatment satisfaction and age. No association between co-morbidities and satisfaction was found, despite the high prevalence of co-morbidities. A possible explanation is that the most common comorbidity was hypertension a silent disease. Treatment satisfaction was found to be associated with lower HbA1c values. 1,4 No such association was found in this study, although lower DTSQ scores were found in patients with HbA1c values greater than 7.0%. This might be related to the small number of patients who remembered their HbA1c value. Only 20% of the patients remembered their last HbA1c. Beckles et al. 15 found, in their population-based study of 2118 diabetic patients from 22 American states, that only 25% were aware of the term HbA1c. Our results support this finding and perhaps point to a wider problem of awareness of the importance of this measurement. Narayan et al. 13 found a positive relationship between treatment satisfaction and ease of accessing care, as was found in this study.

6 Satisfaction of diabetic patients 107 Some studies have demonstrated improved clinical outcomes in patients treated in diabetic clinics. 3,11,16 This could be related to the patient s subjective perception of feeling more secure, when treated by a competent specialist, in the hospital setting, or to an objective reality of a multi-professional team, providing more time and extra attention to the patient. In our study, we found lower satisfaction among patients treated in diabetic clinics. This could reflect the Israeli health care system of treating diabetic patients, which takes place in the primary care setting, with referral of complex patients to diabetes clinics for consultation. As discussed above, we found a strong association in the multivariate analysis between severity of the disease (i.e. those patients with complications or on insulin) and low satisfaction. Therefore, it seems that the effect of the severity of the disease on satisfaction is stronger than the type of clinic and being treated by a diabetes specialist. As in other studies, 17 we found that lower self-rated mental and physical health status on the SF-12 scores were correlated with lower treatment satisfaction, although they were not found significant in the multiple regression model. Strengths and limitations of the study Although this study has advantages over former studies regarding satisfaction, it does have some limitations. One hundred interviewees indicated that they did not have diabetes mellitus and were excluded from the study. Brown et al. 18 noted that lack of acceptance of diabetes as a chronic illness was a major barrier in patients adherence. The exclusion of these patients may have caused a selection bias, as more adherent patients may have been over-represented in this study. In addition, since the study was conducted in a crosssectional design, we cannot conclude whether treatment satisfaction is influenced by the different independent variables or vice versa. Other parameters such as depression and anxiety, which might effect patient satisfaction, were not addressed in this study, although the SF-12 was used to assess the patients mental status. Our results are also limited by relying upon the participants self-reports and not upon objective clinical data, with the possibility of recall bias. A recent study of a large sample of diabetic patients from Italy 11 was based on data collected from patients questionnaires, from their medical records and their physicians. This study found quite similar results regarding treatment satisfaction: Insulin treatment and the presence of diabetic complications were the major factors associated with low DTSQ scores. Thus, we assume that this limitation is marginal. Despite the large patient sample from all parts of Israel, our findings may not be generalizable to other countries where diabetic patients treatment is not as community based as in Israel. In conclusion, of all the many factors we examined, only a handful of parameters were found to be statistically associated with treatment satisfaction. Therefore, patients who have difficulties in adherence with medications and follow-up visits, as well as patients who are treated with insulin and have diabetes complications, should be targeted as a unique group among diabetic patients. Those patients require more careful attention in their medical treatment in order to improve their treatment satisfaction and, hopefully, achieve better clinical outcomes. Future research to confirm our results should use a larger sample, a broader age range and a prospective methodology. Declaration Funding: The Israel National Institute for Health Policy and Health Services Research. Ethical approval: This study was approved by the local institutional ethics committee. Conflicts of interest: None. References 1 Ken W, Koopmanshap MA, Stolk RP, Rutten GEM, Wolffenbuttel BHR, Niessen LW. Health-related quality of life and treatment satisfaction in Dutch patients with type 2 diabetes. Diabetes Care 2002; 25: Blone L, Dey J, Tesa MA, Guthrie RD. Defining and measuring quality of diabetes care. Prim Care 1999; 26: Greenfield S, Kaplan SH, Kahn R, Ninomiya J, Griffith JL. Profiling care provided by different groups of physicians: effects of patient case-mix (bias) and physician-level clustering on quality assessment results. Ann Intern Med 2002; 136: Bradley C, Lewis KS. Measures of psychological well-being and treatment satisfaction developed from the responses of people with tablet-treated diabetes. Diabet Med 1990; 7: McCormick D, Himmelstein DU, Woolhandler S, Wolf SM, Bor DH. Relationship between low quality-of-care scores and HMO s subsequent public disclosure of quality-of-care scores. J Am Med Assoc 2002; 288: Finkel MI. The importance of measuring patient satisfaction. Empl Benefits J 1997; 22: Bradley C. Diabetes treatment satisfaction questionnaire. In Bradly C (ed.). Handbook of Psychology and Diabetes: A Guide to Psychological Measurement in Diabetes Research and Practice, Chur, Switzerland: Harwood Academic Publisher, 1994: Bradly C, Gamsu DS. Guidelines for encouraging psychological well-being: report of a working group of the World Health Organization Regional Office for Europe and International Diabetes Federation Europe Region St Vincent Declaration Action Programme for Diabetes. Diabet Med 1994; 11: American Diabetes Association. Standards of medical care for patients with diabetes mellitus. Diabetes Care 2002; 25 (suppl 1): S33 S The QuED Study Group-Quality of Care and Outcomes in Type 2 Diabetes. Correlates of satisfaction for the relationship with their physician in type 2 diabetic patients. Diabetes Res Clin Pract 2004; 66: Nicolucci A, Cucinotta F, Squatrito S et al. for the QuoLITy Study Group. Clinical and socio-economic correlates of quality of life

7 108 Family Practice an international journal and treatment satisfaction in patients with type 2 diabetes. Nutr Metab Cardiovasc Dis 2009; 19: Jonsson PM, Sterky G, Gafvals C, Ostman J. Gender equity in health care: the case of Swedish diabetes care. Health Care Women Int 2000; 21: Narayan KM, Gregg EW, Fagot-Campagna A et al. Relationship between quality of diabetes care and patient satisfaction. J Natl Med Assoc 2003; 95: Trief PM, Wade MJ, Britton KD, Weinstock RS. A prospective analysis of marital relationship factors and quality of life in diabetes. Diabetes Care 2002; 25: Beckles GL, Engelgau MM, Narayan KM, Herman WH, Aubert RE, Williamson DF. Population-based assessment of the level of care among adults with diabetes in the U.S. Diabetes Care 1998; 21: Varroud-Vial M, Mechaly P, Joannidis S et al. Cooperation between general practitioners and diabetologists and clinical audit improve the management of type 2 diabetic patients. Diabetes Metab 1999; 25: Westaway MS, Rheeder P, Van Zyl DG, Seager JR. Interpersonal and organizational dimensions of patient satisfaction: the moderating effects of healthstatus. Int J QualHealth Care 2003; 15: Brown JB, Harris SB, Webster-Bogaert S, Wetmore S, Faulds C, Stewart M. The role of patient, physician and systemic factors in the management of type 2 diabetes mellitus. Fam Pract 2002; 19: