Audit of a shared-care program for persons with diabetes: baseline and 3 annual follow-ups

Transcription

1 Acta Diabetol (2004) 41:9 13 DOI /s z Springer-Verlag 2004 ORIGINAL A.V. Ciardullo M.M. Daghio M. Brunetti M. Bevini G. Daya G. Feltri D. Novi C.A. Goldoni A. Guerzoni A. Messori N. Magrini Audit of a shared-care program for persons with diabetes: baseline and 3 annual follow-ups Received: 10 October 2002 / Accepted in revised form: 3 November 2003 Abstract In Italy, data on shared-care programs for diabetes are lacking. We described the characteristics of type 2 diabetic population assisted in general practice and evaluated 3 years of follow-up outcomes and performance indicators in a shared-care program in Modena, Italy ( ); only well-controlled diabetic patients were considered. Forty-nine percent of territorial GPs adhered to the project (257 out of 521) and 77% of them sent 6409 paired baseline and follow-up datasheets. Altogether, 97.8% patients had type 2 diabetes, mean age 68.6±11.7 years, disease duration 9.6±7.5 years, BMI 28.6±4.8 kg/m 2, HbA 1c 7.6%±1.6%, 16.1% of them were disabled. Among the non-disabled patients, 23.6% had optimal glycemic control (HbA 1c 6.5%); at baseline the prevalence of micro- and macrovascular diabetic complications was: 8.2% microalbuminuria and 2.4% macroalbuminuria plus nephropathy, 11.0% nonproliferative and 3.0% preproliferative retinopathy, 7.0% neuropathy, 1.8% diabetic foot; 8.5% angina, 6.9% TIA or stroke, 6.3% infarction, 5.2% intermittent claudication, 4.1% heart failure. Among the disabled patients 27.9% had optimal glycemic control, but they had more diabetic complications. The performance indicators significantly improved over the 3-year study period: glycemic control indicators increased from 66% 75% to 83% 90% and micro- and macrovascular indicators from 59% 65% to 75% 81%. The outcome indicators also improved: mean HbA 1c value changed from 7.6%±1.6% to 7.3%±1.3% and the percentage of people with HbA 1c 6.5% significantly improved over time. Similar trends were observed in both disabled and non-disabled diabetic patients. Key words Diabetes mellitus Medical audit Family practice Health services research Introduction A.V. Ciardullo ( ) M.M. Daghio M. Brunetti N. Magrini Center for the Evaluation of the Efficacy of Healthcare (CeVEAS) AUSL Modena, Local Health Unit Viale Muratori 201, I Modena, Italy a.ciardullo@ausl.mo.it A.V. Ciardullo M.M. Daghio M. Bevini G. Feltri D. Novi C.A. Goldoni A. Guerzoni A. Messori Progetto Diabete Modena, Italy G. Daya Italian GPs Trade Union (FIMMG) Modena, Italy Healthcare systems worldwide are facing a shift in the care of people with diabetes from hospital to primary care [1 3], and systems for regular review of patients are crucial in delivering a good standard of diabetes care [1, 4, 5]. In Italy, observational data from general practitioners (GPs) involved in the integrated diabetes care are lacking. In 1998, the healthcare authority in Modena designed a shared-care program for people with well-controlled type 2 diabetes mellitus; the program is called Progetto Diabete. In this paper, we describe the characteristics of the diabetic population assisted by GPs in Progetto Diabete and report patient outcome and GP adherence over 3-year period.

2 10 A.V. Ciardullo et al.: Shared-care program for diabetes Patients and methods Modena Province comprises inhabitants, of whom about thousand have diabetes. The shared-care program was designed only for patients with well-controlled type 2 diabetes. A well-controlled patient was defined as one having a stable form of disease and a glycohemoglobin (HbA 1c) level lower than 8.0%. A diabetic patient was an individual with a fasting blood glucose greater than 126 mg/dl on two occasions or a patient already treated for diabetes. Of the 521 GPs in Modena Province, 257 (49%) adhered to the shared-care program; of these, 198 (77%) sent baseline and follow-up datasheets for 6409 patients. We analyzed 6267 paired patient datasheets (97.8%) because 142 of them (2.2%) referred to type 1 diabetic patients. The Progetto Diabete committee adopted two sheets for data collection, i.e. baseline and yearly follow-up datasheets recording individual patient data on age, gender, type of diabetes, disease duration, weight, height, HbA 1c, diabetes therapy, presence of microvascular diabetic complications (4 stages of nephropathy, 4 stages of retinopathy, peripheral neuropathy, and diabetic foot), presence of macrovascular diabetic complications (angina pectoris, myocardial infarction, transient ischemic attack or stroke, chronic heart failure, and intermittent claudication). No active HbA 1c determinations were performed, however HPLC is the method for HbA 1c measurement used by territorial labs. The patients gave written consent to be included in the shared-care program. Participating GPs sent baseline and annual follow-up datasheets to a diabetes registry located in the territorial public healthcare department. Descriptive statistics from diabetes registry of baseline and follow-up data were calculated according to disability or not. A disabled patient is one who is unable to attend to his own personal care and also unable to walk without assistance. Disabled diabetic patients have to be cared by GPs regardless of the cause of their disability. Beta coefficients adjusted for age, sex, body mass index (BMI) and HbA 1c for comparisons between disabled vs. non-disabled persons were computed by multivariate regression analysis. Time trends were analyzed by means of a nonparametric test for linear trends which is an extension of Wilcoxon s rank-sum test by Cuzick and Altman; the modified formula incorporates a correction for ties. The observation period lasted from 1998 to Levels of significance were expressed by means of 95% confidence intervals (CIs) or as p<0.05. STATA for Windows release 7.0 statistical software package was used. Results We observed baseline and 3 years of follow-up data for 6267 subjects with type 2 diabetes, mean age 68.6±11.7 years, disease duration 9.6±7.5 years, BMI 28.6±4.8 kg/m 2, HbA 1c 7.6%±1.6%, 16.1% of them were disabled. Baseline characteristics of disabled (1007 of 6267) and non-disabled (5260 of 6267) diabetic patients and the comparison between the two groups are shown in Table 1. The disabled diabetic patients were significantly older than the non-disabled ones, there were significantly more women among them and they had a lower BMI. The significant differential characteristics of the disease in disabled patients consisted of a longer disease duration and a poorer glycemic control with a higher mean HbA 1c value. The prevalences of major cardiovascular risk factors in the two groups are shown in Table 2, and the prevalences of micro- and macrovascular complications are shown in Table 3. Overall, 24.2% of diabetic patients had optimal glycemic control (HbA 1c 6.5%), while 34.7% had acceptable control (HbA 1c from 6.5% to 7.5%). As to the disabled patients, 27.9% had optimal glycemic control and 30.3% had acceptable control. As to the non-disabled patients, 23.6% had optimal glycemic control and 35.4% had acceptable control. They showed fairly low rates of diabetic complications both microvascular (8.2% microalbuminuria, 0.2% macroalbuminuria, 2.2% renal failure, 11.0% nonproliferative, 1.3% preproliferative, 1.7% proliferative retinopathy, 7.0% neuropathy, 1.8% diabetic foot), and macrovascular (8.5% angina, 6.9% TIA or stroke, 6.3% myocardial infarction, 5.2% intermittent claudication, 4.1% heart failure) (Table 3). Table 1 Baseline characteristics of 6267 subjects with type 2 diabetes Diabetic subjects Disabled vs. non-disabled, p value Total Disabled Non-disabled (n=6267) (n=1007) (n=5260) Age, years a 68.6 (11.7) 75.6 (11.6) 67.3 (11.2) (0.009 to 0.010) c <0.01 Women, n (%) (51.8) (62.5) (50.4) (0.067 to 0.097) c <0.01 BMI, kg/m 2a (4.8) (5.0) (4.8) ( to ) c <0.01 Diabetes duration, years a 9.6 1(7.5) (8.2) (7.4) 0.47 (0.42 to 0.51) c <0.01 Family history of diabetes, n (%) (51.9) (45.4) (51.4) 0.09 ( to ) c >0.05 HbA 1c, % a 7.6 1(1.6) 7.7 1(1.7) 7.6 1(1.6) (0.002 to 0.014) b <0.01 a Values are mean (SD); b values are adjusted for age, sex and HbA 1c; c values are adjusted for age, sex and BMI BMI, body mass index

3 A.V. Ciardullo et al.: Shared-care program for diabetes 11 Table 2 Cardiovascular risk factors in 6267 subjects with type 2 diabetes, by disability status. No difference between groups is significant Disabled (n=1007) Non-disabled (n=5260) Disabled vs. non-disabled, Hypertension, n (%) 676 (67.1) 3150 (59.9) ( to 0.032) High blood cholesterol, n (%) 281 (27.9) 1660 (31.6) ( to 0.010) High blood triglycerides, n (%) 240 (23.8) 1452 (27.6) ( to 0.027) Cigarette smoking, n (%) 104 (10.3) 1799 (15.2) ( to 0.045) Table 3 Prevalence of diabetic complications at baseline in 6267 subjects with type 2 diabetes, by disability status Disabled (n=1007) Non-disabled (n=5260) Disabled vs. non-disabled, p value Microvascular Microalbuminuria 178 (17.7) 431 1(8.2) (0.10 to 0.15) <0.01 Macroalbuminuria 8 1(0.8) 10 1(0.2) (0.09 to 0.15) <0.01 Renal failure 56 1(5.6) 116 1(2.2) (0.07 to 0.10) <0.01 Nonproliferative retinopathy 193 (19.2) 579 (11.0) (0.054 to 0.10) <0.01 Preproliferative retinopathy 39 1(3.9) 68 1(1.3) (0.06 to 0.11) <0.01 Proliferative retinopathy 59 1(5.9) 89 1(1.7) (0.01 to 0.08) <0.01 Diabetic neuropathy 192 (19.1) 368 1(7.0) (0.13 to 0.19) <0.01 Diabetic foot 84 1(8.4) 95 1(1.8) (0.23 to 0.33) <0.01 Macro-ascular Angina pectoris 157 (15.6) 447 1(8.5) (0.04 to 0.10) <0.01 Myocardial infarction 82 1(8.2) 331 1(6.3) ( to 0.041) >0.05 Heart failure 171 (17.0) 216 1(4.1) (0.18 to 0.25) <0.01 TIA or stroke 271 (26.9) 363 1(6.9) (0.22 to 0.28) <0.01 Intermittent claudication 136 (13.5) 273 1(5.2) (0.14 to 0.21) <0.01 TIA, transient ischemic stroke Table 4 GPs prescribing patterns at baseline for 6267 subjects with type 2 diabetes, by disability status Therapy for diabetes, n (%) Disabled (n=1007) Non-disabled (n=5260) Disabled vs. Non-disabled p value Diet alone 175 (17.4) 1562 (29.7) ( to ) <0.01 Sulphonylureas 351 (34.9) 1662 (31.6) (0.022 to 0.056) <0.01 Sulphonylureas plus biguanides 248 (24.6) 1183 (22.5) ( to 0.070) >0.05 Biguanides 34 1(3.4) 268 1(5.1) ( to 0.050) >0.05 Insulin 61 1(6.1) 16 1(0.3) (0.67 to 0.78) <0.01 Insulin plus oral hypoglycaemic drugs 17 1(1.7) 16 1(0.3) (0.46 to 0.72) <0.01 Glinides 34 1(3.4) 184 1(3.5) (-0.01 to 0.08) >0.05 Sulphonylureas plus glinides 4 (0.45) 20 (0.38) (-0.16 to 0.10) >0.05 Biguanides plus glinides 1 (0.09) 2 (0.04) (-0.20 to 0.29) >0.05 Disabled patients showed no significant differences in cardiovascular risk factors (Table 2), but they presented a higher prevalence of both micro- and macrovascular diabetic complications with the exception of myocardial infarction (Table 3). The GPs therapeutic prescribing pattern - both in disabled and non-disabled patients showed a predominant use for the former ones of sulphonylureas alone (35% and 32%, respectively); the use of metformin alone was limited (3% and 5%, respectively). The disabled patients significantly used more insulin treatment than non-disabled ones (6% vs. 0.3%, p<0.01) (Table 4). The follow-up data referring to the performance and outcome indicators are reported in Table 5. Overall, we found that 83% 90% of practices adhered with glycemic control indicators (number of blood glucose and HbA 1c measurements per year) and 75% 81% with micro- and macrovascular indicators (number of dilated eye examinations and ECG per year), respectively. GPs adherence to the follow-up protocol significantly improved over time (p for trend, Table 5). In particular we observed that the percentage of patients with no measurements dramatically halved from about one-third to one-sixth of cases

4 12 A.V. Ciardullo et al.: Shared-care program for diabetes Table 5 Follow-up data on performance and outcome indicators for 6267 subjects with type 2 diabete Year 1 Year 2 Year 3 p for trend Performance indicators, n (%) of subjects 3 blood glucose measurements annually (66) (80) (83) < HbA 1c measurements annually (75) (86) (90) < dilated eye examinations annually (65) (75) (81) < ECG examinations annually (59) (68) (75) <0.01 Outcome indicators a HbA 1c, % 7.6 1(1.6) 7.5 1(1.3) 7.3 1(1.3) <0.01 Subjects with HbA 1c 6.5%, n (%) 1454 (23.2) 1517 (24.2) 1717 (27.4) <0.05 Weight, kg a 77.1 (14.9) 76.2 (14.7) 76.6 (14.6) >0.05 BMI, kg/m 2a (4.8) (4.8) (5.0) >0.05 a Values are mean (SD) ECG, electrocardiography; BMI, body mass index (Table 5). As to the outcome indicators, we found that both mean HbA 1c value and the percentage of subjects with HbA 1c 6.5% significantly improved over time (Table 5). The body weight and BMI values did not improve over time (Table 5). Similar trends were observed in both disabled and non-disabled diabetic patients. Discussion The number of patients with type 2 diabetes is continuing to rise due to the increasing number of elderly people, the better recognition of prevalent undiagnosed diabetes and the better care for and survival of people with clinically diagnosed diabetes [6 9]. Although the rise in prevalence of patients with diabetes mellitus is mainly due to type 2 diabetes, alarming increases in type 1 diabetes have also been observed [10 12]. Achieving good glycemic control is important and widely recommended to prevent the occurrence of diabetic complications [13 16]. The main purpose of our shared-care program was to involve people with well-controlled type 2 diabetes. Therefore, it included by definition only well-controlled type 2 diabetic patients but this was true only for walking people. In fact, due to problems in the organization of diabetic services, our GPs also have to care the disabled diabetic patients regardless of the cause of disability. For this reason, we have also shown the statistical differences between the two groups in order to check that appropriate differences did exist. If they did not, this would have implied that there was a major healthcare problem and that subsequent actions should be undertaken for patients safety. Overall, our diabetic patients had on average good glycemic control (mean HbA 1c, 7.6%±1.6%). Furthermore, 24.2% of them had optimal control (HbA 1c <6.5%) while 34.7% had acceptable control (HbA 1c from 6.5% to 7.5%). The glycemic control also improved at follow-up. Similar patterns of glycemic control categories at baseline and follow-up were observed in both disabled and non-disabled diabetic patients. Our type 2 diabetic patients were aged and overweight and about two-thirds of them had hypertension or dislipidemia. Nevertheless, the baseline prevalences of both micro- and macrovascular diabetic complications in the non-disabled patients were not high and lower than those observed in the disabled patients, as expected. The comparison of therapeutic patterns between the two groups of patients showed, as expected, that the use of insulin therapy was more frequent in disabled people both in monotherapy (+27-fold) and in association with oral hypoglycemic drugs (+5- fold). This did not mean that insulin represents a risk factor for disability; on the contrary, the association should be interpreted in the sense that the disabled people were appropriately cared by GPs since insulin therapy was suitable for them. The follow-up performance indicators improved over time, and a parallel improvement of glycemic control was observed in both disabled and nondisabled people. Meeting the complex needs of patients with chronic illness or impairment is the single greatest challenge facing organized medical practice. Patients and families struggling with chronic illness have different needs, and these needs are unlikely to be met by an acute care organization and culture. They require planned, regular interactions with their caregivers, with a focus on function and prevention of exacerbations and complications. This interaction includes systematic assessments, attention to treatment guidelines, and support for the patient s role as selfmanager. These interactions must be linked through time by clinically relevant information systems and continuing follow-up initiated by the medical practice [1 5, 17]. Data from the present clinical audit suggest that family physicians are doing a good job of providing care for their patients with type 2 diabetes. The satisfactory results

5 A.V. Ciardullo et al.: Shared-care program for diabetes 13 we found even among the disabled patients cared by GPs are of potential interest for public health. In fact, universal healthcare systems have to face the progressive increase of disability due to aging while suffering from limited resources. Our data are in line with results from both randomized and non-randomized trials, i.e. structured care by GPs with an interest in diabetes and supported by a specialist liaison team produces comparable levels of care to those provided in hospital [1 5]. The Modena Progetto Diabete shares features with successful district-wide shared-care schemes according to current models for improving chronic illness care in health systems [17]: (a) an extensive planning phase in which objectives were carefully defined and the facilities, expertise and commitment of individual general practices were assessed; (b) locally developed, written evidence-based guidelines for diabetic management and their implementation through provider education, reminders, and increased interaction between generalists and specialists; and (c) a well-developed monitoring system gifted with performance and outcome indicators. The last feature is crucial because maintaining standards depends on regular audit, as has been shown before with chronic disease management programs in general practice [4, 5, 17]. Continuing medical education for physicians involved in Progetto Diabete was part of our program. We organize educational meetings addressed to diabetic patients and guided by both diabetologists and GPs. We have also paid attention to patients empowerment through their involvement in producing an easy-toread version of local guidelines [18]. In conclusion, our data showed a low prevalence of diabetic complications and a progressive improvement of follow-up indicators among well-controlled type 2 diabetic patients who accepted to be cared by the GPs adhering to our shared-care program. Acknowledgments We are grateful to all the diabetic patients and to the GPs participating in the Progetto Diabete in Modena Province. We thank Dr. Giuseppe Fattori (Azienda USL Communication and Marketing Unit) who financed the press release of our easy-to-read material and included it among the Local Healthcare Plan s information tools. The Progetto Diabete and the present work are supported by public healthcare authorities. References 1. Griffin S, Kinmonth A (1997) The management of diabetes by general practitioners and shared care. In: Pickup J, Williams G (eds) Textbook of diabetes. Blackwell Scientific, Oxford 2. Vrijhoef HJM, Spreeuwenberg C, Eijkelberg IMJG, Wolffenbuttel BHR, van Merode GG (2001) Adoption of disease management model for diabetes in region of Maastricht. BMJ 323: Laine C, Caro JF (1996) Preventing complications in diabetes mellitus: the role of the primary care physician. Med Clin N Am 80: Greenhalgh PM (1994) Shared care for diabetes: a systematic review. RCGP Occasional Paper 67. Royal College of General Practitioners, London 5. Griffin S (1998) Diabetes care in general practice: metaanalysis of randomised control trials. BMJ 317: Amos AF, McCarty DJ, Zimmet P (1997) The rising global burden of diabetes and its complications: estimates and projections to the year Diabetic Med 14[Suppl 5]:S1 S85 7. Burke JP, Williams K, Gaskill SP, Hazuda HP, Haffner SM, Stern MP (1999) Rapid rise in the incidence of type 2 diabetes from 1987 to 1996: results from the San Antonio Heart Study. Arch Intern Med 159: Meneilly GS, Tessier D (1995) Diabetes in the elderly. Diabetic Med 12: Roman SH, Harris MI (1997) Management of diabetes mellitus from a public health perspective. Endocrinol Metab Clin N Am 26: Gardner SG, Bingley PJ, Sawtell PA, Weeks S, Gale EA (1997) Rising incidence of insulin dependent diabetes in children aged under 5 years in the Oxford region: time trend analysis. The Bart s-oxford Study Group. BMJ 315: Libman IM, LaPorte RE, Becker D, Dorman JS, Drash AL, Kuller L (1998) Was there an epidemic of diabetes in nonwhite adolescents in Allegheny County, Pennsylvania? Diabetes Care 21: Onkamo P, Vaananen S, Karvonen M, Tuomilehto J (1999) Worldwide increase in incidence of type I diabetes the analysis of the data on published incidence trends. Diabetologia 42: The Diabetes Control and Complications Trial Research Group (1993) The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 329: UK Prospective Diabetes Study Group (1998) Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 352: van der Does FE, De Neeling JN, Snoek FJ et al (1996) Symptoms and well-being in relation to glycemic control in type II diabetes. Diabetes Care 19: Reichard P, Pihl M, Rosenqvist U, Sule J (1996) Complications in IDDM are caused by elevated blood glucose level: the Stockholm Diabetes Intervention Study (SDIS) at 10-year follow up. Diabetologia 39: Von Korff M, Gruman J, Schaefer J, Curry SJ, Wagner EH (1997) Collaborative management of chronic illness. Ann Intern Med 127: Ciardullo AV, Fattori G, Carrozzi G, Daghio MM (2004) Evaluation of an information tool for diabetic patient education. Diabetic Med (in press)