Adoption of disease management model for diabetes in region of MaastrichtBMJ 2001; 323 doi: https://doi.org/10.1136/bmj.323.7319.983 (Published 27 October 2001) Cite this as: BMJ 2001;323:983
- H J M Vrijhoef, research fellowa (, )
- C Spreeuwenberg, professora,
- I M J G Eijkelberg, research fellowb,
- B H R Wolffenbuttel, associate professorc,
- G G van Merode, associate professorb
- a Department of Health Care Studies, Faculty of Health Sciences, University of Maastricht, PO Box 616, 6200 MD Maastricht, Netherlands
- b Department of Health Organisation, Policy and Economics, Faculty of Health Sciences, University of Maastricht, PO Box 616, 6200 MD Maastricht, Netherlands
- c Department of Endocrinology, University Hospital Maastricht, PO Box 5800, 6202 AZ Maastricht, Netherlands
- Correspondence to: H J M Vrijhoef
- Accepted 5 September 2001
Chronic diseases and associated conditions will always pose a challenge to healthcare systems. New healthcare models are being introduced in Western countries in response to a set of problems that are evident to some degree in all health services—for example, uncoordinated arrangements for delivering care, bias towards acute treatment, neglect of preventive care, and inappropriate treatment.1 These models take account of the pressure on quality and costs of chronic care and originate from the overlapping approaches of integrated care (United States) and shared care (western Europe). 2 3 In the Netherlands, shared care models have acted as a precursor of the recently introduced concept of disease management. Although several disease management initiatives are emerging, the model is not being adopted as fast as might be expected from the benefits that are claimed to result from it.4 In this article we describe the process of moving from a shared care model to a disease management model by considering recent developments in diabetes care in the region of Maastricht. We also describe the use of health technology assessment to evaluate the model.
The concepts of integrated care and shared care can be regarded, in the Netherlands at least, as the precursor of disease management
Implementation of a shared care model for diabetes in the region of Maastricht ensured that necessary conditions were met for a disease management model
Widespread use of disease management models is hampered by lack of evidence
Evidence from health technology assessment is necessary to justify large scale use of disease management models but will not be sufficient by itself
Shared diabetes care in Maastricht region
Shared care for patients with stable diabetes mellitus type 2 who were receiving care from an endocrinologist in an outpatient clinic was implemented in the region of Maastricht in 1997. Two changes were made from usual care: patients were seen by a nurse specialist rather than an endocrinologist and in general practice rather than an outpatient clinic.
Activities traditionally done by doctors were taken over by nurse specialists, who had the highest qualification level in nursing care and were specialised in diabetes mellitus. The nurse specialist provided direct patient care (taking the medical history, physical examination, interpretation of laboratory results, administration, and assimilation of findings), organised and coordinated care for individual patients (identification of shortcomings in care, referral to and communication with care providers), and provided advice and education to patients and other care providers. Thus, instead of quarterly consultations with the endocrinologist in the hospital, patients received three quarterly consultations with the nurse in general practice and one annual consultation with the endocrinologist.
Appropriate care was formulated in a protocol based on existing guidelines for diabetes care. If complications or other problems arose, the nurse specialist consulted the general practitioner or the endocrinologist. The first point of contact for patients was the nurse during office hours and the general practitioner outside office hours. The general practitioner provided the nurse specialist with a workplace in the practice and was ultimately responsible for patients' care.
The change in the structure of diabetes care was evaluated in terms of outcomes of care as well as change processes. The effects on patient outcomes were assessed by comparing the shared care model with the traditional care model in a 12 month non-equivalent control group study.5 General practitioners referring diabetic patients to Maastricht University Hospital were asked to choose one of the models of care, thus precluding random allocation of patients and a randomised controlled trial. Among patients of the 22 general practitioners choosing the shared cared model, 105 patients with stable diabetes type 2 were eligible and 74 gave informed consent. Twenty nine general practitioners continued to use the traditional model. From these practices, 60 patients were eligible and 47 gave informed consent. Between group analyses were done for patients receiving oral hypoglycaemic drugs or insulin, resulting in an intervention group of 52 patients. The study found no difference between the groups in quality of life, knowledge of diabetes, self care behaviour, patient satisfaction, or consultations with caregivers; glycaemic control improved in patients receiving shared care but deteriorated in patients receiving traditional diabetes care (mean (SD) glycated haemoglobin concentrations were 8.3% (1.5%), 8.2% (1.2%), and 8.2% (1.0%) at 0, 6, and 12 months in the intervention group and 8.2% (1.1%), 8.4% (1.1%), and 8.5% (1.4%), respectively, in the traditional care group.5
Factors that particularly influenced the processes of change during the development and implementation of the shared care model were investigated by case study methods.6 Evaluation was based on in-depth interviews with participants, written questionnaires sent to several care providers, observation of project meetings, and relevant documents and reports.7 From a management perspective, the predominant factors affecting care seemed to be project management, commitment, and local context (culture, power, and structure). This implied that existing rules about the right and wrong ways of doing things were questioned and redefined into rules on “doing the right things.” 7 8 The skills and intentions of the project managers also had a role. Finally, long term strategies and tactics that enabled learning processes seemed the most effective, especially when the small scale shared care model changed into a large scale disease management model.7
From shared care to disease management
The implementation of the shared care model meant that the conditions were met for the introduction of the disease management model. All participants, with their unique values, goals, and interests, trusted each other sufficiently and were willing to share data. Furthermore, consensus was reached between patients, doctors, and nurses on the design of the model for diabetes care, treatment guidelines, and target outcomes.
The roles of the caregivers were sufficiently changed to allow a shift to a diabetes disease management model. Nurse specialists had taken on roles traditionally done by doctors. Medical specialists had fewer consultations than in usual care and, like general practitioners, had more of a consulting role to both patients and other partners in diabetes care. General practitioners viewed the nurse as an interface between primary and secondary care as well as a caregiver with specialised knowledge. Furthermore, the model was accepted by outpatients with stable diabetes.
The evidence showed that the shared care model was beneficial in terms of both process and outcomes, and general practitioners were asking for the role of the nurse specialist to be extended to include care of other diabetic patients. In January 2000, the shared care model for stable type 2 diabetic patients was therefore superseded by a disease management model for all diabetic patients in the Maastricht region.
Diabetes disease management
Diabetes care within the disease management model is delivered by a core team comprising general practitioner (with nurse practitioner or practice assistant), endocrinologist, and nurse specialist. Based on the intensity of care required and the patient's preference, all diabetic patients of participating general practitioners are assigned to one of three routes: high, medium, and low intensity. Patients who require high intensity care fall under the prime responsibility of the endocrinologist. The nurse specialist is the main caregiver for patients requiring medium intensity care, and the general practitioner manages patients requiring low intensity care.
The medium care route is comparable with care in the shared care model, but because nurse specialists have a role in the care of patients in all routes they have a stronger linking role with endocrinologists (hospital care) and general practitioners (primary care). The members of the core team also cooperate closely with other caregivers such as dieticians, community nurses, ophthalmologists, and podiatrists.
Routes, responsibilities, and tasks are formulated in a protocol. Furthermore, care is provided according to accepted guidelines. Prevention and timely detection of complications are given a high priority. Supported by patient management tools, the teams encourage patients to take a more active role in their care. Data on patient contacts are systematically collected and analysed, facilitating follow up of patients and monitoring of care as well as providing opportunities to improve care by disseminating data on variation between caregivers.
The disease management model aims to optimise the quality of diabetes care. In January 2001, 42 general practitioners with 2100 patients were taking part. The involvement of different parties and disciplines in the model means that various approaches and strategies for achieving optimal care are combined.9
Evidence of effectiveness
Wide scale use of disease management has been hampered by lack of data on its effectiveness. 4 10–13 Existing evidence mainly concerns efficacy (impact in optimal circumstances) rather than effectiveness (impact in daily practice). In addition, almost no data regarding costs have been gathered, and it is unclear which outcome measures should be used.4 We are using health technology assessment to evaluate the diabetes disease management model in Maastricht in order to provide evidence on its short and long term effects.
Since disease management is a complex multifaceted strategy, we are using a method that includes several approaches.14 As well as assessing cost effectiveness, the method will enable us to identify the requirements of disease management in terms of organisation, governance, and incentives. It will also provide recommendations for adoption, application, and dissemination of disease management models in daily practice. The challenge is to identify which patients will benefit from disease management and which outcomes will be improved. 15 16 The study should also yield a method of evaluating disease management models that can be accomplished quickly and take advantage of emerging databases and information systems.17
The shift from shared care to a disease management model in Maastricht came about partly through the demand from general practitioners for the diabetes nurses to expand their care from stable type 2 diabetic patients to other diabetic patients. This demand was supported by evidence that the shared care model was beneficial. Similar changes have taken place in the management of patients with chronic obstructive pulmonary disease in the region of Maastricht.
Although the evidence obtained through health technology assessment should help increase the use of disease management models, the technique is itself faced with barriers to implementation.15 These barriers may be placed by policymakers (differing perspectives, timeliness and accessibility of health technology assessment findings, reliability of study findings, incentives, and uncertainties), healthcare professionals (practice environment, knowledge and beliefs, lack of consensus, autonomy, and uncertainty), and the general public (financial barriers, information asymmetry, attitudes, and behaviour).15
Although evidence is essential to increase dissemination of disease management models, legal, ethical, and organisational aspects as well as the social implications of switching to disease management also have a role. 4 14 17 The applied strategy of health technology assessment contains many elements for successful implementation, but it remains to be seen whether its findings will be put into practice.
Contributors: HJMV and CS had the original idea. All the authors participated in writing the paper. CS and BHRW will act as guarantors.
Competing interests None declared.