Interdisciplinary diabetes care teams operating on the interface between primary and specialty care are associated with improved outcomes of care: findings from the Leuven Diabetes Project

Despite its multi-system effects, diabetes is a controllable disease, and there is unequivocal evidence that early and proactive, continuous monitoring and treatment can significantly reduce its human and economic toll [1‐3]. Many guidelines provide targets that are desirable for most patients with diabetes[4]. Literature demonstrates however that many patients with type 2 diabetes mellitus (DM) still don't receive the care they need[5], as physicians overrate the quality of the care they already deliver and substantially underestimate the number of patients in need of intensified pharmacotherapy and lifestyle interventions[6]. We therefore need quality improvement programs that promote comprehensive and proactive diabetes care at different levels of the health care system. Components of such care that many physicians find difficult to provide include risk factor reduction, periodic examination for early signs of complications, ongoing education and behavioral interventions and health promotion[7]. The failure to initiate or to intensify therapy in a patient not at evidence-based treatment targets is defined as 'clinical inertia'[8]. Besides physician factors the principal sources for clinical inertia include patient factors (e.g. resistance to adopting lifestyles that support optimal disease care) and organizational factors (e.g. a lack of a team approach to care)[9, 10]. These three sources interact in complex ways and quality improvement interventions to reduce clinical inertia therefore need to be multifactorial in nature[11]. A multifactorial approach to chronic care delivery is advocated by multiple authors that have developed conceptual models for this purpose [12‐14]. A widely used conceptual model for improving chronic illness care is the Chronic Care Model (CCM)[15]. Despite a diabetes prevalence rate of 7.9% in Europe[16], very few countries in Europe have national or even local programs that target this chronic condition building on multiple dimensions of the CCM. Previous studies have shown that it is not the intensity of the program, i.e. the number dimensions of the CCM that are applied which determines outcomes of care[17]. Important predictors of success are the type of interventions and the implementation strategies applied[18]. One such important component of quality improvement programs in diabetes care are considered interdisciplinary diabetes care teams[19]. These teams often focus on clinical guidance to primary care physicians and the provision of a broad range of educational services to patients[20]. There is a particular paucity of evidence with regard to the effectiveness of interdisciplinary diabetes care teams operating on the interface between primary and specialty care[21]. This interface refers to collaborative relationships between primary and specialty care aimed at improvements in efficiency and effectiveness of care. This study reports on the effectiveness of primary care-based interdisciplinary diabetes care teams that are actively guided and supported by a specialist team from secondary care. The teams are embedded in a two-arm multifaceted quality improvement program targeting adherence to guidelines and clinical inertia in primary care physicians.

A first study hypothesis was that the implementation of an interdisciplinary diabetes care team (IDCT) in support of primary care physicians is feasible in a health care setting with limited tradition in shared care. A second hypothesis was that patients who make use of an IDCT would have significantly better outcomes compared to non-users of the IDCT (irrespective of the intervention arm) and after an 18-month intervention period. A third hypothesis is that patients who used the IDCT in an advanced quality improvement program (AQIP) would have better outcomes compared to users in a usual quality improvement program (UQIP).

This study (The Leuven Diabetes Project) is intended to create the basis for the development of a national diabetes care program that set out national standards for the care of people with diabetes with the aim to raise the quality of services and to promote interdisciplinary shared care.

The study protocol of our trial has been peer-reviewed and published previously[22]. We summarize information on the study design, participants and key interventions of the IDCT. For detailed information on the sample size, randomization, allocation concealment, method for data collection and detailed descriptions on the interventions provided we refer to our study protocol and the CONSORT Checklist of items of cluster randomized trials[23], as presented in Table 1.

The results of this trial demonstrate that the use of interdisciplinary diabetes care teams which are actively guided and supported by a specialist team from secondary care are associated with important improvements in outcomes of care. Another important finding of our study is that the high number of physicians that made use of these teams demonstrates that the implementation of interdisciplinary shared care in a health care setting with limited tradition in shared care is feasible. The expected participation rate of 80% in physicians was easily attained, with physicians in the AQIP to make more use of the IDCT compared to physicians of the UQIP. The latter is possibly explained since physicians in the AQIP were actively encouraged by the program manager to make use of the IDCT whereas physicians in the UQIP were not. These findings highlight the need for specific implementation strategies in settings that have limited tradition in shared care. The higher referral rate in AQIP was however not associated with better patient outcomes which possibly indicates that these differences were not large enough to induce a significant effect. The relative absence of clinical significant differences between users of the IDCT when comparing the two intervention arms might also be related to the limited use of supplementary interventions as provided to physicians and patients of the AQIP. For example, very few physicians or their patients consulted the health psychologist or the travelling educator. These service offerings were relatively unknown to physicians and patients prior to the start of the project. Another possible reason for the aforementioned finding is that patients in the UQIP that used the IDCT had significantly worse levels in their outcomes at baseline that allowed for greater improvements over the 18-month intervention period.

In contrast to the overall participation rate of physicians, patients less often made use of the IDCT (12,5%) which means the expected participation rate of 30% in patients was not attained. It is not clear to what extent this is related to either physician referral patterns, patient behaviours or both. Physicians generally refer to a diabetes care team once their threshold for comfort with diabetes management has been surpassed [33]. In our study higher levels of HbA1c, LDL-cholesterol and insulin initiation indeed seemed important triggers for referral to the IDCT. As physicians in our study were recommended to refer to the IDCT in case the treatment targets were not met, they have focused on those patients that showed an important need for interdisciplinary shared care. Data show however that a substantial proportion of the patients who did not reach treatment targets did not make use of the IDCT. This might in part be explained by clinical inertia in physicians, but it is reasonable to assume that physicians had difficulty to motivate their patients to consult the IDCT despite their own convictions on the added value of the team in the co-management of diabetes. More than 95% of diabetes care is done by the patient, and health professionals have very little control over how patients manage their illness between office visits. Patients manage their diabetes on a daily basis within the context of the other goals, priorities, health issues, family demands, and other personal concerns that make up their lives[34]. Previous studies have shown that patients with diabetes often show poor compliance with diet, exercise, medication regimen and even appointment compliance[35]. These behaviours are in their turn affected by numerous variables, such as the nature of the patient-physician relationship, complexity of regimen, disruption of lifestyle, emotional support, financial resources, education in self-management skills, cues to action, perceived barriers, locus of control and motivation[36]. In our study patients who were not at target might have been unwilling to consult the IDCT, despite recommendations of their PCPs. These findings highlight the need for ongoing patient empowerment strategies in the context of the physician-patient relationship as well as clear referral indications in order to prevent over and under use of interdisciplinary shared care services in both physicians and patients. The broad referral indication that was applied in our project was probably not specific enough to guide PCPs in their decision to refer patients to the IDCT. The American Diabetes Association (ADA) standards currently do not indicate a minimum frequency for the provision of diabetes education, including dietary advice[4]. Other and clearer referral criteria for specialist services defined by the ADA are e.g. recurrent hypoglycaemia, poor glycaemic control and persistent hypertension and/or hyperlipidaemia despite intensive management and the need for psychosocial/counselling support to overcome barriers to self-care.

As an explanation to the positive results that are associated with the use of the IDCT we assume that particular team characteristics such as quality task orientation and shared leadership with clearly defined and valued group goals that the teams were trained at prior to the intervention might possibly have contributed to the success of the team. There is an extensive basis of research demonstrating the aforementioned team characteristics are associated with improved outcomes of care [37‐39].

Limitations to our study design are twofold. The UK Medical Research Council (MRC) Framework that we used for the development of our complex intervention demonstrated the difficulty of balancing optimum study design with designing interventions that were practical enough to be applied in family practice. The MRC Framework provides a methodological rather than an explanatory approach to evaluating complex interventions. This means that in order to fully understand and predict problems of workability and integration of complex interventions the additional use of psychological and sociological models is important in the design and evaluation of complex interventions. The Normalization Process Model [40] is a recent sociological model that asks what people do to make a complex intervention workable, and to integrate it in practice. The model proposes that complex interventions are implemented in processes in which the collective action and interaction of patients, professionals and others are governed by four factors. These factors are interactional and skill-set workability and relational and contextual integration. Since we did not make use of such additional social and psychological models in the design of our program, we recognize this is a limitation to our study protocol. We also have not integrated qualitative methods within the context of a pilot trial that could have helped to interpret the quantitative results of the pilot trials by clarifying process and outcome results. What we consider a particular strength of the study is it provides answers to hypotheses on innovative types of change interventions (such as interdisciplinary teams operating on the primary-specialty care interface). These hypotheses were tested using a large group of physicians and patients during an 18-month period. Most quality improvement programs reduce the intervention period to six months and are limited to smaller groups. A six-month period is short considering the Hawthorne effect that probably has not been washed out. Another strength of our study is that the primary and secondary care interface is strongly valued since we recognize a well-operationalized interface as an important attribute to high quality diabetes care [41]. More specific, the clinical leadership and coaching provided by a diabetologist to both the primary care physicians and the interdisciplinary diabetes care team is of particular importance in fragmented health care systems including the one from Belgium. Another strength of the study is the explicit project focus on multiple cardiovascular risk factors. A systematic review published in 2001 observed that most interventions did not pay enough attention to patient outcomes and if so, only changes in glycemic control were evaluated [42]. A last strength of the study is the use of all six dimensions of the Chronic Care Model (CCM), which is to our knowledge the fourth study in the field of diabetes care to do so [43‐45]. The use of all six dimensions of the CCM allows for an evaluation how some of the CCM components are associated with improved outcomes and thus it provides evidence to support the validity of this model.

Future research that will complement this present study will include 1. a qualitative analysis on perceived barriers to high quality diabetes through PCPs eyes, 2. a quantitative study on how patients' experiences with perceived quality of patient-centered care and self-management support are associated with documented diabetes outcome of care indicators and 3. a health economic assessment of the cost-effectiveness of the trial.

Overall, we assume that the increasing prevalence of diabetes, combined with its complexity, rapidly evolving medical therapies and the requirement for patient self-management will lead to a growing demand for multifaceted and interdisciplinary shared care services that target improvements in diabetes care[46].

Interdisciplinary care teams that are part of a multifaceted quality improvement program are associated with improved outcomes of care. More research is however required on what type of team and program characteristics contribute to improvements in diabetes care. There is a particular need for clear referral indications in order to prevent over and underuse of interdisciplinary care services in persons with type 2 DM.