Implementation of recommended type 2 diabetes care for people with severe mental illness – a qualitative exploration with healthcare professionals

The prevalence of diabetes is increasing globally, affecting an estimated 8.8 % of adults – 415 million people worldwide – and accounting for 12 % of international health expenditure [1,1]. In the UK, 6.2 % of the adult population are estimated to have diabetes and as in other high income countries around 90 % of have type 2 diabetes [1]. Although there is no single cause of the condition, one important risk factor is diagnosis of a severe mental illness (SMI), such as schizophrenia, bipolar disorder and other psychoses. Research indicates a 2–3 fold increased incidence of type 2 diabetes among people with SMI [2]. This increased risk has been attributed to a number of factors, including the effects of anti-psychotic medications [3], lifestyle factors such as poor diet, obesity and physical inactivity [2, 4] and high rates of smoking [5]. As a consequence, in people diagnosed with diabetes, those with SMI have been found to die significantly younger than people without SMI [6] and experience a greater risk of diabetes complications that require specialist treatment [7].

These significant health inequalities could be explained by the inequity of diabetes care between people with and without SMI. People with SMI are reported to be less likely to receive recommended diabetes care, including retinopathy screening; foot examinations; HbA1c, renal and cholesterol checks; and diabetes education. They are also less likely to meet recommended blood glucose targets than those without SMI [8‐13]. Although some studies have failed to find any significant differences in diabetes care between people with and without SMI [14‐18], this has been attributed to the greater number of medical visits made by people with SMI compared to those without SMI [14, 15], a bias towards data that includes only service users who are registered with a general practitioner (GP) [18] or studies that recruit primarily male veterans [13‐16] making generalisability of these latter findings problematic.

The reasons for these disparities in care are largely unknown, which means that the development of interventions to ensure that people receive recommended care are not yet possible. Care pathways for people with a SMI and diabetes are often complex and fragmented, yet co-location of mental and physical services seems to have had variable impact - increasing the likelihood of receiving diabetes-related foot examinations for example, but having no effect on rates of retinopathy screening [19]. There also appears to be confusion and role ambiguity about responsibility for monitoring and supporting type 2 diabetes in SMI, between primary care and specialist mental health care services [20]. This is despite the fact that guidelines for the management of psychosis and schizophrenia state that specialist mental health teams should assume the role of monitoring service users’ physical health for at least the first 12 months or until the person’s condition has stabilised, thereafter responsibility should be transferred to primary care [21]. The knowledge and skills of mental health nurses to deliver and support diabetes care has also been questioned [20, 22, 23].

The purpose of this study was to explore, using an established theoretical framework, a broad range of barriers and enablers that may affect the practice of primary care, specialist mental health and diabetes specialist clinicians in their management of type 2 diabetes in people with SMI. It is hoped that the study will increase knowledge and understanding of individual and organisational level interventions that may support healthcare professionals in this endeavour and inform further research on this important topic.

In using the TDF [25], our study has provided a comprehensive understanding of the enablers and barriers to delivering effective care for type 2 diabetes in people with SMI, from the perspective of various healthcare providers. The importance of all theoretical domains illustrates the complexity of diabetes management in this population. Informed by the Medical Research Council (MRC) guidelines [35], use of the TDF provides a foundation for developing a theoretical- and evidence-based intervention by providing a basis on which to model these relationships in a larger sample of participants and subsequently select key behaviour change techniques to be implemented in an intervention.

Knowing how to manage and monitor type 2 diabetes was a significant barrier to implementation. There was poor awareness, particularly amongst mental health professionals, about national and local guidelines for managing type 2 diabetes. As a consequence training on effective diabetes management was identified as a need by all mental health professionals, not just mental health nurses as reported elsewhere in the literature [22]. Theoretical- and evidence based diabetes educational interventions for mental health professionals are scarce. A single group pre-post study [36] reporting the development and evaluation of a diabetes educational package for mental health nurses found significant improvements in diabetes knowledge. However, the quality of the evaluation was poor with non-validated measures, small sample size and no pre-defined hypothesis or power calculation employed.

In order to engage service users in their diabetes care it was recognised that specialist communication and negotiation skills were required, and training needed. The inability of some service users to engage or have insight into their illness and treatment was seen as a significant barrier and therefore development of these skills was deemed vital to engaging the population. Difficulties engaging service users due to limitations in cognitive and executive functioning are also a barrier to preventing diabetes in people with SMI [37, 38]. These findings indicate a clear need for educational and skills based training in order to develop competences in effective communication and behaviour change, as well as implementation of guidelines for managing type 2 diabetes.

As found in the broader diabetes literature [39, 40] time constraints were seen as a significant barrier to delivering effective diabetes care to people with SMI. Despite these time constraints a majority of the sample did however, acknowledge that certain aspects of diabetes care were within their professional role. These specific roles and responsibilities did differ by profession. All participants saw it as their role to refer service users on to relevant diabetes-related services; support and educate people to manage their diabetes and lead healthier lifestyles; and monitor or help service users to monitor their diabetes.

Diabetes medication initiation and titration was deemed solely the responsibility of primary care and specialist diabetes services, as reported by mental health nurses elsewhere in the literature [22]. Whereas, mental health specialists were clear about the importance of physical health monitoring and their role in this, suggesting a clear demarcation of responsibilities, rather than the suggested confusion and role ambiguity [20].

There does however, seem to be discrepancy between participants’ identity as a professional within their workplace, and individual goal priorities. Whilst participants were clear about their role and the policies of the organisation, diabetes was not considered a priority over the management of mental health by any of the sample, with mental health more often the primary focus. The concept of goal priority, the likely prioritization of one goal over another, has been relatively under-researched in relation to health-related behaviours. Recent research does however indicate that when levels of goal priority are high, the intention-behaviour relationship is stronger [41]. This is particularly salient for this study, as participants lacked any intention to manage type 2 diabetes in people with SMI. Although an explicit lack of intention should be addressed by developing or increasing intentions [42], such endeavours are unlikely to lead to changed behaviour unless establishing diabetes management as a healthcare priority is also achieved.

Reviewing past behaviour, either as individuals or part of their clinic teams, helped participants to manage and monitor diabetes. These reviews tended to be in relation to poor performance or practices, in order for staff to reflect, learn and then develop more appropriate action plans for future efforts. Encouraging healthcare professionals to focus on past successes is a mechanism for increasing perceived behavioural control [43], but there has been little investigation of how reflecting on past failures may also be beneficial. This study indicates that doing so may enable healthcare professionals to develop action plans that are more relevant to their individual practices and past performances. Action planning, one of the most prolific behaviour change techniques, has been found to be effective in changing behaviour across a range of contexts [44], and could therefore be a route to improving the care offered to this population.

Integration of mental and physical health care services had the potential to be both a barrier and enabler to delivering diabetes care to people with SMI. This was either as a result of poorly integrated IT systems that prevented easy referrals between services or the inability to access patient healthcare records in order to confirm diagnosis, obtain test results or communicate with other members of the healthcare team. Being able to easily refer service users to other diabetes-related services as a result of good organisational structures was an enabler to effective care. Lack of cooperation between psychiatry and primary care and poor knowledge about the organisation of diabetes care have previously been identified as barriers to preventing diabetes in people with SMI [37, 38]. The current data suggests that when integration and easy access existed this allowed healthcare professionals to seek support and advice from their respective physical and mental health colleagues. The nature of this integration and access was however unclear, and since co-location of mental and physical services seems to have variable impact [19], it maybe not just be proximity that prevents integration but other barriers to engaging healthcare professionals in integrated care, which need further exploration.

Supportive family members and carers made it easier for healthcare professionals to manage type 2 diabetes in people with SMI, as found in the prevention of diabetes [37, 38]. This is in line with recommendations that mental health services employ the Triangle of Care initiative [45], which encourages mental health organisations to ensure staff involve and work closely with family members and cares, where possible, to improve care.

Feelings of hope and optimism about managing diabetes and the health of service users were mixed. This may be an important focus for interventions aimed at health professionals as hope and optimism for the future have been identified as important concepts in the facilitation of recovery in people with SMI [46] and may potentially impact on other health behaviours. There may be a complex interaction between hope and recovery with an important role for social context and interpersonal relationships, particularly for those healthcare professionals who occupy a powerful position in relation to service users' hope. Healthcare professionals must carefully consider how they communicate their own hopefulness about people’s recovery [47], including how they ‘hold’ hope for service users during times of struggle and despondency.

Expecting negative physical and psychological outcomes in patients if diabetes was not effectively managed acted as an enabler to action. Fear of discipline, if anything should occur as a result of absent or poor care, and pay-for-performance schemes were thought to be important reinforcing factors. Blame and punishment are felt by healthcare professionals to be part of health service culture, particularly when someone is involved in an error, near miss or incident [48]. However, considering that the evidence for pay-for-performance systems, such as QOF, in changing healthcare professional behaviour is limited and for their efficiency scarce and inconclusive [49‐51] these beliefs appear to be unsupported in the literature.

There are several limitations to our study. Type 2 diabetes is a complex condition that requires a range of monitoring and management; as a result participants were reflecting on a range of targeted behaviours. In order to explore how these beliefs impact on specific diabetes-related behaviours, work is currently underway to understand the enablers and barriers to specific diabetes-related behaviours in a larger sample of healthcare professionals, using these beliefs statements as the basis for a large scale survey. The qualitative nature of the study precludes generalisability to the wider population, as does sampling within one inner city service provider. The snowball recruitment strategy may also mean that healthcare professionals particularly interested and involved in the management of diabetes in people with SMI were approached to take part, which may have biased these findings. Although the sample contained small numbers of each professional group, saturation of themes was achieved. As opposed to the actual causes of behaviour these findings reflect participants’ perceptions of the barriers and enablers to implementing the target behaviours [52], which may not necessarily translate into performance of the behaviour [53]. Finally, there was significant variation in inter-rater reliability. Although use of the TDF provides significant advantages including its theoretical underpinnings, synthesis of concepts from a range of behaviour change theories and systematic approach to identifying a broad range of barriers and enablers [54] it comes with challenges. A lack of clear operational definitions for each of the domains [54] often makes coding of the transcripts difficult, particularly for those without a background in psychology. This is likely to explain the variation in inter-rater reliability found within the study.

The findings of this study indicate a need for individual and organisational level interventions to support physical and mental health professionals in the delivery of diabetes care for people with SMI. At an individual level, increasing knowledge and optimism, improving skills and supporting healthcare professionals to develop action plans and intentions for their own practice could be effective in addressing disparities in care. At an organisational level, creating links and pathways between mental health and physical services to promote integrated care, prioritizing diabetes within mental health settings and clearly defining the roles and responsibilities are identified as potentially important avenues for intervening. Future work will involve exploring these specific barriers and enablers in a larger sample of healthcare professionals in order to model these outcomes more robustly in line with the MRC guidance [35] for developing complex interventions.

CQUIN, commissioning for quality and innovation; GP, general (medical) practitioner; MRC, Medical Research Council; NICE, National Institute of Health and Social Care Excellence; QOF, quality outcomes framework; SMI, severe mental illness; TDF, theoretical domains framework