Many individuals attending a diabetes clinic have comorbid mental health problems; therefore, treatment of diabetes needs an integrated approach, incorporating both physical and psychological perspectives. The National Service Framework for Diabetes emphasises the importance of psychological well-being for individuals diagnosed with diabetes (Department of Health, 2001). Psychiatrists and psychologists have not been routinely integrated into the multidisciplinary diabetes team as, for example, pain clinics have. There are very few specialist diabetes centres in which a psychiatrist is a part of the diabetes team. The Beta Cell Diabetes Centre at the Chelsea and Westminster Hospital Foundation Trust was one of the first in the UK to establish a liaison psychiatry placement within a diabetes centre.

Several psychological conditions may occur in individuals with diabetes. Depression is the most common, but it is often unrecognised and untreated (Reference Anderson, Clouse and FreedlandAnderson et al, 2001). Also common are schizophrenia and substance misuse which can impair the individual's ability to manage diabetes. Alcohol interferes with the capability of the liver to store and release glucose and can complicate diabetes (Reference Saudek, Rubin and ShumpSaudek et al, 1997). A person with phobic disorders may be unable to do self-injections. The pressure of the need to eat the right food in the right amounts may lead to a feeling of food deprivation; anger and frustration may follow.

Methods

This prospective case-note survey covered the 12-month period from April 2006 to April 2007. At 12 months, case-note data were collected in a standardised manner. Information included the source and reason for referral, service user characteristics, type of diabetes, DSM–IV diagnosis at the end of assessment, treatment offered and outcome regarding attendance and discharge. The data were compared with the previous report from the clinic (Mitchell et al, 2000).

Service user characteristics

Thirty-two individuals were referred from April 2006 to April 2007 to the liaison psychiatry clinic. Seven failed to attend their appointment and three were still waiting to be assessed at the time of the study. One appointment was cancelled by the service user. Our results, therefore, show the outcome of 21 assessments. Out of the total number of individuals referred, 15 were male (46.8%) and 17 were female (53.1%), within the age range of 31–71 years. Out of the total 32 individuals referred (including those who did not attend their appointment, those who awaited appointments and those who cancelled their appointment), 14 individuals (43.7%) had type 1 diabetes mellitus and 18 (56.2%) had type 2 diabetes mellitus. Of the 21 individuals who were assessed in the clinic, 12 were married (57.1%) and 8 were single (38.09%; 2 were single parents and 1 was divorced (4.7%)). Other service user characteristics are given in Table 1.

Table 1. Baseline characteristics of individuals referred to a liaison psychiatry diabetes clinic – comparison with Mitchel et al's survey

Baseline characteristics	1997-1998 survey1	2006-2007 survey
Age range, years	26-72	31-71
Marital status, n (%)
Married	4 (17)	12 (57.1)
Single	16 (67)	8 (38.2)
Divorced	2 (8)	1 (4.7)
Gender, n (%)
Male	10 (42)	15 (46.8)
Female	14 (58.3)	17 (53.1)
Individuals referred, n (%)
Patients seen	24 (77.5)	21 (65.7)
Did not attend	7 (22.5)	11 (34.3)
Individuals with type 1 diabetes referred, n (%)	22 (70.9)	14 (43.7)
Individuals with type 2 diabetes referred, n (%)	9 (29.0)	18 (56.2)

Two individuals were seen on the day of the referral and the remaining 21 individuals received appointments within 2 weeks of the referral. Each of the seven individuals who failed to attend their appointments was sent two further appointments and every effort was made to establish contact with them personally (by telephone). Where this was not possible, letters were sent to them informing them of their next appointment. Out of these seven individuals, one had already established contact with her community psychiatric team, three had alcohol-related problems which they minimised, one had a needle phobia and the remaining two had poor adherence with insulin.

Reasons for referral

A common reason for referral was low mood (n=12, 37.5%) which led to non-adherence with diabetes medication and poor glycaemic control. The next most common reason for referral was anxiety (n=3, 9.3%), followed by personality difficulties (n=2, 6.3%). Most individuals were referred to the liaison clinic by a diabetes specialist nurse (n=17; Table 2).

Table 2. Referral by profession

Referral by profession	Individuals referred 2006-2007, n
Diabetes specialist nurse	17
Consultant diabetologist	6
Medical specialist registrar	3
Associate specialist	5

Psychiatric diagnosis

Out of the 21 individuals who attended their appointments at the clinic, the most common psychiatric diagnosis was adjustment disorder (n=6, 28.5%). Two of the individuals with adjustment disorder (9.5%) had a comorbid diagnosis of alcohol dependency and anxiety respectively. Five individuals had depression (23.8%) and two had personality difficulties (9.5%). One person (4.7%) each had a binge eating disorder, alcohol dependence, schizophrenia, needle phobia and anger problems. One individual did not have any kind of psychiatric problem and was referred to the clinic because of non-adherence of insulin which was thought to be caused by depression. We have compared the results of our survey with that by Mitchel et al (Reference Mitchel, Catalan and McIntosh2000) (Table 3).

Table 3. Principal psychiatric diagnosis comparing the two reports

	Individuals diagnosed with a psychiatric problem, n (%)
Diagnosis	1997-19981	2006-2007
Adjustment disorder	3 (12.5)	6 (28.5)
With comorbid anxiety	-	2 (9.5)
With comorbid alcohol dependence	-	2 (9.5)
Depression	9 (37.5)	5 (23.8)
Personality disorder	5 (21)	2 (9.5)
Needle phobia	1 (4.1)	1 (4.7)
Eating disorder not otherwise specified	2 (8.3)	1 (4.7)
Bulimia	1 (4.1)	-
Hypoglycaemic anxiety	-	1 (4.7)
Schizophrenia	1 (4.1)	1 (4.7)
Bipolar disorder	-	1 (4.7)
Sexual dysfunction	-	1 (4.7)
Anger problems	-	1 (4.7)
Alcohol dependence	2 (8.3)	1 (4.7)
Alcohol-induced mood disorder	1 (4.1)	-
Alcohol-induced anxiety disorder	1 (4.1)	-
Adverse effects of medications	1 (4.1)	-
Bereavement	1 (4.1)	-
Personality traits or mental disorder affecting medial condition	9 (37.5)	-
No psychiatric problems	3 (13)	1 (4.7)
Total number of individuals referred to the liaison psychiatry diabetes clinic	31	32

Discussion

In this survey, adjustment disorder was the most common psychiatric disorder diagnosed in the liaison psychiatry diabetes clinic (n=6, 28.5%). Common mental disorders are more prevalent in our clinic than severe mental illness which is more likely to be encountered in an exclusive psychiatric setting.

In this cohort, individuals with an adjustment disorder were distressed by problems of living with diabetes - they were ashamed to inject insulin at the workplace and frustrated with unpredictable blood glucose levels, which led to demotivation for monitoring. Many harboured inaccurate assumptions about insulin, relating it to end-stage disease.

Fear about taking insulin was common. It is important to recognise erroneous assumptions about diabetes, like those mentioned above, as they may be treated by psychoeducation and cognitive-behavioural therapy. Diagnosing depression in diabetes can pose a challenge, as high blood glucose level in itself can lead to depressive symptoms, like fatigue, changes in sleep patterns, changes to appetite and weight. The feelings of helplessness and hopelessness that manifest in depression may contribute to poor self-management of diabetes. For many participants in our survey, diabetes was a life event which triggered depression, leading to only partial adherence to medication and worsening of diabetes.

There was no significant difference between service users’ baseline characteristics in both periods. The level of non-attendance to the clinic was 34.3% in our study, as compared with 22% in the previous study. There were more individuals with type 2 diabetes referred to the clinic between 2006 and 2007 (56.2%) than in 1997–1998 when the majority had type 1 diabetes (70.9%).

Psychoeducation formed an important aspect of the treatment provided in this clinic and has a good evidence base in the treatment of diabetes (Reference Rubin, Peyrot and SaudekRubin et al, 1989), as many individuals are distressed by the demands of living with diabetes such as careful eating, drawing blood and checking insulin level several times a day. This was seen to lead to dysphoric feelings and a fear of taking insulin, followed by a lack of motivation to monitor one's blood glucose levels, which in turn led to complications. Six of the referred individuals were treated with medications – five with antidepressants and one with an antipsychotic.

This liaison psychiatry diabetes clinic is a unique pragmatic model, with a psychiatrist embedded within the multidisciplinary diabetes team. This is unique in that in most liaison clinics, such as cardiology or dermatology, it will be a psychologist who is part of the team. Working with the liaison psychiatrist, apart from providing support to the diabetes team, allowed for psychiatric assessment, diagnosis and treatment in the diabetes centre, thereby reducing stigma of attendance to a mental health clinic. It was also no longer necessary to refer service users to busy and overstretched inner city community mental health teams. The psychiatrist was in direct contact with the diabetes team and they discussed each individual referred to the clinic. Psychiatric problems could be early detected and treated.

Liaison psychiatric teams also educate and support medical teams in recognising symptoms of psychiatric illness; such was the case in the liaison psychiatry diabetes clinic. Despite the focus on psychiatric diagnosis and treatment, a large majority of service users (18 out of the 21 assessed) were also seen by clinical psychologists from the psychological medicine team after initial assessment by the psychiatrist in the liaison psychiatry diabetes clinic, as they required psychotherapeutic intervention.

The diagnostic differences between the present and Mitchel et al's (Reference Mitchel, Catalan and McIntosh2000) report could be attributed to different diagnostic approaches.

Declaration of interest

None.