Abstract
Background
Individuals with diabetes and co-existing psychiatric disorders have more diabetes complications and lower life expectancy than those with diabetes but no co-existing psychiatric disorders. Psychiatric health professionals may have a role in improving these outcomes but often lack diabetes knowledge and skills. This study aims to examine the effectiveness of a diabetes training course for psychiatric health professionals on their diabetes knowledge and skills and clinical outcomes, diabetes support and diabetes distress among individuals with diabetes and psychiatric disorders treated in psychiatric outpatient clinics.
Methods
A pragmatic non-randomized controlled cluster trial was conducted in eight psychiatric outpatient clinics in Denmark. All psychiatric health professionals from four clinics participated in the diabetes training course (the intervention) and completed a questionnaire on experience of the training course and a 20-item pre- and post-test to measure diabetes knowledge and skills. Difference in pre- and post-tests were analyzed using t-tests. From August 2018 – June 2019, individuals with diabetes were recruited from the intervention clinics (n = 49) and from four control clinics continuing usual clinical practice (n = 57). Differences in clinical outcomes, diabetes support and diabetes distress between the intervention and control groups at six and 12 months after the training course, were analyzed using logistic and linear regression models adjusted for baseline levels.
Results
Psychiatric health professionals (n = 64) had more correct answers after completing the course, with a mean increase of 6.3 [95% CI 5.6 to 7.0] correct answers. A total of 49 and 57 individuals were recruited for the intervention and control group, respectively. At follow-up, individuals treated in the intervention group had lower levels (clinical improvement) of systolic blood pressure, but had lower receipt of annual assessment of blood pressure, and body mass index (BMI) (worsening of process measures). While there were observed differences in odds and means for several other outcomes, none of these received statistical significance (see Table 2 and Fig. 2).
Conclusions
Training psychiatric health professionals in diabetes care improved their diabetes knowledge and skills and improved clinical levels of systolic blood pressure in individuals treated in the intervention group. However, this training intervention was associated with a lower likelihood of receiving annual assessment of blood pressure and BMI.
Trial registration
ISRCTN registry registration number ISRCTN15523920, registration date: 02/10/2019.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12909-024-06288-z.
Keywords: Diabetes education, Psychiatric health professionals, Effectiveness, Patient outcomes
Background
Life expectancy for individuals with psychiatric disorders is 10–15 years shorter than that of the general population [1–3]. The most common causes of death for individuals with psychiatric disorders are cardiovascular disease, type 2 diabetes, chronic obstructive lung disease, and suicide [1–3]. Diabetes is two to three times more common among individuals with psychiatric disorders than in the general population [4], due in part to weight gain arising from the use of psychotropic drugs and unhealthy lifestyle behaviors and inadequate diabetes management [5]. Among individuals with diabetes, those with comorbid psychiatric disorders have a higher incidence of diabetes-related complications and higher mortality than those with no psychiatric disorders [2, 6, 7]. Potential explanations for the disparities in outcomes include a primary focus on psychiatric disorders among both health professionals [5] and individuals living with diabetes and psychiatric disorders [5, 8], lower quality of diabetes care [9, 10], and difficulty with daily diabetes self-management [8, 11] and keeping diabetes care appointments [12].
Previous lifestyle and individually focused interventions aiming to improve diabetes care or self-management have had a limited effect on hemoglobin A1c (HBA1c) and lipid levels and diabetes management in this vulnerable population [13, 14].
In Denmark, general practitioners or endocrinologists are responsible for the diabetes treatment. Nevertheless, individuals with co-existing psychiatric disorders have lower quality of diabetes care and often more frequent contact with psychiatric health professionals. Previous studies have suggested that psychiatric health professionals may play a key role in improving mental- and physical health problems among individuals with diabetes and psychiatric disorders [5, 15, 16]. For example, annual screening for metabolic syndrome in individuals treated with antipsychotic medication in psychiatric clinics is recommended.
Research has identified the lack of knowledge about diabetes and diabetes support skills among psychiatric health professionals as a barrier to improving diabetes outcomes [16]. Psychiatric health professionals have expressed a need for diabetes training to improve care for individuals with diabetes and psychiatric disorders [17, 18]. To increase psychiatric health professionals’ knowledge about diabetes and related care for individuals with diabetes and psychiatric disorders, we developed a three-day diabetes training course. The study aims were to: 1) assess psychiatric health professionals’ need for diabetes knowledge, experiences of the training, and diabetes knowledge before and after attending it and 2) evaluate the effect of the training on clinical outcomes related to diabetes, experiences of support from psychiatric health professionals, and diabetes distress among individuals with diabetes and psychiatric disorders treated in psychiatric outpatient clinics.
Methods
Study design and population
We conducted a pragmatic non-randomized controlled cluster trial to examine the real-world effectiveness of the intervention [19], ISRCTN registry (ISRCTN15523920) registered 02/10/2019. All fifteen psychiatric out-patient clinics in the Capital Region of Denmark were invited to participate. Six clinics declined due to contemporaneous participation in other trials. Clinics were allocated to the intervention or control group based on their ability to allow three days off for all health professionals to participate in the training course. One clinic elected to participate was later excluded because all clinic health professionals could not participate in the training. Thus, the intervention and control groups each comprised four clinics.
The staff at the psychiatric out-patient clinics included psychiatric nurses, psychologists, psychiatrists, occupational therapists, health care assistants, social workers, health coordinators, health care assistants, and clinical supervisors.
The intervention
The intervention consisted of a tailored diabetes training course provided over three consecutive days in February-June and December 2019. To tailor the content to the expressed needs of psychiatric health professionals, the course instructors performed a needs assessment that included interviews with psychiatric health professionals from each clinic and participant observations of clinical visits between psychiatric health professionals and individuals with diabetes and psychiatric disorder. The course instructors included an endocrinologist, two diabetes nurses, a dietitian, and a podiatrist, all with between ten to twenty-five years of clinical experience in diabetes treatment and training of health professionals. Between 2–4 participant observations were conducted in each clinic.
The course instructors identified the learning components and objectives of the training course based on the needs assessment. The objective of the training course was to increase psychiatric health professionals’ knowledge and skills related to: 1) assessing diabetes and the risk of diabetes, 2) monitoring and acting on physical, emotional, or social changes that could impact diabetes care, and 3) supporting and following up on diabetes care (e.g., annual assessments of HbA1c or foot screening) among individuals with diabetes treated in psychiatric outpatient clinics.
The training included ten learning components: type 1 and type 2 diabetes, medical treatment of type 1 and type 2 diabetes, blood glucose measurements, acute complications, late complications, the impact of other illnesses and pregnancy on blood sugar and diabetes management, links between psychiatric disorder and diabetes, alcohol and drugs, diet and physical activity, and organization of diabetes care in Denmark (Additional file 1). Pedagogical principles, learning objectives, materials and educational principles for each learning component are further described in additional file 1. The training course was delivered by instructors specialized in diabetes training for health professionals: an endocrinologist, two diabetes specialist nurses, a dietitian, and a podiatrist. Psychiatric health professionals with primary patient contact in the four clinics in the intervention group attended the training.
The practical application of the course included face-to-face education in plenum followed by small group and individual reflection to enhance the learning process. As an example, training in blood glucose measurement included a lecture on information about how and why blood glucose is measured. All participants were given a graph with an example of fluctuations in blood glucose levels to discuss possible causes of the fluctuations and how to address them. This was followed up in plenum. Afterwards, each participant was instructed on how to measure glucose with a blood glucose device first shown by the instructors. The participants were then encouraged to perform the measurement on themselves in the class with help from the instructors if needed and accordingly the next days by themselves. This allowed for practical training.
Setting and recruiting individuals with diabetes treated in psychiatric outpatient clinics
In Denmark around 80% of all individuals with diabetes are treated in general practice and 20% are treated by diabetes specialists at hospitals in case of type 1 diabetes or complicated type 2 diabetes. While, psychiatric disorders are treated in general practice, out- or in-patient clinics depending on the severity of the psychiatric symptoms.
The eight psychiatric out-patient clinics were all governed by the Capital Region of Copenhagen. Each clinic provides treatment for any psychiatric disorders with severe psychiatric symptoms. The eight clinics are responsible for patients according to geographical areas.
Individuals aged ≥ 18 years with a psychiatric disorder and diabetes (type 1, type 2, or unspecified) were recruited from each psychiatric out-patient clinic in August 2018-June 2019 through psychiatric health professionals as previously described [11]. The psychiatric health professionals were asked to screen all their patients for study eligibility based on one or more of the following inclusion criteria: 1) any diabetes diagnosis, 2) treatment with any diabetes medication, or 3) two or more HbA1c measurements ≥ 48 mmol/mol. They provided a verbal and written invitation about the study, and eligible participants gave verbal and written consent.
Characteristics of individuals with diabetes treated in the intervention and control group were measured at baseline and included age, sex, and diabetes, and psychiatric diagnoses from hospital electronic records. Items assessing diabetes duration, self-reported diabetes treatment provider, level of education, and smoking status were included on baseline participant questionnaires.
Outcome measures
Psychiatric health professional outcomes
A 20-item questionnaire was developed to assess participating health professionals’ experience of the training course and need for more knowledge after completing it (Additional file 2). It consisted of written statements such as “I had sufficient knowledge about diabetes care before taking the course,” “I was interested in diabetes care and treatment before taking the course,” and “the course content was relevant to my work as a psychiatric health professional.” Response options for each item were “strongly agree,” “somewhat agree,” “neither agree nor disagree,” “somewhat disagree,” “strongly disagree,” and “don’t know.”
A 20-item multiple-choice pre- and post-test was developed to measure psychiatric health professionals’ diabetes knowledge in five subthemes covering all ten learning components (Additional file 3). Each item had five response options and one correct answer. All participants were asked to complete the test before the training course began on the first day and after it ended on the last day.
Outcomes on individuals treated in psychiatric outpatient clinics
The effect of the training course on individuals’ clinical outcomes, experience of support from psychiatric health professionals, and diabetes distress was measured at baseline and at six and 12 months after psychiatric health professionals completed the training course (Fig. 1).
Fig. 1.
Flow chart of individuals with diabetes treated in the intervention and control group
Clinical outcomes included the proportion of individuals with annual assessments in the last 12 months and values of the most recent clinical measurements. Annual assessments were defined according to Danish guidelines for diabetes care [20–22] as the proportion of individuals with annual clinical assessments, including self-reported foot examinations. Clinical measurements included HbA1c, body mass index (BMI), blood pressure, Urine albumin creatinine ratio and low-density lipoprotein (LDL) cholesterol and were obtained from electronic medical records.
Participants’ experience of general diabetes support from psychiatric health professionals was measured with one item inspired by the validated DAWN Support for Diabetes Self-Management Profile [23]. Response options to the question “How much support do you get to manage your diabetes from the psychiatric health professional?” included “Don’t have any,” “Not supportive,” “Somewhat supportive,” and “Very supportive.” Responses were dichotomized into high support (very supportive) vs. no to some support (all other responses).
Diabetes-specific support was also measured by three items from the Patient Assessment of Chronic Illness Care (PACIC) [24]: “Helped to plan ahead so I could take care of my diabetes even in hard times” (item 14), “Contacted after a visit to see how things were going” (item 16), and “Referred to a dietitian or another health care professional who can support my diabetes” (item 18). Response options for each item were “None of the time,” “A little of the time,” “Some of the time,” “Most of the time,” “Always,” or “Not relevant.”
Diabetes distress was measured with the five-item Problem Areas in Diabetes Scale (PAID-5) [25, 26]. A questionnaire comprising all items assessing diabetes support and diabetes distress was distributed and collected by psychiatric health professionals as part of routine care at baseline and six and 12 months after the training. Participants completed the questionnaires independently or, as needed, with help from relatives or psychiatric health professionals.
Data security
All questionnaire data were anonymized and entered manually into REDCap using double data entry by AJ and LK [27, 28]. A data entry specialist blinded to group allocation collected clinical data from participants’ electronic medical records and entered them into REDCap.
Statistical analysis
Psychiatric health professional outcomes
Descriptive statistics were used to summarize information on psychiatric health professionals’ experiences of the training course and need for diabetes knowledge. Pre- and post-test differences were visualized for each psychiatric health professional in Spaghetti plots (Additional file 4) and pre-and post-test differences for all psychiatric health professionals were analysed with t-tests.
Outcomes on individuals treated in psychiatric outpatient clinics
Differences between participants dropping out of the study in the intervention and control groups were analysed using descriptive statistics, t-tests and chi-square tests (Additional file 5).
Participant characteristics were analyzed as proportions for categorical and means with standard deviations (SD) or medians with interquartile ranges (IQR) for non-normal distributed continuous variables. Differences in baseline characteristics between the intervention and control groups were analysed using chi-square tests and t-tests. Spaghetti plots [29] for all clinical measures were used to visualize individual differences at baseline and after six and 12 months grouped by intervention and control group (Additional file 6). As per requirement, we added a small amount of random noise to the data used in the Spaghetti plots to ensure full anonymity. The random noise was performed by adding small uniformly distributed jittering to the values. The jittering change was found by finding appropriate numbers for change without changing the results.
Between-group differences in clinical outcomes and diabetes distress were analysed using logistic regression for binary outcomes (presence or absence of annual assessments of HbA1c, BMI, blood pressure, Urine albumin creatinine ratio, LDL cholesterol, annual foot screening, and high diabetes support) and linear regression models for continuous outcomes (HbA1c, BMI, blood pressure, Urine albumin creatinine ratio, LDL cholesterol, diabetes distress). All models were adjusted for baseline levels. Differences in ordinal outcomes were visualized in Sankey plots and analysed with the Mantel–Haenszel test (Additional file 7). To examine the potential impact of missing data on our findings, we performed both complete case analysis and multiple imputation analysis when estimating between-group differences for HbA1c and diabetes distress (Additional file 8). All analyses were conducted using the statistical software R version 3.6.0 (R Foundation for Statistical Computing).
Ethics
The Danish Data Protection Agency approved the trial P-2019–338. According to the Danish Act on Research Ethics Review of Health Research Projects, the Danish Committee on Health Research Ethics evaluated the trial should not be registered (H-18009189).
Results
Psychiatric health professional outcomes
A total of 66 psychiatric health professionals from four psychiatric outpatient clinics participated in the training course. They included psychiatric nurses (n = 51), psychologists (n = 3), occupational therapists (n = 3), health care assistants (n = 3) and others (n = 6), including social workers, health coordinators, health care assistants and clinical supervisors. Of these, two participants were unable to complete the entire course due to illness, leaving 64 psychiatric health professionals in analyses.
Among psychiatric health professionals participating in the training, 43 (69%) highly or somewhat disagreed and 11 (18%) highly agreed or agreed with the statement, “I had sufficient knowledge about diabetes care and treatment before taking the course.” In response to the statement, “The course content was relevant to my work as a psychiatric health professional,” 44 (69%) strongly agreed and 19 (30%) somewhat agreed. All questionnaire results are available in Additional file 2.
Diabetes knowledge among psychiatric health professionals increased by a mean 6.3 items (95% confidence interval [CI] 5.6 to 7.0) out of 20 questions. Of 20 questions, the mean number of correct responses before the training was 9 (standard deviation [SD] 3), increasing to 16 (SD 2) after the training. Scores in diabetes knowledge increased significantly in all subtheme scores (Additional file 4).
Outcomes on individuals treated in psychiatric outpatient clinics
A total of 106 individuals with diabetes and psychiatric disorders were recruited; 49 and 57 individuals were treated in the intervention and control groups, respectively. The follow-up rate in the intervention group decreased to 80% at six months and 69% at 12 months, while the follow-up rate in the control group was 84% at both time intervals (Fig. 1). We found no differences in age, sex, diabetes duration, or other baseline characteristics between individuals who were lost to follow-up and those who completed the study (Additional file 5). More than 75% of all individuals were diagnosed with type 2 diabetes, and the most frequent psychiatric diagnoses were schizophrenia and other psychotic disorders (Table 1). The primary reason for study dropout was discharge from the psychiatric clinic during follow-up (Fig. 1). Missing data analyses are presented in Additional file 8.
Table 1.
Baseline characteristics of individuals with diabetes and psychiatric disorder treated in the intervention and control group, n = 106
| Intervention group (n = 49) | Control group (n = 57) | |
|---|---|---|
| Age in years, mean (SD) | 50.5 (12.2) | 52.5 (11.6) |
| Men, n (%) | 27 (54.0) | 33 (57.9) |
| Diabetes diagnosis, n (%) | ||
| Type 1 | 5 (10.0) | 7 (12.3) |
| Type 2 | 40 (80.0) | 43 (75.4) |
| Unspecified or other | 1 (2.0) | 6 (10.5) |
| Diabetes duration in years, mean (SD) | 10.6 (9.8) | 12.6 (10.0) |
| Missing responses, n (%) | 17 (34.7) | 18 (31.6) |
| Diabetes treatment provider, n (%) | ||
| General practice | 26 (53.1) | 22 (38.6) |
| Endocrinologist at hospital | 10 (20.4) | 22 (38.6) |
| None | 1 (2.0) | 3 (5.3) |
| Missing responses | 12 (24.5) | 10 (17.5) |
| Psychiatric diagnoses, n (%) | ||
| Schizophrenia and other psychotic disorders | 25 (50.0) | 37 (64.9) |
| Affective mental disorders | 13 (26.0) | 10 (17.5) |
| Nervous and stress-related disorders | 7 (14.0) | 8 (14.0) |
| Psychiatric disorders due to psychoactive drugs | 10 (20.0) | 3 (5.3) |
| Personality and behavioral disorders | 6 (12.0) | 3 (5.3) |
| Other psychiatric disorders | 10 (20.0) a | 5 (8.8) a |
| Clinical outcomes | ||
| HbA1c, %, mean (SD) | 7.1 (1.5) | 7.7 (1.7) |
| HbA1c, mmol/mol, mean (SD) | 53.4 (15.8) | 60.3 (18.8) |
| Body mass index, kg/m2, mean (SD) | 32.3 (6.4) | 31.5 (7.4) |
| Missing responses, n (%) | 18 (36.7) | 23 (40.4) |
| Systolic blood pressure, mmHg, mean (SD) | 131 (14) | 131 (19) |
| Diastolic blood pressure, mmHg, mean (SD) | 84 (13) | 82 (10) |
| Missing responses, n (%) | 7 (14.3) | 2 (3.5) |
| Urine albumin creatinine ratio, median (IQR) | 9.0 (5.0, 20.8) | 13.5 (6.3, 38.3) |
| Missing responses, n (%) | 23 (46.9) | 19 (33.3) |
| LDL cholesterol, mmol/l, mean (SD) | 2.2 (0.9) | 2.0 (0.9) |
| Missing responses, n (%) | 3 (6.1) | 1 (1.8) |
| Receiving annual assessment, n (%) | ||
| HbA1c | 48 (98.0) | 56 (98.2) |
| Body mass index | 28 (57.1) | 51 (89.5) |
| Blood pressure | 32 (65.3) | 53 (93.0) |
| Urine albumin creatinine ratio | 20 (40.8) | 25 (43.9) |
| Low-density lipoprotein | 39 (79.6) | 46 (80.7) |
| Foot screening, self-reported | 28 (57.1) | 37 (64.9) |
| Missing responses | 12 (24.5) | 9 (15.8) |
| Blood glucose control, n (%) | ||
| HbA1c ≤ 53 mmol/mol | 29 (59.2) | 24 (42.1) |
| HbA1c ≥ 70 mmol/mol | 7 (14.3) | 14 (24.6) |
| Reported high diabetes support from psychiatric health professionalsa, n (%) | 8 (25.8) | 7 (15.2) |
| Diabetes distress (PAID-5) score, mean (SD) | 7.8 (5.3) | 7.7 (5.0) |
| Highest level of education, n (%) | ||
| Low | 14 (28.6) | 15 (26.3) |
| Medium | 12 (24.5) | 16 (28.1) |
| High | 10 (20.4) | 16 (28.1) |
| Missing responses | 13 (26.5) | 10 (17.5) |
Abbreviations: HbA1c hemoglobin A1C, LDL cholesterol = low density-lipoprotein cholesterol
aStatistically significant between-group difference (p < 0.05)
Clinical outcomes
Spaghetti plots of all clinical measures are presented in Additional File 6 and display individual differences over time for individuals in the intervention and control groups. At baseline, 98% of participants in the intervention and control groups had an annual HbA1c assessment (Table 2).
Table 2.
Differences in the intervention and control group at six and 12 months, adjusted for baseline levelsa
| OR (95% CI) | ||
|---|---|---|
| 6 months | 12 months | |
| Received annual assessment, n (%) | ||
| HbA1c | 0.5 (0.1, 2.9) | |
| Body mass index | 0.2 (0.1, 0.7) | |
| Blood pressure | 0.2 (0.1, 0.7) | |
| Urine albumin creatinine ratio | 1.2 (0.5, 2.9) | |
| LDL cholesterol | 0.7 (0.2, 2.0) | |
| Foot screening, self-reported, | 1.0 (0.2, 6.1) | |
| Blood glucose control | ||
| HbA1c ≤ 53 mmol/mol | 2.3 (0.9, 5.9) | 1.5 (0.5, 4.4) |
| HbA1c ≥ 70 mmol/mol | 0.4 (0.1, 1.7) | 0.4 (0.1, 1.8) |
| Level of diabetes support from psychiatric health professionals | ||
| High vs. none or low, n (%)a | 0.6 (0.1, 2.4) | 0.2 (0.0, 2.3) |
| Diabetes distress, mean between-group difference (96% CI)a | 0.0 (-2.0, 1.9) | -2.0 (-4.3, 0.3) |
Abbreviations: HbA1c hemoglobin A1C, LDL cholesterol = low density-lipoprotein cholesterol
aMissing data (n; intervention group, control group):
Diabetes support: baseline (6, 3); 6 months (2, 0); 12 months (2, 2)
Diabetes distress: baseline (12, 9); 6 months (13, 8); 12 months (12, 8)
At 12 months, individuals in the intervention group had lower odds of receiving annual assessment for BMI and blood pressure at 12 months, with 80% (OR 0.2, 95% CI 0.1 to 0.7) lower odds of receiving an assessment for either measure in the intervention group than in the control group.
We also found lower levels of systolic blood pressure at 12-months follow-up in the intervention compared to the control group.
While there were observed differences in odds and means for several other outcomes, none of these received statistical significance (see Table 2 and Fig. 2).
Fig. 2.
Mean 95% CI difference in clinical measures between individuals treated in the intervention and control group at 6 months and 12 months follow-up adjusted for baseline levels. Foot note: The x-axis shows the mean (95% CI) difference in units specified for each measure. Units are shown on the y-axis. HbA1c measures are presented in % and mmol/mol. Abbreviations: HbA1c, hemoglobin A1C, LDL cholesterol = low density-lipoprotein cholesterol
Experience of support from psychiatric health professionals
We found a difference in the experience of support between baseline and 6-months follow-up, in the intervention group compared to the control group. This was found for the items “helped to plan ahead so I could take care of my diabetes even in hard times” and “Referred to a dietitian or other health professional, who can support my diabetes”. However, we find no clear pattern of worse or better support at follow-up in the individual responses in the intervention and control group (Additional file 7).
Diabetes distress
We found a tendency towards a lower mean level of diabetes distress of -2.0 [-4.3; 0.3] on the PACIC-score from 0–20, at 12 months in the intervention compared to the control group, but it did not reach statistical significance (Table 2).
Discussion
This study is the first to examine the effect of tailored diabetes training of psychiatric health professionals on diabetes-related clinical outcomes, experience of diabetes support from psychiatric health professionals, and diabetes distress among individuals with diabetes and psychiatric disorders.
Psychiatric health professionals found the course relevant to their clinical practice and indicated a lack of sufficient diabetes knowledge of before attending it. After the training course, the psychiatric health professionals had statistically significantly improved their diabetes-related knowledge and skills.
We found lower odds of annual assessment of BMI and blood pressure. Simultaneously, we found lower mean of systolic blood pressure in the intervention group at 12 months and no other statistically significant difference in clinical measures but a consistent tendency to lower mean levels of all clinical measures at six and 12 months in the intervention group, compared to the control group, and a lower mean level of diabetes distress at 12 months, although neither of these findings reached statistical significance.
Our results contrast with previous lifestyle and individually focused interventions that have found limited effect on HbA1c and lipid levels and diabetes management [13, 14]. This could be due to our intervention’s focus on psychiatric health professionals, rather than individuals. Research has suggested that psychiatric health professionals could play a key role in improving physical health problems in individuals with psychiatric disorders [5, 15, 16]. However, despite the significant increase in psychiatric health professionals’ diabetes knowledge, we found lower odds of receiving annual assessment of clinical outcomes and tendencies toward improved clinical outcomes.
One explanation could be our very complex and heterogeneous group of individuals with several comorbidities in addition to diabetes and psychiatric disorders. We also experienced a high drop-out rate at 12 months, primarily due to discharge from the psychiatric clinic as a result of improved psychiatric symptoms. In Denmark, individuals treated in outpatient clinics comprise a group with severe psychiatric symptoms. When psychiatric symptoms improve, patients are discharged and further treatment is in general practice. This means that our results are disproportionately based on participants with severe psychiatric symptoms, for whom it may be more difficult to improve diabetes outcomes within 12 months. A previous study found that diabetes care and management is often under-prioritized during periods of severe psychiatric symptoms by both psychiatric health professionals and patients with both conditions [8]. On the other hand, interventions targeting individuals treated at psychiatric inpatient clinics were more effective than those targeting individuals treated at outpatient clinics [14]. A potential explanation is that individuals in inpatient clinics have more support from psychiatric health professionals throughout the day than those in outpatient clinics.
Our lack of statistically significant findings related to clinical measures could be due to the fact that we were unable to recruit the number of participants indicated in sample size calculations. Another explanation could be that, at baseline, most individuals treated in the intervention and control group had both well-regulated clinical measures and annual assessments, leaving little room for improvement. However, it is also possible that the statistically significant increase in psychiatric health professionals’ diabetes knowledge and skills was not translated into clinical practice. This could also explain why we found paradoxically lower annual assessments and levels of clinical measures and diabetes support in the intervention group. The tendency toward improved levels of clinical measures may be related to less assessment at follow-up in the intervention compared to the control group. A decreasing proportion receiving care could result in improved clinical measures as it is likely that the lower proportion of individuals receiving annual assessments in the intervention group comprised a select group whose diabetes was in better control.
The Danish health care system has increasingly focused on diabetes in psychiatric care over the past years, and psychiatric health professionals are encouraged to ensure annual assessments of diabetes-related clinical measures, which may explain the well-regulated baseline levels we observed. However, it could also be possible that our measurements of diabetes support may not have captured other areas of improved support or contact between our participants with diabetes and health professionals.
Methodological considerations
A strength was the combination of clinical and psychosocial outcomes in this hard-to-reach population. In addition, although we had high drop-out rates, we retained most study participants at six and 12 months in both the intervention and control groups, even though this vulnerable population is notoriously hard to reach. We obtained clinical data for all participants. We found no obvious differences between participants who completed the study and those who dropped out on the baseline characteristics, indicating a lack of bias due to missing responses.
A limitation of our study was that it was not possible to conduct it as a randomized trial because four participating clinics could not provide three days off from clinical practice for all psychiatric health professionals to participate in the intervention. Consequently, clinics allocated themselves to the intervention and control groups, potentially introducing bias in the form of greater organizational willingness to support diabetes training in the intervention group, enhancing the effect of the intervention. The non-randomization may also have led to selection bias related to geographical uptake of patients or difference in the clinical staff or work conditions.
We may also have introduced geographical bias, as the four clinics in the control group were all located in the Capital city.
Nevertheless, we found no differences in baseline characteristics between participants treated in the intervention and control groups.
Another study limitation was that recruitment of participants was very difficult, limiting the statistical power of our findings and increasing the risk of type II error. Additionally, we recruited participants through psychiatric health professionals, which may have introduced selection bias because they invited only the patients who they knew had diabetes. It is also possible that the study population consisted of their highest functioning patients with the resources for participating in a trial.
Possible intervention effects due to changes in diabetes treatment related to general practitioners or diabetes specialists were not within the scope of this study. However, diabetes treatment providers in intervention- and control group were obliged to follow the same national treatment guidelines, which should limit the possibility of intervention effects due to this.
This study did not include a pre- and post-test of diabetes knowledge in the control group. This means that we could not assess a possible learning effect independent of the training course with a difference-in-difference analysis. Additionally, it was beyond the scope of this study to examine the training’s impact on long-term competency. Future studies should examine the impact of training on health professionals’ competencies, for example, 12 or 24 month after training, in a randomized intervention and control group.
Future training should also incorporate a greater focus on knowledge transfer to enhance its impact on patient clinical outcomes. This could be achieved through more scenario-based learning or simulations that mirror real world clinical situations, as well as incorporating continuous feedback loops and post-training follow-up sessions.
Our analyses indicated that missing data were not randomly distributed, which could both be due to improvement or worsening of the psychiatric disorder or other factors. Thus, our analyses may underestimate or overestimate the effect of the intervention. However, when we repeated our analyses for levels of HbA1c and diabetes distress with multiple imputations of missing values, the pattern of results was similar (Supplementary Table 5), although the effect estimates were smaller.
Conclusion
Psychiatric health professionals expressed a need for more diabetes knowledge and skills, which were improved after attending a three-day training course. The effectiveness of the training course among psychiatric health professionals had diverse effects on individuals with diabetes and psychiatric disorders. We found improved levels of systolic blood pressure, but fewer annual assessments of BMI and blood pressure. This may be due to high quality care at baseline, leaving little room for improvement, and higher drop-out rates in the intervention group due to being discharged from psychiatric care. Future studies must include randomized design, larger sample sizes to examine the effectiveness of similar interventions and measure the implementation and transfer of knowledge and skills into clinical practice and the effect on health care professionals’ long-term competencies should be further examined.
Supplementary Information
Additional file 1. Training course components, learning objectives and education methods for the diabetes training course. Contains a detailed overview of each session of the three-day training.
Additional file 2. Psychiatric health professionals’ experience of the training course and need for more knowledge. Depicts all items and responses on a 20-item questionnaire.
Additional file 3. Pre- and post-test of diabetes knowledge among psychiatric health professionals. All items and response options on a 20-item pre- and post-test covering five domains of knowledge about diabetes and diabetes treatment.
Additional file 4. Changes in diabetes knowledge and skills among all participating psychiatric health professionals (n = 64). Means, standard deviations, and difference with 95% confidence intervals for psychiatric health professionals’ scores on the pre- and post-test. Spaghetti plot of pre/post changes in diabetes knowledge and skills among individual participating psychiatric health professionals.
Additional file 5. Baseline characteristics of participating individuals with diabetes treated in the intervention and control group who were lost to follow-up at any time and those completing the study. Compares completers and non-completers on 15 variables and identifies statistically significant differences only in some psychiatric diagnoses and questionnaire completion.
Additional file 6. Spaghetti plots of clinical measures at baseline, six, and 12 months for each individual with diabetes treated in psychiatric outpatient clinics grouped by intervention and control group. Provides data visualization for each individual with diabetes treated in the intervention and control groups for levels of clinical measures.
Additional file 7. Difference at six and 12 months in individuals with diabetes on specific diabetes support from psychiatric health professionals. Sankey plots providing data visualization of each individual’s responses to two items assessing the level of diabetes support provided by psychiatric health professionals.
Additional file 8. Models with and without multiple imputation for missing data for individuals with diabetes treated in the intervention and control group. Provides comparative results for models that do and do not account for missing data for HbA1c and diabetes distress
Acknowledgements
This study would not have been possible without the great contribution from all the participants with diabetes and psychiatric disorders and the psychiatric health professionals from the included psychiatric out-patient clinics in Copenhagen Denmark, who made a great effort in recruitment and handing out questionnaires.
Abbreviations
- HbA1c
Hemoglobin A1c
- BMI
Body mass index
- LDL -cholesterol
Low-density lipoprotein cholesterol
- PACIC
Patient Assessment of Chronic Illness Care
- PAID-5
Five item Problem Areas in Diabetes Scale
- SD
Standard deviation
- IQR
Interquartile ranges
Authors’ contributions
LK, GSA, LEJ, LLN, AJ, KL, MEJ and DLH led the conception, design and planning of the study. LK led data management with support from LJD and KKBC. LK led the analyses with support from GSA and LJD. LK led drafting of the work. All authors contributed to the interpretation of the data and revising it critically for important intellectual content and read and approved the final manuscript. LK is responsible for the overall content of the manuscript as guarantor. LK, GSA, LJD, KKBC and DLH had access to the data and LK, GSA, MEJ and DLH controlled the decision to publish.
Funding
This study has been funded by Copenhagen University Hospital—Steno Diabetes Center Copenhagen through an unrestricted grant from Novo Nordisk Foundation and Jascha Foundation; case number 6994.
Data availability
The datasets used and/or analysed during the current study available from the corresponding author Lenette Knudsen, on reasonable request.
Declarations
Ethics approval and consent to participate
This study was conducted according to the guidelines in the Declaration of Helsinki. The study was approved by the Danish Data Protection Agency P-2019–338. The Danish Committee on Health Research Ethics assessed that ethical approval was not required for this study according to the Danish Act on Research Ethics Review of Health Research Projects (H-18009189). Interested and eligible individuals with diabetes and psychiatric disorder were given written and verbal information about the study before giving oral and written informed consent to participate.
Consent for publication
Not applicable.
Competing interests
LEJ, LLN, AJ, MAN, KL, and DLH have no conflicts of interest to declare. LK, GSA and KKBC owns shares in Novo Nordisk A/S. LJD, KKBC and GSA are currently employed by Novo Nordisk A/S. During contribution to the manuscript, LJD, KKBC, and GSA were employed by Steno Diabetes Center Copenhagen. MEJ holds shares in Novo Nordisk, and has received research grants from AMGEN, Astra Zeneca, Boehringer Ingelheim, Novo Nordisk, and Sanofi Aventis.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Laursen TM. Life expectancy among persons with schizophrenia or bipolar affective disorder. Schizophr Res. 2011;131(1–3):101–4. [DOI] [PubMed] [Google Scholar]
- 2.Ribe AR, Laursen TM, Sandbaek A, Charles M, Nordentoft M, Vestergaard M. Long-term mortality of persons with severe mental illness and diabetes: a population-based cohort study in Denmark. Psychol Med. 2014;44(14):3097–107. [DOI] [PubMed] [Google Scholar]
- 3.Wahlbeck K, Westman J, Nordentoft M, Gissler M, Laursen TM. Outcomes of Nordic mental health systems: life expectancy of patients with mental disorders. The British journal of psychiatry : the journal of mental science. 2011;199(6):453–8. [DOI] [PubMed] [Google Scholar]
- 4.Lindekilde N, Scheuer S, Rutters F, Knudsen L, Lasgaard M, Rubin K, et al. Prevalence of type 2 diabetes in psychiatric disorders: an umbrella review with meta-analysis of 245 observational studies from 32 systematic reviews. Diabetologia. 2022(65(3)):440–456. [DOI] [PubMed]
- 5.De Hert M, Cohen D, Bobes J, Cetkovich-Bakmas M, Leucht S, Ndetei DM, et al. Physical illness in patients with severe mental disorders. II. Barriers to care, monitoring and treatment guidelines, plus recommendations at the system and individual level. World psychiatry : official journal of the World Psychiatric Association (WPA). 2011; 10(2):138–151. [DOI] [PMC free article] [PubMed]
- 6.Toender A, Vestergaard M, Munk-Olsen T, Larsen JT, Kristensen JK, Laursen TM. Risk of diabetic complications and subsequent mortality among individuals with schizophrenia and diabetes - a population-based register study. Schizophr Res. 2020;218:99–106. [DOI] [PubMed] [Google Scholar]
- 7.Scheuer SH, Kosjerina V, Lindekilde N, Pouwer F, Carstensen B, Jørgensen ME, et al. Severe mental illness as a risk factor for incident micro- and macrovascular complications in persons with type 2 diabetes. Journal of Endocrinology and Metabolism. 2022; (in press).
- 8.Stenov V, Joensen LE, Knudsen L, Lindqvist Hansen D, Willaing TI. "Mental health professionals have never mentioned my diabetes, they don’t get into that”: A qualitative study of support needs in adults with type 1 and type 2 diabetes and severe mental illness. Can J Diabetes. 2020;44(6):494–500. [DOI] [PubMed] [Google Scholar]
- 9.Jorgensen M, Mainz J, Carinci F, Thomsen RW, Johnsen SP. Quality and predictors of diabetes care among patients with schizophrenia: A Danish nationwide study. Psychiatr Serv. 2018;69(2):179–85. [DOI] [PubMed] [Google Scholar]
- 10.Knudsen L, Scheuer SH, Diaz LJ, Jackson CA, Wild SH, Benros ME, et al. Indicators of quality of diabetes care in persons with type 2 diabetes with and without severe mental illness: a Danish nationwide register-based cohort study. 2022(2666–7762 (Electronic)). [DOI] [PMC free article] [PubMed]
- 11.Knudsen L, Hansen D, Joensen LE, Wibaek R, Benros M, Jørgensen M, et al. Need for improved diabetes support among people with psychiatric disorders and diabetes treated in psychiatric outpatient clinics: results from a Danish cross-sectional study. BMJ Open Diabetes Res Care. 2022;10(1): e002366. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Nasrallah HA, Meyer JM, Goff DC, McEvoy JP, Davis SM, Stroup TS, et al. Low rates of treatment for hypertension, dyslipidemia and diabetes in schizophrenia: data from the CATIE schizophrenia trial sample at baseline. Schizophr Res. 86(1–3):15–22. [DOI] [PubMed]
- 13.McBain H, Mulligan K, Haddad M, Flood C, Jones J, Simpson A. Self management interventions for type 2 diabetes in adult people with severe mental illness. The Cochrane database of systematic reviews. 2016; 4:Cd011361. [DOI] [PMC free article] [PubMed]
- 14.Gron AO, Dalsgaard EM, Ribe AR, Seidu S, Mora G, Cebrian-Cuenca AM, et al. Improving diabetes care among patients with severe mental illness: A systematic review of the effect of interventions. Prim Care Diabetes. 2018;12(4):289–304. [DOI] [PubMed] [Google Scholar]
- 15.Whicher CA, Price HC, Holt RIG. Mechanisms in endocrinology: Antipsychotic medication and type 2 diabetes and impaired glucose regulation. Eur J Endocrinol. 2018;178(6):R245–58. [DOI] [PubMed] [Google Scholar]
- 16.Hennessy S, Cocoman AM. What Is the Impact of Targeted Health Education for Mental Health Nurses in the Provision of Physical Health Care? An Integrated Literature Review. Issues Ment Health Nurs. 2018;39(8):700–6. [DOI] [PubMed] [Google Scholar]
- 17.Nash M. Mental health nurses' diabetes care skills - a training needs analysis. British journal of nursing (Mark Allen Publishing). 2009; 18(10):626, 628–630. [DOI] [PubMed]
- 18.Howard L, Gamble C. Supporting mental health nurses to address the physical health needs of people with serious mental illness in acute inpatient care settings. J Psychiatr Ment Health Nurs. 2011;18(2):105–12. [DOI] [PubMed] [Google Scholar]
- 19.Loudon K, Treweek S, Sullivan F, Donnan P, Thorpe KE, Zwarenstein M. The PRECIS-2 tool: Designing trials that are fit for purpose. BMJ. 2015;350: h2147. [DOI] [PubMed] [Google Scholar]
- 20.Danish Endocrine Society. Treatment guidelines for type 1 diabetes. https://endocrinology.dk/nbv/diabetes-melitus/type-1-diabetes-mellitus/ (2019). Accessed 11 Nov 2021.
- 21.Danish Endocrine Society. Treatment guidelines for type 2 diabetes. https://endocrinology.dk/nbv/diabetes-melitus/behandling-og-kontrol-af-type-2-diabetes/ (2019). Accessed 1 Nov 2021.
- 22.The Danish Clinical Quality Program (RKKP). Danish diabetes database national annual report 2019/2020. https://www.rkkp.dk/siteassets/de-kliniske-kvalitetsdatabaser/arsrapporter/arsrapport_diabetes_2019_20_endelig.pdf (2020). Accessed 6 June 2022.
- 23.Nicolucci A, Kovacs Burns K, Holt RI, Comaschi M, Hermanns N, Ishii H, et al. Diabetes Attitudes, Wishes and Needs second study (DAWN2): cross-national benchmarking of diabetes-related psychosocial outcomes for people with diabetes. Diabetic medicine : a journal of the British Diabetic Association. 2013;30(7):767–77. [DOI] [PubMed] [Google Scholar]
- 24.Glasgow RE, Whitesides H, Nelson CC, King DK. Use of the Patient Assessment of Chronic Illness Care (PACIC) with diabetic patients: relationship to patient characteristics, receipt of care, and self-management. Diabetes Care. 2005;28(11):2655–61. [DOI] [PubMed] [Google Scholar]
- 25.Polonsky WH, Fisher L, Earles J, Dudl RJ, Lees J, Mullan J, et al. Assessing psychosocial distress in diabetes: development of the diabetes distress scale. Diabetes Care. 2005;28(3):626–31. [DOI] [PubMed] [Google Scholar]
- 26.McGuire BE, Morrison TG, Hermanns N, Skovlund S, Eldrup E, Gagliardino J, et al. Short-form measures of diabetes-related emotional distress: the Problem Areas in Diabetes Scale (PAID)-5 and PAID-1. Diabetologia. 2010;53(1):66–9. [DOI] [PubMed] [Google Scholar]
- 27.Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O’Neal L, et al. The REDCap consortium: Building an international community of software platform partners. J Biomed Inform. 2019;95: 103208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)–a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Wilkinson L: The Grammar of Graphics, Second Edition edn. Statistics and Computing: Springer; 2005.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Additional file 1. Training course components, learning objectives and education methods for the diabetes training course. Contains a detailed overview of each session of the three-day training.
Additional file 2. Psychiatric health professionals’ experience of the training course and need for more knowledge. Depicts all items and responses on a 20-item questionnaire.
Additional file 3. Pre- and post-test of diabetes knowledge among psychiatric health professionals. All items and response options on a 20-item pre- and post-test covering five domains of knowledge about diabetes and diabetes treatment.
Additional file 4. Changes in diabetes knowledge and skills among all participating psychiatric health professionals (n = 64). Means, standard deviations, and difference with 95% confidence intervals for psychiatric health professionals’ scores on the pre- and post-test. Spaghetti plot of pre/post changes in diabetes knowledge and skills among individual participating psychiatric health professionals.
Additional file 5. Baseline characteristics of participating individuals with diabetes treated in the intervention and control group who were lost to follow-up at any time and those completing the study. Compares completers and non-completers on 15 variables and identifies statistically significant differences only in some psychiatric diagnoses and questionnaire completion.
Additional file 6. Spaghetti plots of clinical measures at baseline, six, and 12 months for each individual with diabetes treated in psychiatric outpatient clinics grouped by intervention and control group. Provides data visualization for each individual with diabetes treated in the intervention and control groups for levels of clinical measures.
Additional file 7. Difference at six and 12 months in individuals with diabetes on specific diabetes support from psychiatric health professionals. Sankey plots providing data visualization of each individual’s responses to two items assessing the level of diabetes support provided by psychiatric health professionals.
Additional file 8. Models with and without multiple imputation for missing data for individuals with diabetes treated in the intervention and control group. Provides comparative results for models that do and do not account for missing data for HbA1c and diabetes distress
Data Availability Statement
The datasets used and/or analysed during the current study available from the corresponding author Lenette Knudsen, on reasonable request.