Background
In the Netherlands, 66 in 1000 persons were diagnosed with type 2 diabetes mellitus in 2017 [
1]. This rate is expected to increase to 80 in 1000 persons by 2025 [
2]. There will be no comparable increase of health care providers, which compromises the availability of face-to-face time between patients and health care providers. In the Dutch health system, 95% of patients with type 2 diabetes are treated in primary care by a general practitioner (GP) and a practice nurse (PN) [
3]. To lower the risk of cardiovascular complications and manage the increasing number of type 2 diabetes patients, encouraging self-management could be part of the solution. Self-management can be defined as the active participation of patients in their treatment [
4] to minimize the impact of type 2 diabetes on their physical health and functioning and help them to cope with the psychological effects of their illness [
5]. In the Netherlands, the quality of care for patients with type 2 diabetes has improved considerably in recent decades [
6]. In 1998, 55% of patients with type 2 diabetes younger than 75 years had an HbA1c higher than 53 mmol/mol. In 2013, this decreased to 29% [
5]. For patients of 75 years and older, similar trends were found for HbA1c. Poor health behavior places patients with type 2 diabetes at an increased risk of disease progression, impacting their quality of life and increasing their risk of additional health problems and premature death [
7‐
9].
The effectiveness of diabetes self-management ultimately depends on patients’ adherence with lifestyle recommendations and treatment [
10]. Patients need to understand the principles and importance of self-management activities [
11], which makes diabetes self-management education a key component of diabetes care. Diabetes self-management education can be helpful for patients with type 2 diabetes in managing their condition [
12]. It aims not only to enhance the patients’ medical understanding but also to improve their intrinsic motivation, belief in their innate ability to achieve goals (self-efficacy), illness perception, self-management skills, and behavior [
13]. Moreover, a recent systematic review showed that diabetes self-management education resulted in improved HbA1c [
14]. Group-based diabetes self-management education seems to be more effective than usual care and individual education at improving clinical, lifestyle, and psychosocial outcomes [
15].
One example of an evidence-based diabetes self-management education program is the group-based Proactive Interdisciplinary Self-MAnagement (PRISMA) training program. This program is based on the DESMOND (Diabetes Education and Self-Management for Ongoing and Newly Diagnosed) program, which was developed and tested in the UK [
12,
16,
17]. DESMOND has proven to be (cost-)effective in patients newly diagnosed with type 2 diabetes [
16,
17]. PRISMA was adapted for use in type 2 diabetes in primary care in the Netherlands and, in a previous study, seemed to improve self-management behavior in terms of dietary behaviors, foot care, and action planning three months posttraining [
18]. However, the effects of PRISMA on self-reported and clinical outcomes in a controlled setting are still unknown. Therefore, the present study aimed to investigate the effects of PRISMA on self-reported (knowledge, skills and confidence for self-management, diabetes self-care behavior, health-related quality of life, and emotional well-being) and clinical outcomes in patients with type 2 diabetes treated in general practice.
Discussion
We hypothesized that offering the PRISMA program would result in better self-reported and clinical outcomes in patients with type 2 diabetes treated in primary care. However, no effects were found on the following self-reported outcomes: knowledge, skills and confidence for self-management, diabetes self-care behavior, health-related quality of life, and emotional well-being. In addition, it was not possible to make a statement about the clinical effects.
Previous observational research showed that PRISMA appeared to improve self-management behavior in terms of dietary behaviors, foot care, and action planning after 3 months [
18]. The lack of effects in self-reported outcomes in our study may be explained by the already high scores at baseline. These high scores indicated that the patients included in our study had generally well managed type 2 diabetes, so there was limited room for improvement. In addition, Van Vught et al. (2016) observed patients for only 3 months. Changes in outcomes could have been diminished after 6 or 12 months. However, PRISMA was expected to change self-management skills by achieving personal goals (e.g., losing weight), which usually would be visible in clinical measures in the intermediate term or longer. Therefore, in our study, 6 and 12 months were logical terms. Diabetes self-management education interventions appear most effective when group and individualized interventions are combined [
27]. Therefore, patients probably need additional encouragement to change their behavior, such as with follow-up education in either group or individual settings.
Some evidence also suggests that contact exceeding 10 h for diabetes self-management education interventions are more likely to result in additional decreases in HbA1c [
14]. However, PRISMA consisted of two interactive group meetings totaling 7 h [
12]. Thus, additional meetings may have been needed. However, the day-to-day management of type 2 diabetes requires major responsibility from patients. Group education (combined with access to an online care platform) would be an extra activity and, as a result, patients could become overwhelmed. Moreover, our target group might not be interested in this type of interventions, which is the reality of usual care.
A strength of this study was that it was embedded in routine primary diabetes care, which means that it was designed to test PRISMA in the full spectrum of everyday usual care to maximize applicability and generalizability. Furthermore, randomization was performed at a patient level, which generally decreases influences of health care providers. Some limitations of this study must be mentioned as well. First, the eight participating general practices diverged in the inclusion rate: of the 1476 invited patients, most were registered in three general practices. Thus, the extent to which the health care providers motivated their patients to participate could have played a role. However, the characteristics of the patients (sex, age, education) do reflect the general Dutch population with type 2 diabetes. Second, in spite of the expected positive effects, in other studies about 30–50% of the eligible patients do not participate in diabetes education [
28,
29]. In our study this rate was even lower. Despite our efforts to enthuse patients about the PRISMA program, only 12% of the approached patients participated. Therefore, selection bias was very likely. Patients who experience difficulties with self-management behavior probably do not voluntarily take part in studies or do not show up at interventions. Moreover, recruiting from a clinical sample may exclude the patients in greatest need of self-management, i.e. people who actually do not visit a general practice. Our target group may have been uninterested in this type of intervention, which could explain our low participation rate. If especially motivated patients participated, this could also explain why the self-reported and clinical measures were already quite good. However, despite their motivation, 28% of the patients from the intervention group who agreed to participate in the study did not attend at least one meeting of PRISMA. Patients changed their mind because were persuaded to enroll by their healthcare provider despite lack of interest. Patients were invited for PRISMA per letter signed by their own GP. Possibly, more personal attention in terms of reminding patients by telephone a couple of days before PRISMA would have prevented dropouts. Third, in an attempt to avoid a type III error, we monitored as many process factors as possible during the implementation of PRISMA. A type III error occurs when evaluating a program that has not been adequately implemented [
30]. Such low intervention fidelity could decrease the interpretability of the data collected. As mentioned, PRISMA was guided by different PNs and dieticians, and their work experience in the diabetes field varied. However, all trainers used the same protocol and reported deviations from this protocol after the training. Notable deviations were not reported. In an ideal situation, the researcher would personally attend or record all training sessions in order to check intervention fidelity, however, this was not possibly due to a lack of time. Fourth, the great number of missing values in the clinical outcomes should be acknowledged. As a result, it is not possible to make a statement about the clinical effects. This could be explained by the fact that the study was carried out in daily practice. The clinical outcomes were collected annually during the routine diabetes checkups as part of usual care, and no extra laboratory tests were done. As a result, we were dependent on the data delivered as part of usual care. The missing values in the self-reported data were due to non-responses on the questionnaires. Sending reminders may have been a solution for this problem but would have interfered with the restricted time frame in which questionnaires had to be completed. Fifth, because of organizational reasons, the patients completed the baseline questionnaire at the end of the two PRISMA meetings rather than at the start. We realize this is inconsistent with RCTs, however, the two PRISMA meetings should be seen as a starting point to motivate/activate patients, with behavior change as the final objective. Because we were primarily interested in the outcomes during the months, the influence of the intervention on completing questionnaires before or after the intervention was less relevant. Sixth, a sample size calculation was not specifically performed for examining effects on the secondary outcomes for the present study [
19].
Conclusion
In previous observational research, the diabetes self-management education program PRISMA seemed to improve self-management behavior after 3 months. However, in this study, PRISMA did not improve self-reported outcomes (knowledge, skills and confidence for self-management, diabetes self-care behavior, health-related quality of life, and emotional well-being) after 6 or 12 months. In addition, it was not possible to make a statement about the clinical effects given the large number of missing values. The lack of effects on the outcomes in the current study does not automatically mean that PRISMA is not effective in improving self-management skills. Our target group may have been uninterested in this type of intervention, which is the reality of usual care. Therefore, additional research is necessary. In pragmatic trials such as the current study, it is essential to monitor all possible information during the implementation phase of the intervention. This will improve the reliability of the data collected.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.