A total of 66 practices participated and an interview with one member of all disciplines present (GPs, nurses, dieticians) was conducted. The response rate differed across subgroups: 16% among monodisciplinary practices, 49% among multidisciplinary fee-for-service practices and 94% among multidisciplinary capitation practices. Table 2 shows participating practices’ characteristics. We will first outline the quantitative data and then explain our results qualitatively, by each element of the CCM.
Table 2: Practice Characteristics (n=66)
Variable
|
Value
|
n
|
Percent
|
Region
|
Antwerp
|
26
|
39%
|
|
Ghent
|
23
|
35%
|
|
Campine
|
17
|
26%
|
Type of GP practice
|
Monodisciplinary, fee-for-service, solo
|
16
|
24%
|
|
Monodisciplinary, fee-for-service, group
|
14
|
21%
|
|
Multidisciplinary, fee-for-service
|
19
|
29%
|
|
Multidisciplinary, capitation
|
17
|
26%
|
Nurse
|
Yes
|
27
|
41%
|
|
No
|
39
|
59%
|
Dietician
|
Yes
|
24
|
36%
|
|
No
|
43
|
65%
|
Secretary
|
Yes
|
45
|
68%
|
|
No
|
21
|
32%
|
Age OF interviewed GP (years)
|
<35
|
29
|
44%
|
|
35–50
|
16
|
24%
|
|
>50
|
21
|
32%
|
Gender OF interviewed GP
|
Male
|
33
|
50%
|
|
Female
|
33
|
50%
|
ACIC scores by region
There was no significant difference between the regions on the total ACIC score and its separate elements. All elements, except the first two, were measured on the practice level, so p-values could be calculated (Table 3). The first two elements were scored using the answers of the meso-level stakeholders (so that there was no variation between the practices in one region, resulting in the statistical analysis not making sense). The practices in the region of Ghent scored the highest, followed by the regions of the Campine and Antwerp.
Table 3: Average ACIC scores by region
ACIC Elements
|
URBAN
|
rural
|
|
|
Ghent
|
Antwerp
|
Campine
|
|
|
x
|
sd
|
x
|
sd
|
x
|
sd
|
p-value*
|
1. Organisation of the healthcare delivery system
|
8.17
|
N/A
|
4.67
|
N/A
|
5.83
|
N/A
|
N/A
|
2. Community linkages
|
3.84
|
N/A
|
2.53
|
N/A
|
4.24
|
N/A
|
N/A
|
3.a Self-management support
|
5.22
|
1.99
|
4.19
|
1.82
|
4.18
|
1.60
|
0.10
|
3.b Decision support
|
3.63
|
1.42
|
3.62
|
1.09
|
3.97
|
0.97
|
0.58
|
3.c Delivery system design
|
4.35
|
2.64
|
3.74
|
1.66
|
4.26
|
2.11
|
0.57
|
3.d Clinical information system
|
4.38
|
2.73
|
4.42
|
2.16
|
3.50
|
1.82
|
0.39
|
TOTAL acic SCORE
|
4.93
|
1.64
|
3.86
|
1.26
|
4.33
|
1.10
|
N/A
|
*ANOVA test; sd: standard deviation; N/A: not applicable
ACIC scores by type of GP practice
Capitation practices scored significantly higher than multidisciplinary fee-for-service practices, which scored, on their turn, higher than monodisciplinary practices on all the ACIC elements, except the first one (see Table 4 and Additional file 1).
Table 4: Average ACIC scores by type of GP practice
ACIC ELEMENTS
|
Fee-for-service
|
|
|
Mono-disciplinary
|
Multi-disciplinary
|
Multi-disciplinary
|
|
|
x
|
sd
|
x
|
sd
|
x
|
sd
|
p-value*
|
1. Organisation of the healthcare delivery system
|
6.22
|
1.48
|
5.77
|
1.37
|
6.59
|
1.75
|
0.28
|
2. Community linkages
|
2.53
|
1.41
|
3.35
|
1.65
|
5.08
|
1.13
|
<0.001
|
3.a Self-management support
|
3.19
|
1.51
|
4.82
|
0.97
|
6.63
|
0.92
|
<0.001
|
3.b Decision support
|
3.06
|
1.09
|
3.84
|
0.96
|
4.72
|
0.75
|
<0.001
|
3.c Delivery system design
|
2.90
|
1.80
|
3.99
|
1.75
|
6.28
|
1.25
|
<0.001
|
3.d Clinical information system
|
2.27
|
1.22
|
4.48
|
1.14
|
7.16
|
0.99
|
<0.001
|
Total score
|
3.36
|
1.02
|
4.38
|
0.83
|
6.08
|
0.77
|
<0.001
|
|
|
|
|
|
|
|
|
|
*ANOVA test; sd: standard deviation
In Table 5, the results of the multiple regression analyses are presented. The financing system is significantly related to the ACIC score. After taking the other practice characteristics into account, GP practices within the capitation system had a significantly higher overall ACIC score compared to GP practices within the fee-for-service system. The strength of this relation, however, decreases in particular when controlling for whether the practice had a nurse, indicating a mediating effect in addition to the direct effect of the financing system. Bivariate statistics showed that GP practices within the capitation system more often have a nurse, and in model 2, we observed that GP practices with a nurse scored significantly higher on the ACIC. The total ACIC score was not related to the inclusion of a dietician in the practice. A total of 14% of the variance in ACIC scores between the practices can be ascribed to the presence or absence of these paramedics. GP practices with a secretary also scored significantly higher on the ACIC, irrespective of the financing type of the practice and whether or not there was a dietician and/or a nurse. By adding ‘secretary’ to the model, an additional 3% of the variance was explained. Finally, 75% of the variance in the ACIC scores was explained by the financing system, region, and having a nurse, dietician or secretary. When looking at the last model, nurses have the highest impact, followed by being a capitation practice and having a secretary. This last model was re-estimated for the separate dimensions of the ACIC (see Additional file 2).
Table 5: Multivariate regression analysis on the total ACIC score
|
Model 1
|
Model 2
|
Model 3
|
|
b
|
se
|
sign
|
b
|
se
|
sign
|
b
|
se
|
sign
|
Financing system (ref. ‘fee-for-service’)
|
|
|
|
|
|
|
|
|
|
Capitation
|
2.392
|
0.269
|
<0.001
|
1.130
|
0.313
|
<0.001
|
1.073
|
0.299
|
<0.001
|
Region (ref. ‘Antwerp’)
|
|
|
|
|
|
|
|
|
|
Ghent
|
0.779
|
0.264
|
0.004
|
0.946
|
0.218
|
<0.001
|
1.092
|
0.216
|
<0.001
|
Campine
|
0.973
|
0.291
|
0.001
|
0.961
|
0.237
|
<0.001
|
0.936
|
0.226
|
<0.001
|
Nurse (ref. ‘No’)
|
|
|
|
1.423
|
0.275
|
<0.001
|
1.149
|
0.282
|
<0.001
|
Dietician (ref. ‘No’)
|
|
|
|
0.279
|
0.203
|
0.175
|
0.129
|
0.202
|
0.526
|
Secretary (ref. ‘No’)
|
|
|
|
|
|
|
0.650
|
0.247
|
0.011
|
Adjusted R²
|
0.588
|
|
|
0.726
|
|
|
0.751
|
|
|
se: standard error
Organisation of the HC delivery system
The region of Ghent scored higher than the region of the Campine, which scored higher than the region of Antwerp. As these scores were based on the interview with one single respondent for each region—a meso-level stakeholder—no statistical analysis could be performed. The reason the region of Ghent scored higher than the other regions is mainly due to the fact that they have set up a health council for a decade. This council devised an improvement strategy, with measurable goals, which are reviewed routinely. In the other regions, the improvement strategy was rather implicit. Additionally, senior leaders visibly participate in the efforts for the improvement of chronic care in Ghent. The region of the Campine scored higher than Antwerp because they have set up a prevention centre, in which multiple behaviour-change interventions are available. The revenues from the GPs are used to fund the centre. In all the regions, scores for incentives and regulations for chronic illness care were low. Respondents mentioned that the current Belgian financing system does not provide incentives to improve quality of care, and that they cannot influence this. One respondent mentioned that in their region, they try to stimulate professional satisfaction instead; to stimulate the quality of chronic care in a different way.
Community linkages
Having a nurse and being a capitation practice had a significant impact on element 2 of the CCM: community linkages. Practices with nurses more often referred patients to initiatives organised by other partners in the community. These could be exercise classes, cooking classes, peer-support programmes, exercise coaching or social care. In practices where nurses were given extensive responsibility and also actively sought to change patient behaviour, referring patients to other initiatives was often their task and they sometimes actually had a list of places to which they could refer patients. In practices where the role of the nurse was mainly technical, they more often considered this to be the task of the GPs. In most capitation practices, the tasks of the nurses were elaborated. While in some fee-for-service practices that was also the case; however, in others, they mainly performed technical tasks. Most GPs, with few exceptions, often did not have the time or were not aware of community initiatives. Real collaboration with these initiatives was rare, and when it did occur, it was in some of the capitation centres that had dedicated staff for this purpose—a health promoter.
Self-management support
Being a capitation practice, having a nurse, dietician and secretary impacted the score for self-management support, with the nurse having the biggest impact. Nurses often adopt a more structured approach than GPs and, on the basis of a protocol, discuss all aspects that can be affected by T2D. Some participants also indicated that patients found it easier to tell their concerns to the nurse than to the GP. Explaining to patients how they should check their blood pressure or glucose level themselves was often part of the nurse’s job. Not all nurses were responsible for effective interventions related to behavioural change; especially when their tasks were purely technical, they left this to other healthcare providers. Dieticians are eminently in charge of changing behaviour, but in contrast to the nurse, they mostly served as an optional provider; not all patients went to see the dietician. Some dieticians even said that they received very few referrals from the GP and mainly provided care to patients visiting them independently. Secretaries helped in arranging appointments with educators and other providers and in such a way, supported self-management.
Decision support
Capitation practices scored higher on decision support compared to multidisciplinary fee-for-service practices, which scored higher than monodisciplinary practices. However, the difference between the practices in the decision support element was the smallest of all elements. In the regression analysis, the different parameters were all borderline significant and only explained 40% of the differences between practices. It was observed that using guidelines, involving specialists or taking education also happens in monodisciplinary practices. However, multidisciplinary practices made more use of guidelines, and in capitation practices, it was sometimes observed that guidelines were adapted into a specific practice protocol to cooperate across disciplines. Many nurses and dieticians follow regular training on diabetes and a few nurses had followed the general practice nurse training, where the organisation of chronic conditions is discussed in detail. Finally, informing patients about guidelines is also a task that nurses and dieticians take on—they use leaflets or other material for education more often than doctors.
Delivery system design
Concerning the delivery system design element, compared to all other elements, the parameters in the regression analysis were able to explain most (81%) of the differences between the practices. Being a capitation practice, having a nurse and having a secretary had a large impact, whereas having a dietician did not have an impact. When a GP practice decides to work with a nurse or secretary, tasks can be better divided between them. Practices with the highest scores hold meetings regularly and also have a prior agreement on which care provider will take on which tasks. When the group of caregivers grows larger, cooperation runs more smoothly if one caregiver takes the lead in organising the care for the patients with T2D. This dialogue between healthcare workers (HCWs) is one of the points where practices with nurses differ from practices with only a dietician as the paramedic. The dietician is often more external to the group, where, in the fee-for-service practices, the dietician rents a consultation room from the doctors as an independent provider and is, therefore, less accountable to them, in contrary to the nurse and secretary, who are paid by the GP budget. Dieticians also often spend fewer hours in the GP practice and can combine it with consultations in other places. Practices with a nurse often have more concrete agreements about when the patient should come for a consultation, usually every three months. Nurses and secretaries are also more likely to actively contact or call the patient if they do not show up and assure appointments in the hospital are planned, if needed. Planned consultations, in which only the chronic disease is managed and who are prepared by certain activities, such as a blood test, are more common in practices with nurses. Different models are possible, such as a consultation with the nurse and shortly afterwards with the doctor, or a blood collection beforehand via the lab. The nurses used their protocol to plan the content of their consultations; in some practices, this was even arranged periodically, with a different element of follow-up each year. Nurses also often had more time per consultation than doctors. Most importantly, when patients visit their doctors, they often bring other complaints (apart from the chronic disease), while only asking for a prescription for their chronic medication at the end of the consult. Obviously, in such a way it is difficult to manage diseases such as T2D, respondents said.
Clinical information system
Being a capitation practice and having a nurse had a significant effect on the score on the clinical information system dimension. We observed that the electronic medical record is often better developed in practices with a nurse. This is often an essential tool for communication between healthcare providers within a practice. Practices with nurses are more likely to be able to extract a list of the diabetes patients from the system and work more often with schedules and reminders. Developing and recording a clear care plan with goals, both clinical and self-management, was more common in practices with a nurse, with some practices having developed their own template for this. Although, there was also a big difference between practices where the nurse merely performed technical tasks and practices in which nurses did take up more advanced tasks, the latter being more common in capitation practices. Also, there was a notable difference with practices that only had a dietician: in those practices, the medical records were usually not shared, but referral letters and reports were used. Lastly, in the capitation practices, specific medical software was used, which has the function of planning and reminders in a prominent place, which was used by all the capitation practices.