Absolute Cardiovascular Disease Risk and Shared Decision Making in Primary Care: A Randomized Controlled Trial

Tanja Krones; Heidemarie Keller; Andreas Sönnichsen; Eva-Maria Sadowski; Erika Baum; Karl Wegscheider; Justine Rochon; Norbert Donner-Banzhoff

doi:10.1370/afm.854

Article Figures & Data

Figures

Tables
Additional Files

Download figure
Open in new tab
Download powerpoint
Figure 1.
Study flowchart—CME groups, practices, and patients.
CME = continuing medical education.

View popup

Table 1.

Decision Aid Steps: ARRIBA-Herz – The Script

Step	Components
1. Agree on task	Calculate individual absolute risk for stroke and/or myocardial infarction, involving patient in decision making
2. Explore subjective risk	Address fears, expectations, preferences, questions of patients
3. Calculate and show objective risk	Assess individual risk factors (eg, age, blood pressure, total cholesterol), calculate risk, compare with population with identical sex and age, provide probabilities in absolute numbers
4. Present preventive options	Include behavioral change and medication
5. Discuss pros and cons	Include nonaction
6. Agree on course of action	Include date of next consultation

View popup

Table 2.

Baseline Characteristics of Participating Family Doctors by Intervention (n = 44) and Control Group (n = 47)

Baseline Variables	Intervention Group No. (%)	Control Group No. (%)	χ²	P Value
GP age, years			3.8	.279
31–40	2 (4.5)	1 (2.1)
41–50	24 (54.5)	24 (51.1)
51–60	17 (38.6)	16 (34.0)
>60	1 (2.3)	6 (12.8)
Male	27 (61.4)	26 (55.3)	3.4	.671
Geographical location (% urban)	20 (45.5)	20 (42.6)	0.07	.837
Practice size (practice attendance per 3-month period)			11.4	.01
<500	1 (2.3)	3 (6.4)
500–1,000	6 (13.6)	19 (40.4)
1,000–1,500	21 (47.7)	18 (38.3)
>1,500	16 (36.4)	7 (14.9)

View popup

Table 3.

Baseline Characteristics of Patients by Intervention and Control Group.

Variable	Intervention Group (n=550)	Control Group (n=582)	χ² (df)	P Value
^a Mean cardiovascular risk in % for 10 years at baseline, calculated by Framingham risk points, is displayed for patients in primary prevention only, including patients with diabetes.
Mean age, years (SD)	59.1 (12.3)	58.6 (12.5)	0.69	.49
Male, No. (%)	231 (42.0)	265 (45.5)	1.4 (1)	.26
Nationality German, No. (%)	528 (96.0)	566 (97.3)	0.83 (1)	.46
Education, No. (%)			2.1 (4)	.71
No or basic education	284 (51.6)	320 (54.9)
Medium education	169 (30.7)	168 (28.9)
Higher education	91 (16.6)	91 (15.6)
Smokers, No. (%)	102 (18.5)	108 (18.6)	0.0 (1)	1.0
Diabetic, No. (%)	63 (11.5)	140 (24.1)	30.1 (1)	<.001
Previous cardiovascular disease, No. (%)	101 (18.4)	130 (22.3)	2.7 (1)	.10
Mean cardiovascular risk at baseline^a (SD)	10.7 (9.5)	10.1 (9.2)	098 (899)	.33
Subjective health status, No. (%)			5.2 (4)	.27
Very good/good	266 (48.4)	260 (44.7)
Satisfactory	242 (44.0)	278 (47.8)
Very bad/bad	36 (6.5)	40 (6.9)
Preference of participation in decision, No. (%)			13.4 (4)	.009
Patient only	6 (1.1)	11 (1.9)
Patient mainly	14 (2.5)	16 (2.7)
Patient and doctor	405 (73.6)	383 (65.8)
Doctor mainly	99 (18.0)	114 (19.6)
Doctor only	15 (2.7)	39 (6.7)

View popup

Table 4.

Outcome Variables by Intervention, All Clusters (14 CME Groups) Contributing to Every Analysis

Variable	Intervention Group^a	Control Group^a	Estimated ICC of CME Group/Practice	Difference Intervention – Control (95% CI)	Adjusted F Statistic	P Value^a
CME = continuing medical education; CVD = cardiovascular disease; ICC = intraclass correlation; SDM = shared decision making.
^a Values are adjusted estimated means, numbers of patients (n) and P values in mixed models adjusted for confounders (see Methods) and cluster structure (CME group, practice nested within CME group, patients nested within practice).
^b Lower scores denoting higher participation and satisfaction.
^c Mean change of CVD risk (%) was calculated by Framingham risk points for all patients, including patients with diabetes and secondary prevention (focus on relative change).
Evaluation after index consultation
Patient participation and satisfaction^b	6.76 (501)	7.56 (536)	0.00/0.06	−0.80 (−1.23 to −0.37)	13.55	<.001
SDM steps reported	9.48 (407)	7.49 (442)	0.00/0.07	1.99 (1.27 to 2.71)	30.21	<.001
Knowledge	2.03 (535)	1.92 (576)	0.001/0.06	0.11 (−0.01 to 0.24)	3.36	.07
Follow-up examination
Mean change of CVD risk^c	−3.00 (415)	−3.33 (407)	0.006/0.02	0.32 (−0.30 to 0.95)	1.07	.31
Decisional regret	14.69 (372)	18.08 (372)	0.00 /0.02	−3.39 (−6.26 to −0.53)	6.58	.02

View popup

Table 5.

Steps of Shared Decision-Making Process During Consultation as Reported by Patients Directly After Consultation

Steps and Statements	Intervention Group % Agree (n)	Control Group % Agree (n)	P Value^a	Estimated ICC of CME Group/Practice^a
^a After adjustment for confounders, see also methods section.
^b Because item expresses the opposite, high percentages representing less involvement of patients, it has been inverted before summing up shared decision making steps in Table 4.
Step 1. Disclosure, that a decision needs to be made
My doctor told me that a treatment decision is necessary	42.9 (236/550)	39.0 (227/582)	.160	0.035/0.125
Step 2. Formulation of equality of partners
My doctor asked me, if I want to participate in decision making	64.2 (353/550)	46.1 (268/582)	<.001	0.000/0.083
Step 3. Equipoise statement^b
Due to my medical condition, treatment decision based on physicians’ recommendation is already clear	53.8 (296/550)	58.4 (340/582)	.867	0.000/0.000
Step 4. Informing on the options’ benefits and risks
My doctor has informed me about a variety of alternatives	64.0 (352/550)	47.9 (279/582)	<.001	0.000/0.203
The possibility to choose no treatment was also discussed	55.3 (304/550)	36.9 (215/582)	<.001	0.000/0.076
Step 5. Investigation of patient’s understanding and expectations
I have mentioned other possibilities that my doctor has not referred to	14.4 (79/550)	8.9 (52/582)	.075	0.000/0.089
My doctor has asked me how I think about different treatment options	62.9 (346/550)	37.6 (219/582)	<.001	0.000/0.281
Step 6. Identification of preferences (both)
I have communicated to my doctor which decision I do prefer	55.5 (305/550)	34.7 (202/582)	<.001	0.000/0.145
My doctor has told me which decision he prefers	68.9 (379/550)	57.6 (335/582)	<.001	0.000/0.153
Step 7. Negotiation
In the selection of treatment method, my thoughts were taken into account just as much as the considerations of my doctor	77.1 (424/550)	61.3 (357/582)	<.001	0.001/0. 184
My doctor and I weighted up the different treatment options thoroughly	67.6 (372/550)	51.2 (298/582)	<.001	0.000/0. 113
Step 8. Shared decision making
My doctor enabled me to actively participate in decision making about treatment	79.1 (435/550)	64.3 (374/582)	<.001	0.000/0.118
My doctor and I selected a treatment together	65.8 (362/550)	55.8 (325/582)	.004	0.000/0.149
Step 9. Arrangement of follow-up
My doctor and I reached an agreement as to how we will proceed	78.9 (434/550)	70.4 (410/582)	.006	0.013/0. 196

Additional Files

Figures
Tables

Supplemental Appendix

Supplemental Appendix. ARRIBA-Herz Decision Aid

Files in this Data Supplement:

Supplemental data: Appendix - PDF file, 3 pages, 455 KB
The Article in Brief

What This Study Found Patients who use the decision aid are more involved in decision making, more satisfied, and have less regret about their health care decisions.

Implications

This decision aid helps doctors and patients discuss individual concerns and pace information according to patients� needs.
Decision aids completed by patients on their own, as well as those used during a visit to the doctor, are important, might be combined, and deserve further study.