Table 3 Study characteristics and main findings

Arterburn 2011 [21]

People considering bariatric surgery

Group Health Cooperative in USA

Prospective, randomised controlled trial

101

50.5

Booklet with general information on severe obesity and surgical weight loss treatments

Video and booklet about bariatric surgery + guidance (list of questions to ask clinician)

Both groups improved significantly in knowledge (P < 0.001), values concordance (P = 0.009), decisional conflict (P < 0.001) and decisional self-efficacy (P < 0.001). DA group had larger improvements than control group in knowledge (P = 0.03), decision conflict (P = 0.03), and outcome expectancies (P = 0.001). No difference in proportion of participants choosing bariatric surgery between groups.

Branda 2013 [22]

People with type 2 diabetes considering changing their antihyperglycaemic drugs or lipid lowering strategies

Rural primary care practices in USA

Cluster randomised controlled trial

103

57.6

Lipid therapy medication discussion

OR

Anti hyperglycaemic medication discussion

Diabetes Medication DA: decision cards with treatment information and usual care for lipid therapy

OR

Statin PtDA: 1-page DA with cardiovascular risk with and without medication, treatment information

DA group were more likely to report discussing medications (p < 0.001), answer knowledge questions correctly (risk reduction with statins, p = 0.07; knowledge about options p = 0.002) and were more engaged by their clinician in decision making (p = .01). No difference in patient satisfaction, decisional conflict, medication starts, adherence or clinical outcome

Jibaja-Weiss 2011 [23]

Women diagnosed with breast cancer considering surgical treatment

Two breast pathology clinics in USA

Randomised controlled trial

100

51

Breast cancer educational materials

Computer based information, values clarification and guidance (step by step process for making the decision)

DA group more likely to choose mastectomy (P = 0.018). No difference in satisfaction between groups. Decreased decisional conflict for both groups (P < 0.001) across assessment periods but DA group more informed about options (P = 0.007) and clearer about values (trend at P = 0.053) at pre-surgery assessment.

LeBlanc 2015 [24]

Women with osteopenia or osteoporosis

Primary care practices in USA

Randomised controlled trial

77

67.5

Clinicians discussed risk of fractures and treatment as usual without any research-related intervention OR clinicians were provided with patients’ individualised 10-year risk of bone fracture for use during the clinical encounter

1-page decision aid with personalised risk of having a fracture with and without medication and information about harms and side effects

DA group had better knowledge (P = 0.01), improved understanding of fracture risk and risk reduction with medication (P = 0.01 and P < 0.0001, respectively), increased patient involvement (P = 0.001) but had no effect on decisional conflict. Consultations using DA were 0.8 min longer. DA arm had more patients receiving and filling prescriptions (P = 0.07), medication adherence was no different across arms at 6 months.

LeBlanc 2015 [25]

People with moderate or severe depression

Primary care practices in USA

Cluster randomised trial

297

43.5

No access to decision aid

Seven laminated cards with information about treatments

DA significantly improved patients’ decision comfort (P = 0.02), knowledge (P = 0.03), satisfaction (P = 0.81 to P = 0.002, depending on domain) and involvement (P < 0.001) and clinicians’ decisional comfort (P < 0.001) and satisfaction (P = 0.02). No differences in consultation duration, adherence or improvement in depression control between groups.

Mann 2010 [26]

People diagnosed with diabetes considering statins to reduce cardiovascular risk

Primary care practices in USA

Cluster randomised trial

150

58

Pamphlet about reducing cholesterol through diet

1-page DA with cardiovascular risk with and without medication and information about statins

DA group more likely to accurately estimate risk of cardiac arrest without statin (OR: 1.9, CI: 1.0–3.8) and with statin (OR: 1.4, CI: 0.7–2.8). DA group reported stronger belief in the need for statins (OR 1.45, CI: 0.89–2.36) and were less likely to have concerns about long-term effects (OR: 0.44, CI: 0.20–0.97). DA resulted in improvements in decisional conflict (P = 0.1). No difference in statin adherence or knowledge between groups.

Mathers 2012 [27]

People with type 2 diabetes considering insulin therapy

General practices in UK

Cluster randomised controlled trial

175

64

No access to decision aid

Booklet containing information on treatment options, values clarification + guidance

DA group had lower total Decisional Conflict Scores (p < 0.001); greater knowledge (p < 0.001); realistic expectations (p < 0.001); and more autonomy in decision making (p = 0.012). No significant difference in the glycaemic control between groups.

Montori 2011 [28]

Postmenopausal women at risk of osteoporotic fractures

General medicine and primary care practises in USA

Multicentre, randomised controlled trial.

100

67

Review of bone mineral density results without fracture risk calculation or graphic representation of treatment benefit + general information booklet

1-page decision aid with personalised risk of having a fracture with and without medication and information about harms and side effects

DA group were 1.8 times more likely to correctly identify fracture risk, 2.7 times more likely to identify risk reduction with bisphosphonates and demonstrated improved involvement in decision making process by 23%. Bisphosphonates started more in DA group, adherence similarly high at 6 months, across both groups but proportion with more than 80% adherence was higher in DA group.

Mott 2014 [29]

War veterans with PTSD

PTSD clinic in USA

Randomised controlled trial

27

29.3

No access to decision aid

Booklet describing treatment options

Greater number of people in DA group preferred an evidence-based treatment and received an adequate dose of therapy compared to control (≥9 sessions). No difference in initiation rates of psychotherapy between groups.

Mullan 2009 [30]

People with type 2 diabetes considering treatment options

Primary care and family medicine sites in USA

Cluster randomised trial

85

62.1

12-page pamphlet on oral antihyperglycaemic medications

Six decision cards with information about treatments

DA group had better knowledge and more involvement in decision making. Similar scores for trust in physician and decisional conflict between groups.

At follow-up, both groups had almost perfect medication use but there was no significant impact on HbA1c levels.

Solberg 2010 [31]

Women considering treatment options for uterine fibroids

Gynaecology clinics in USA

Randomised controlled trial

300

46

Pamphlet about uterine fibroids

DVD and booklet, decision worksheet and nurse coach access

DA group reported more options being mentioned, had better knowledge scores, were more likely to report being adequately informed and decisions were both more satisfactory and consistent with personal values.

Vandemheen 2009 [32]

People with cystic fibrosis considering lung transplantation

Outpatient centres in Canada and Australia

Single-blind, randomised controlled trial

149

30.4

Blank pages and a letter explaining why blank pages were included

Booklet with treatment information

DA group had better knowledge about options (P < 0.0001) and more realistic expectations (P < 0.0001). Decisional conflict was significantly lower in DA (P = 0.0007).

Weymiller 2007 [33]

People with type 2 diabetes

Metabolic clinic in USA

Cluster randomised trial

97

65

Standard educational pamphlet on cholesterol management

1-page DA with cardiovascular risk with and without medication and information about statins

DA group had better knowledge, estimated cardiovascular risk and potential absolute risk reduction with statin drugs, and had less decisional conflict. DA missed less doses than control group at 3-month follow-up.