Skip to main content

Table 3 Summary of findings behaviour and health outcomes: results in intervention group (IG) as compared with control group (CG) (n = 17)a

From: Decision aids to help older people make health decisions: a systematic review and meta-analysis

Short title Choice Adherence with chosen option Preference-linked health outcomes (e.g. anxiety, depression, regret) Health outcomes Health services outcomes
Davison (1997) [28] n.m. n.m. Lower state anxiety levels at 6 weeks.
No differences in levels of depression at 6 weeks.
n.m. n.m.
Dolan (2002) [29] No differences in decision (i.e. the proportion of colorectal cancer screening plans carried out) b n.m. n.m. n.m. n.m.
Fraenkel (2007) [30] n.m. n.m. Greater arthritis self-efficacy. n.m. n.m.
Fraenkel (2012) [31] A small proportion of the IG (n = 5) expressed a preference for medication that was not concordant with their current treatment plan.
No change in treatment plan.
n.m. No differences in anxiety.
No differences in worry about stroke and about bleeding.
n.m. n.m.
Hanson (2011) [32] n.m. After 3 months: Residents in the IG had greater use of some assisted oral feeding techniques (i.e. were more likely to receive a dysphagia diet and showed a trend toward greater staff eating assistance).
No differences in tube feeding after 9 months.
No differences in decisional regret for surrogates. Less weight loss after 9 months.
No differences in mortality.
n.m.
Man-Son-Hing (1999) [35] More IG patients made a definite choice about antithrombotic therapy (aspirin or warfarin).
Slightly more IG patients preferred to continue taking aspirin rather than switch to warfarin.
A similar % of IG and CG reported that they, rather than their physician, made the decision.
After 6 months: A similar % of IG and CG continued to take the therapy that was initially chosen. n.m. n.m. n.m.
Mathers (2012) [36] No sign difference in proportion undecided (although patients in IG were three times more likely to change from undecided to decided). n.m. n.m. No significant difference in the glycaemic control. n.m.
Mathieu (2007) [37] IG women were less likely to be undecided.
Among those women who made a decision, there were no differences in intention to stop or continue screening.
No differences in participation in screening.
n.m. No differences in anxiety.
No differences in breast cancer worry.
n.m. n.m.
McAlister (2005) [38] After 3 months: Increase in the proportion of patients receiving therapy appropriate to their stroke risk (i.e. 12 % absolute improvement in IG as compared to CG). After 12 months: No difference in proportion of patients taking appropriate therapy (i.e. care in both IG and CG had regressed towards baseline levels). n.m. n.m. n.m.
Montori (2011) [39] No differences in distribution of prescriptions (bisphosphonates were started by 44 % of IG patients and 40 % of CG patients). Adherence at 6 months was similarly high across both groups (self-reported), but the proportion with more than 80 % of days covered was higher in the IG (pharmacy records). n.m. n.m. n.m.
Partin (2004) [40] IG1 and IG2 were less likely to intend to have a PSA.
2 weeks post-target appointment: No differences in PSA testing rates.
1 year post-target appointment: No differences in PSA testing rates. n.m. n.m. n.m.
Stirling (2012) [42] n.m. n.m. Less increase in carer burden (however not statistically significant, possibly due to small sample size). n.m. n.m.
Thomson (2007) [44] Participants in the IG not already on warfarin were much less likely to start warfarin than participants not already on warfarin in the CG. n.m. There was a significant fall in anxiety immediately after the clinic, but no differences in reduction between IG and CG. No differences in strokes and bleeds requiring admission. No differences in GP consultations.
No differences in hospital appointments.
Volandes (2009b) [46] IG group was more likely to prefer comfort care as their goal of care. 6 weeks after the intervention: IG had more stable preferences over time.    
Volandes (2011) [47] IG group was more likely to prefer comfort care as their goal of care. n.m. n.m. n.m. n.m.
Weymiller (2007) [48] 30 % of IG patients and 21 % of CG patients not receiving statin at baseline started statin therapy immediately after the visit (not reported whether this difference was significant). IG patients with 10-year cardiovascular risk greater than 15 % most often started statin therapy. Using the decision aid was not associated with stopping statin therapy and was associated with greater statin adherence at 3 months. Of 33 IG patients taking statin drugs at 3 months, 2 reported missing 1 dose or more in the last week compared with 6 of 29 patients in the CG group taking statin drugs.
Overall, there was no difference in adherence to patient choice at 3 months.
n.m. n.m. n.m.
Wolf (2000) [49] No difference in screening interest between the two IGs and the CG b n.m. n.m. n.m. n.m.
  1. n.m. = not measured; IG = intervention group; CG = control group
  2. aUnless otherwise stated are the described results effects in the intervention group (IG) as compared to the control group (CG); see Additional file 2 for description of the CG intervention. Standard font indicates positive results (p < .05 unless otherwise stated) in favour of the IG; italic font indicates no significant results
  3. bIncluded in meta-analysis