Table 1​1 summarises the findings of the systematic reviews. While the results of some reviews were negative (no difference between intervention and control or worse outcome with the intervention) or mixed (positive for some outcome measures and negative for others), most were positive (beneficial effect) for the outcomes selected for our overview.

Evidence from reviews suggests that well designed written information (such as leaflets) can be a useful adjunct to professional consultation and advice and can improve health knowledge and recall, especially if information is personalised. For example, providing information and educational support to cancer patients improved their understanding of their condition and prognosis, leading to reduced anxiety and improved symptom management.³ As another example, preoperative information improved surgical patients' knowledge and sense of empowerment.⁴ Leaflets on their own have little effect, but combined oral and written information can improve patients' experience and, in some cases, reduce use of health service resources.

Other resources, such as websites, can also improve knowledge, and studies of such resources have shown high user satisfaction and beneficial effects on self efficacy and health behaviour. An initiative in the United States that provided computer based support for disadvantaged populations found that they benefited more from this type of intervention than did other groups, perhaps because they have more to gain from health information.⁵ Evidence suggests that interventions designed to mitigate the effects of low literacy can improve knowledge and health behaviour, but few studies have shown an impact on reducing inequalities in health status.⁶

Targeted mass media campaigns have been used to inform patients and the public, often to promote specific health behaviours or patterns of service use—for example, to increase uptake of screening or to reduce inappropriate use of certain drugs, procedures, or services. Well organised campaigns can have a beneficial effect on behaviour, but it is hard to disentangle the effects on patients from those on professionals, as mass media campaigns can reach and potentially influence both groups.^{7 8}

To facilitate shared decision making, patient decision aids have been developed for a wide variety of conditions and treatments. An inventory of such aids is available from the University of Ottawa's Health Research Institute (http://decisionaid.ohri.ca/cochinvent.php). Decision aids have been tested in several randomised controlled trials, and we found 10 systematic reviews evaluating their effectiveness. The reviews show that such aids improve patients' knowledge and understanding of their condition and of the treatment options and outcome probabilities; they also improve agreement between patients' preferences and subsequent treatment decisions.⁹

Decision aids can be cost effective, especially if coupled with face to face counselling. For example, a large randomised trial of decision support for patients with menorrhagia, which included a video of the treatment options and outcomes, an accompanying booklet, and a structured interview with a nurse to help patients express their preferences, resulted in reduced hysterectomy rates and lower mean overall service costs—$1566 (£794; €1178) in the intervention group compared with $2751 in the control group (mean difference $1184; 95% confidence interval $684 to $2110).¹⁰

Educational interventions designed to increase participation of patients in treatment decisions have also been shown to be effective. Coaching and question prompts for patients and training in communication skills for health professionals can improve patients' knowledge and information recall and help to increase participation in decision making.

Approaches that provide information only are mostly unsuccessful, but educational and self help programmes that are actively supported by clinicians improve health outcomes for patients with depression, eating disorders, asthma, diabetes, and hypertension. A review that looked at group based educational programmes for patients with type 2 diabetes found that they improved blood pressure, fasting blood glucose concentrations, glycated haemoglobin, patients' knowledge of their condition, and quality of life.¹² Short self management courses run by voluntary groups seem to improve knowledge, coping behaviour, adherence, self efficacy, and cost effectiveness, but the effects may diminish over time.¹³ Educational programmes involving health professionals seem to be more effective for disadvantaged populations.¹⁴ Interactive computer based support systems also have positive effects on a range of outcomes, including knowledge (standardised mean difference 0.46, 0.22 to 0.69), social support (0.35, 0.18 to 0.52), and clinical outcomes (0.18, 0.01 to 0.35), but it is not clear whether they are more cost effective than face to face delivery.¹⁵

Patient held records can enhance patients' knowledge and sense of control.¹⁶ Self monitoring of blood pressure, blood glucose, and oral anticoagulation, as well as remote telemonitoring, can be both effective and cost effective.¹⁷ Research into home based telecare is still in its infancy. Studies so far indicate potential benefits in terms of social support, quality of life, use and cost of health services, and, in some cases, physiological outcomes.¹⁸

Various strategies for improving adherence to treatment have been evaluated. The most effective involve simplifying dosing regimens.¹⁹ The results of other interventions—such as patient education, information, and counselling—are less conclusive. Evidence suggests that encouraging patients to ask health workers if they have washed their hands can improve hygiene if the provision of hand washing facilities is also improved.²⁰

The systematic reviews we identified varied in how they described the interventions studied, the outcomes measured, and whether outcomes were presented qualitatively or quantitatively. Some provided no clear descriptions of the interventions or the study context. Patients' demographic and clinical characteristics were sometimes missing or poorly reported; the length of follow-up was often relatively short, so the longer term effectiveness of many of the interventions is unknown; and few studies evaluated the cost effectiveness of interventions or considered their potential opportunity costs.

Nevertheless, we believe there is a substantial evidence base, albeit imperfect, on which to build strategies to strengthen patient engagement. Most reviews reported improvements in important outcomes, and several promising avenues to pursue have been identified.

Because health literacy is central to enhancing involvement of patients in their care, all strategies to strengthen patient engagement should aim to improve health literacy. Many people will have difficulty taking advantage of these new opportunities if the problem of health literacy is not dealt with. This could widen health inequalities, or even create new ones.

Thus, patients' knowledge and understanding can be improved, at least at the individual level. Patients with acute and chronic health problems benefit when they are involved in their care, both at home and in clinical settings, and evidence suggests that this can lead to better use of resources. Shared decision making and self management are mutually supportive approaches, which should be given equal importance and implemented consistently.