Outcomes Research in Review

A Decision Aid Did Not Improve Patient Empowerment for Setting and Achieving Diabetes Treatment Goals

Denig P, Schuling J, Haaijer-Ruskamp F, Voorham J. Effects of a patient oriented decision aid for prioritising treatment goals in diabetes: pragmatic randomised controlled trial. BMJ 2014;349:g5651.



Study Overview

Objective. To determine if a patient-oriented decision aid for prioritizing treatment goals in diabetes leads to changes in patient empowerment for setting and achieving goals and in treatment.

Design. Randomized controlled trial.

Setting and participants. Study participants were recruited from 18 general practices in the north of the Netherlands between April 2011 and August 2012. Participants were included if they had a diagnosis of type 2 diabetes and were managed in primary care. Participants were identified from the electronic medical record system and at least 40 patients were selected from each practice to be contacted for participation. Subjects were excluded if they had myocardial infarction in the preceding year, experienced a stroke, had heart failure, angina, or a terminal illness, or were more than 65 years of age when they received their diabetes diagnosis. Other exclusion criteria include dementia, cognitive deficits, blindness, or an inability to read Dutch. Eligibility criteria were confirmed with the health care provider from each practice. Practices that were included in the study had several features: (1) each had an electronic medical record system supporting structured care protocols; (2) most practices have a nurse practitioner or specialized assistant for diabetes care who carries out the quarterly diabetes checks and is trained to conduct physical examinations, risk assessments, patient education, and counseling; (3) all practices received training in motivational interviewing.

The decision aid format was either a computer screen or printed version, and presented as either a short version, showing treatment effects on myocardial infarction risk only, or as an extended version, including effects on additional outcomes (stroke, amputation, blindness, renal failure). Practices were randomly assigned to use the computer screen or printed version, stratified by practice size (< 2500 patients or > 2500 patients) and number of GPs (solo or several). Within each practice, consenting patients were randomized to receive the short version aid, the extended version, or to the control group.

Intervention. The decision aid presents individually tailored information on risks and treatment options for multiple risk factors. The aid focuses on shared goal setting and decision making, particularly with respect to the drug treatment of risk factors including hemoglobin A1c, systolic blood pressure, low density lipoprotein cholesterol, and smoking. The decision aid is designed to be used by patients before a regular check-up and discussed with their health care provider during a visit to help prioritize treatment that will maximize outcomes; the aid helps to summarize effects of the various treatment options. The patients were asked to come to the practice 15 minutes in advance to go through the information, either in print or on the computer; health care providers were expected to support patients to think about treatment goals and options. Patients in the control received care as usual.

Main outcome measures. The primary outcome measure was the empowerment of patients for setting and achieving goals, which was measured with the Diabetes Empowerment Scale (DES-III). Other outcome measures included changes in treatment, including intensification of drug treatment and treatment with ACE inhibitors.

Main results. A total of 344 patients were included in the study and were randomized to the intervention ( n = 225) or usual care group ( n = 119). Patients in the intervention group were comparable to usual care patients in terms of age, sex, and educational level. However, there were several differences between the 2 groups: intervention patients were more likely to have well-controlled HbA1c level at baseline and less likely to have well-controlled blood pressure at baseline. Among participants in the intervention group, only 46% reported to have received the basic elements of the intervention. The mean empowerment score increased 0.1 point on a 5-point scale in the intervention group, which was not different from the control group (mean adjusted difference, 0.039 points [95% confidence interval {CI}], −0.056 to 0.134). Lipid lowering medication treatment was intensified in 25% of intervention and 12% of control participants (odds ratio [OR], 2.5 [95% CI, 0.89–7.23]). Explorative analyses comparing printed version of the aid with control did find that lipid lowering medication treatment was more intensified although the confidence interval was wide (OR, 3.90 [95% CI, 1.29–11.80]). No other differences in treatment plan were observed.

Conclusions. The treatment decision aid for diabetes did not improve patient empowerment or substantially alter treatment plan when compared to usual care. However, this finding is limited by the uptake of use of the decision aid during the study period.


Patient engagement through shared decision making is an important element in chronic disease management, particularly in diseases such as diabetes where there are a number of significant tasks, including monitoring and administration of medication, that are key to its successful management. The use of decision aids is an innovation that has demonstrated effects in improving patient understanding of disease, and has potential downstream effect in improving management and control of the disease [1]. However, the use of decision aids is not without limitations—patients with poorer health literacy, and perhaps lower socioeconomic status, may derive less clinical benefit [2], and in older adults cognitive and physical limitations may also limit their use.

This study found that the decision aid used in the study did not significantly improve patient empowerment or alter treatment plan. In comparison with previous studies on decision aids for diabetes [3,4], this study is notable that it did not find any significant clinical impact of the decision aid when compared with usual care. However, it is important to consider reasons that may explain its null finding. First, the study has a rather complicated design, with 4 different intervention groups. The study design attempts to differentiate intervention groups with differences in its delivery (computer screen vs. printed) and content (focused information on myocardial infarction risk outcome only vs. all outcomes). The rationale was that it could provide evidence to perhaps suggest the most effective decision aid, but the drawback is that it has the potential to weaken the power of the study, increasing the likelihood of a false-negative finding. Second, in contrast to other studies, this study also uses a different measurement as its primary outcome—a measurement of patient empowerment. Though an important concept to measure, it is less clear what the expected impact and what the level of clinical significance would be. Third, as noted by the investigators, the decision aid had limited uptake in the intervention group; this may be related to its design and format. The challenge in design of a decision aid is that it needs to be simple and easy to use, consume little time, yet be adequately informative with helpful information for patients. Finally, another unique feature of the study is that the control group was an active control group, in that the providers in the practices had significant training in motivational interviewing and communication, which may have made it more challenging to demonstrate impact in intervention group.

Applications for Clinical Practice

Decision aids remain a potentially important addition for patients in the management of chronic diseases such as diabetes. Most studies have demonstrated significant impact. Despite the limitations of the current study, it does point out that different formats of decision aid may have different effects on patient outcomes. For practices that are adopting decision aids for chronic disease management, they need to take into account the format, the information, and the burden of use of the decision aid. Further studies may help to elucidate how decision aids can be optimized for maximizing clinical impact.

—William Hung, MD, MPH

Next Article: