Helping Patients With Type 2 Diabetes Mellitus Make Treatment Decisions
Audrey J. Weymiller, CNP; Victor M. Montori, MD, MSc; Lesley A. Jones; Amiram Gafni, PhD; Gordon H. Guyatt, MD, MSc; Sandra C. Bryant, MS; Teresa J. H. Christianson, BS; Rebecca J. Mullan, MS; Steven A. Smith, MD Arch Intern Med. 2007;167:1076-1082.
The Clinical Trial
Today’s entry is a follow-up to Patients’ Tool Improves Treatment Decisions, Adherence, covering a recently publish study of the effectiveness of the Statin Choice Tool discussed in that post.
Abstract
Additional Study Information
Patients were given a self-administered written questionnaire immediately that included 7-point scales to explore patient perceptions of the amount, clarity, and helpfulness of the information, willingness to recommend the way statins were discussed with others, and desirability of using the process of sharing information in future decisions. The questionnaire also included 14 knowledge questions to assess patient understanding of the relative merits of using or not using statins. Nine of these questions were addressed in the decision aid; 5 were not. At 3 months, surveys were mailed to patients to determine whether they were taking statins and, using a single-question,9 whether they had missed any doses in the last week. Nonrespondents were telephoned.
Patient Education and Decision-making Results
- Participants receiving either the decision aid or the control pamphlet scored similarly on questions irrelevant to the statin choice.
- Patients allocated to receive the interventions from their clinicians during the visit achieved better knowledge scores when using the decision aid than when using the control pamphlet.
- Patients allocated to receive the interventions from the clinicians during the visit were most accurate when reporting the relevant cardiovascular risk without statins when using the decision aid than when using the control pamphlet.
- Participants receiving the decision aid were more likely to accurately estimate the potential absolute risk reduction afforded by statin use than participants receiving the control pamphlet.
- Compared with the control group, the decision aid group had significantly less postvisit decisional conflict.
- Participants using the decision aid thought they were better informed about the options than did participants using the control pamphlet.
Statin Therapy Starts
- Among participants not receiving statin therapy at baseline, 7 (30%) of 23 in the decision aid group (6 of whom received the decision aid from their clinician during the visit) and 4 (21%) of 19 in the control group decided to start statin therapy immediately after the visit.
- Eight of these starts occurred among participants with 10-year cardiovascular risk greater than 15%.
- Of the 3 starts in the group with cardiovascular risk less than 15%, 2 occurred in the control group.
- At 3 months, 9 (39%) of 23 participants in the intervention group and 6 (32%) of 19 participants in the control group had started statin therapy (OR, 1.5; 95% CI, 0.3-6.8).
- Two of 4 patients with interim starts received Statin Choice from the clinician during the visit.
Adherence At 3 Months
At 3 months, 33 (63%) of the 52 participants in the decision aid treatment arm and 29 (63%) of the 46 participants in the control treatment arm reported taking statins (OR, 1.4; 95% CI, 0.8-2.4). Overall, there was no difference in adherence to patient choice at 3 months (analysis adjusted by sex, cardiovascular risk, and number of medications; OR, 1.9; 95% CI, 0.4-9.8). Of those patients taking statins at 3 months, 2 of 33 participants in the decision aid group reported missing 1 dose or more in the last week compared with 6 of 29 participants in the control group (OR for adherence, 3.4; 95% CI, 1.5-7.5).
Unanswered Questions And Future Research
Elements in the agenda for future research include evaluation of the role of decision aids in chronic conditions requiring decision revisions over time, testing Statin Choice in primary care and with less educated patients, use of multiple measures of adherence to medication regimen, estimation of the costs and burdens (eg, time) of implementing decision aids in practice, use of decision aids as tools to educate physicians-in-training to better enhance patient-clinician communication and decision making, and development of decisional quality as an outcome of clinical trials and as a measure for quality of care.
Commentary
It is significant – and the authors are careful to point out – that this clinical trial was undertaken by the same folks that developed the Statin Choice Tool being evaluated.
Further, it is also significant that the compliance rates were calculated from the patient’s self-report, a notoriously unreliable data source. Again, the authors point this out, specifically suggesting that future studies look at the “use of multiple measures of adherence to medication regimen.”
Otherwise, the study indicates that the statin choice tool is promising as a decision-making aid but far from proven. Perhaps I’ve grown cynical about compliance research, but I find, because self-report was the only data source, little compelling evidence of improved adherence although there is reason for cautious optimism.
I will be interested in the results of future trials, especially if they more rigorously assess the tool’s effect on patient compliance.