What is new?
- •
Researchers all-too-rarely consult patients about their values and preferences in terms of the research agenda.
- •
This is the first report that shows patients' opinion about diabetes trial design and outcomes.
- •
Most patients prefer trials that measure patient-important outcomes and provide pragmatic/practical inferences.
The Patient Protection and Affordable Care Act of 2010, the health care reform law, incorporates a strong agenda supporting the conduct of comparative effectiveness research. The law mandates the constitution and funding of an independent nonprofit organization, the Patient-Centered Outcomes Research Institute, to promote the conduct and dissemination of this research. This form of research has to be most informative to decision makers, a hallmark of practical trials [1]. Furthermore, research should directly address the effects of health care on outcomes that matter to patients (patient-important outcomes, such as death, stroke, myocardial infarction, pain, quality of life, and patient satisfaction). To our knowledge, however, the preferences of patients regarding this research remain unknown.
Mechanistic trials, also known as explanatory trials, refer to experiments that address a biological relationship and provide “proof of concept” evidence that helps researchers make decisions about further research. As a result, these trials often take the participants, use interventions, and measure the physiological or disease-specific objective outcomes that allow researchers to draw mechanistic inferences. On the other hand, practical trials refer to experiments that provide evidence about the effectiveness of interventions that helps either clinician–patient dyads or policymakers make informed decisions. Their structure will reflect this goal (e.g., for clinician–patient dyads, practical trials may need to enroll only highly compliant patients, whereas for public health decision making, “all comers” may be optimal). Common to all practical trials is the measure of patient-important outcomes [1].
Judging by the proportion of public funding that goes to mechanistic research and by the extensive press coverage of early-stage investigations, it may appear as if the public is mostly interested in discovery of new treatments and their effect on mechanisms of health and disease. We have recently measured the extent to which clinical trials in diabetes emphasize surrogate end points. Only 20% of published diabetes trials designated their primary end points as outcomes that are important to patients, that is, outcomes that affect the way patients feel, function, or survive [2]. A review of trial registries that address ongoing yet-to-be published trials showed that only one in five future trials will measure patient-important outcomes as primary end points and less than one in two will assess them at all [3]. Thus, we should not expect a change in this trend in future trials.
In diabetes, hemoglobin A1c (HbA1c) is considered by many patients, clinicians, and quality improvement specialists to be a measure of diabetes metabolic control that signals improvement in patient outcomes. Contrary to this expectation, recent trials in patients with type 2 diabetes have found that interventions that effectively lower this marker fail to improve patient-important outcomes [4], [5]. Thus, to be confident that our interventions positively influence outcomes important to patients, trials must directly measure those outcomes.
We conducted a cross-sectional survey of patients with diabetes to ascertain their preferences on how clinical trials should be conducted in terms of study design and end points. Knowing the needs of the primary user and beneficiary of research, the patient, should guide subsequent trial design and thus ensure the evidence from comparative effectiveness research will optimally address the needs of the patient.