The Rise of Patient-Centered Outcomes Research

Article Outline

• Copyright

The newly formed US Patient-Centered Outcomes Research Institute (PCORI) has been tasked with “helping people make informed health care decisions and improving health care delivery and outcomes by producing and promoting high integrity, evidence-based information that comes from research guided by patients, caregivers, and the broader health care community.” PCORI is currently embarking on funding its first set of pilot research projects, on which it will spend close to $13 million in 2012. In this issue of Health Outcomes Research in Medicine, we are indeed reminded of the rising importance of patient-centered outcomes research. Patient-centered outcomes research is diverse and has to address a large range of issues varying from stakeholder involvement to health care quality, efficiency, and delivery, and our current issue certainly reflects this mixture. With a variety of contributions, we address such diverse topics as patient satisfaction with different types of providers, health-related quality of life (HRQoL) in head and neck cancer, the importance of appropriate outcome measures both in economic evaluations and in safety database studies, and real-world treatment patterns in patients with anxiety disorders.

Understanding which factors influence provider selection is important as we seek to better understand what matters to patients and how this may play out in health care delivery reform. As Adolph and colleagues observe in their Internet survey, the waiting time for visit appointment and time spent with the doctor were both associated with patient satisfaction with the physician. This was more important to this set of responders than whether they saw a primary care or specialty provider.

Evaluating HRQoL is a critical element in assessing what matters to patients, and can provide physicians, patients, and decision-makers with new insights into treatment priorities. Szabo and colleagues elicited preferences for standardized head and neck cancer health states in a sample of the Canadian general public. Unsurprisingly, all health states associated with head and neck cancer had substantially decreased preferences compared with full health, but the authors also observe that the effect on HRQoL is greater than previously reported. This is an important insight into this particular type of cancer that can be devastating to patients.

Choosing appropriate outcome measures to evaluate new treatments is critical, and nowhere more than in oncology, where newer, targeted, but also more expensive treatments are coming to market. As Davies and colleagues observe using the example from a recent trial in metastatic melanoma, the use of the difference in means (the measures needed for economic evaluations) and the median survival gains (which are the measures commonly reported in clinical studies) in outcome measures for cancer therapies may provide quite different results. This could lead to very different decisions about reimbursement for these treatments.

In the safety arena, the definition of outcome measures is a major challenge, as Stang et al report in this issue. While database studies using administrative claims and electronic health records are being used to provide critical information on medical product safety, Stang et al found that “health outcomes of interest” (HOI) were far from well defined in the literature. Reporting on work done for the private–public partnership OMOP (Observational Medical Outcomes Partnership), they call for a more transparent reporting of HOI definitions and warn that the approach of many drug safety studies to use one database and one HOI definition at one moment in time may not be optimal for medical product safety surveillance. It is likely that a broader process of defining events that can be applied across multiple databases and over time as new products enter the market will be more appropriate.

Patient-centered research is also about understanding what is happening in the real world. Berger and colleagues examined the patterns of health care utilization and costs in patients with generalized anxiety disorder initiating treatment with benzodiazepine anxiolytics as add-on therapy using a large US claims database. While they found that most patients appropriately initiated this treatment for a short duration, 13% of patients received more than 90 days of therapy. This treatment pattern warrants further research because of the risk of dependency and side effects associated with this medication.

Our issue highlights the diversity of topics relevant to patient-centered outcomes research, and we expect to see more such contributions in future issues, as this research is critical in supporting and improving health care quality, efficiency, and delivery.