Real-world Imperative of Outcomes Research

Outcomes research, with its focus on the experience of patients, is evolving as a core component of health care delivery.^{1 - 2} Pioneers such as Nightingale, Codman, and Wennberg established the promise of focusing on the end results of health care.^{3 - 5} However, as a still-emerging field, there is a need for clarity about the definition, scope, and goals of outcomes research (Figure).

The Patient Protection and Affordable Care Act characterizes outcomes research as a field that determines how “diseases, disorders, and other health conditions can effectively and appropriately be prevented, diagnosed, treated, monitored, and managed. . . . ”¹ Stated another way, outcomes research is scientific investigation that generates knowledge to guide health care decisions and promote optimal results from preventive, diagnostic, prognostic, and therapeutic health care strategies. The goal is to increase the likelihood that patients achieve the outcomes they desire through better information, better decisions, and better health care delivery. Toward this end, outcomes research uses a diverse array of study designs and data sources. It draws from clinical medicine, statistics, informatics, epidemiology, and the social sciences.

Outcomes research derives directly from real-world practice and policy. It seeks to determine what is being achieved for patients and how clinicians can do better. The emphasis is on outcomes that patients experience, not on measurements that are surrogates for what may happen (eg, biomarkers) because those measures often fail to predict the effect of interventions.⁶ Outcomes research purposely directs attention to effects in typical patients across a spectrum of venues with emphasis on characteristics that may influence outcomes. Thus, it generates information about treatment patterns, risks, benefits, and costs that can be tailored to individuals.

The questions and priorities of outcomes research ideally derive from partnerships with those (patients, clinicians, purchasers, and policy makers) who may use the results. Such relationships can bridge the gap that often exists between the published research and the information needs of decision makers.⁷

Comparative Effectiveness of Clinical Strategies in Real-world Settings. Patients, clinicians, and policy makers face immense gaps in knowledge that impede informed decision making. This domain is concerned with filling those gaps by determining which strategies are most effective—whether overall, for selected patients or populations, or when delivered by particular types of hospitals and practitioners. The effectiveness of a strategy may vary based on a patient's biological, demographic, clinical, psychological, or social characteristics. Moreover, context matters because the effects of complex interventions often depend on how and where a strategy is applied.

Perspective of the Patient. The second domain involves generating knowledge that promotes the alignment of health care decisions with patient preferences, values, and goals. Although studies may show an average advantage of a particular approach, the evidence rarely guarantees a result or mandates a choice for an individual. Each patient's preferences must guide the decision. For example, facing a choice of cancer treatments, some patients value length of survival most, whereas others prefer a higher quality of life.⁸ Research-tested tools and strategies are needed to help maintain the centrality of the patient perspective in clinical decision making about short- and long-term outcomes.

System Performance. The third domain focuses on health system performance, with emphasis on optimally implementing strategies to prevent, diagnose, and treat disease; eliminating waste and ensuring equity in access and care; and achieving optimal return on health care investments. That emphasis promotes effectiveness, efficiency, equity, safety, patient-centeredness, and timeliness—the features of a high-quality health care system as defined by the Institute of Medicine.⁹

Within the 3 domains, 3 overlapping types of investigation are conducted: discovery, application, and surveillance.

Discovery science, rather than being merely an extension of basic laboratory or clinical science, yields novel insights from observation or experimentation. It reveals information about the determinants of outcomes, methods to assess risk, strategies that improve outcomes, and methods that make the research easier to perform and the inferences more valid and useful. For example, the discoveries of small-area practice variation by Wennberg et al⁴ were novel observations that led to a new understanding of the determinants of clinical decisions.

Application studies focus on the development, use, and evaluation of interventions that can improve medicine and health care by identifying and assessing tools and strategies that yield tangible opportunities for patient-centered change. Implementation science and improvement science are relevant to this effort. Application studies can also examine workflow in typical practice, cognitive biases that present barriers to using the best evidence, and patient-clinician interactions.

Surveillance is a critical component that can reveal patterns and trends in care, identify opportunities for improvement, and establish accountability for the results of these efforts. For example, Medicare data have revealed that over time, declines in length of hospital stays for patients with heart failure were associated with increased readmission rates.¹⁰

In the short term, outcomes research faces 3 critical challenges. The first is to build partnerships with those who will use the knowledge that the investigations yield. Just as investigators in community-based participatory research respect and collaborate with members of the communities they study, so are outcomes researchers evolving to be responsive to those whom the research can assist.

The second challenge is to produce studies that are relevant, timely, and methodologically rigorous by advancing methods and study design and expanding the number of capable investigators within the outcomes research community.

Third, the information-sharing infrastructure and technological tools must be improved to support the conduct of high-quality research through better use of existing data and more efficient methods of collecting and analyzing new data. Existing data, including electronic health record data that protect patient privacy, should be made widely available to expand research capacity and facilitate independent validation of findings.

A commitment to the next generation of outcomes research is a vital component of efforts to advance the science of clinical care and the quality of health care systems. Learning from practice and applying those lessons to individual clinical decisions and system planning are essential. Outcomes research is the vehicle by which this learning can occur and, in the process, may fuel innovation in care driven by advances in technology, information science, and systems.

The application of this science to health care practice will require leadership, vision, and the courage to overcome resistance to change. The alternative is to continue to lack critical information about many interventions, to be uncertain about the risks and benefits for individual patients, and to seek solutions to the challenges of medicine without sufficient evidence to inform wise choices. Outcomes research as an integrated and integral part of the health care delivery system can support better decisions for patients and generate higher performance for the health care system.