Introduction to Outcome Analysis and Office-Based Clinical Research
Introduction to Outcome Analysis and Office-Based Clinical Research
Richard M. Rosenfeld
The American psychiatrist Philip Bonnet once noted, “The patient is always the ultimate source of knowledge.” Consequently, every clinical encounter has the potential to improve our knowledge of disease and to offer new insights regarding optimal management. The word potential is emphasized, because knowledge results only from systematic and standardized observation recorded in an equally compulsive manner. In contrast, anecdotal and biased information, no matter how plentiful, cannot produce meaningful insights. As William Osler observed, “The value of experience is not in seeing much, but in seeing wisely.”
In an era of evidence-based medicine, the purpose of each clinical encounter should be to gather data in a standardized way to facilitate inference. Inference is the process of moving from observations to generalizations so that clinical pathways and practice guidelines can be formulated. Systematic data from routine clinical encounters are the heart and soul of outcomes research, which is focused on real-life treatment endpoints that are meaningful to patients and physicians. This chapter explores the importance and scope of outcomes research and emphasizes the role of the physician’s office as the laboratory of the future for this new and exciting approach to clinical investigation.
RISE AND FALL OF RANDOMIZED TRIALS
Medical research has traditionally been dominated by the noble goals of objectivity and scientific purity. The fruit of this obsession was the randomized controlled trial (RCT), introduced soon after World War II. Until that time clinical research was largely observational, making it difficult to separate the effects of patient selection for a new treatment from the effects of the treatment itself. An RCT is a true experimental study, because conditions are strictly controlled by the investigator, who randomly assigns patients to treatment or control groups. Further, investigators and patients often are unaware of their group assignment (they are blinded), so that preconceived expectations about treatment do not interfere with outcome assessment.
Unfortunately, the price of scientific purity was increasing separation from the realities of everyday medical practice. Interpretation of results from an RCT is limited by selective admission criteria, rigorous therapeutic protocols to ensure compliance, and narrowly defined outcome measures. The treatment endpoints in RCTs are chosen for their objectivity and are not necessarily important to patients or practicing physicians. RCTs are expensive and not always ethical, particularly for surgical therapies. Finally, randomized trials can become outdated quickly and often do not address issues of effectiveness and information dissemination.
TABLE 54-1. Comparison of randomized clinical trials and outcomes studies
Characteristic
Randomized clinical trial
Outcomes study
Goal
Establish cause and effect
Demonstrate relationships
Level of investigator control
Experimental
Observational
Treatment allocation
Random assignment
Routine clinical care
Study design
Parallel groups
Longitudinal cohort
Patient selection criteria
Restrictive
Broad
Typical setting
Hospital or university based
Community based
Endpoint definitions
Objective health status
Subjective quality of life
Endpoint assessment
Masked (blinded)
Unmasked
Statistical analysis
Comparison of groups
Multivariate regression
Potential for bias
Potentially low
Potentially high
Beginning in the mid-1980s, a quiet revolution in clinical medicine began with the cryptic title of outcomes research. In some ways the antithesis of RCTs, outcomes researchers tried to show how medical practice affects subjective quality of life and well-being by observing large numbers of patients treated in real-life practice settings under ordinary conditions (Table 54-1). Outcomes research has created new opportunities for individual practitioners to become highly effective clinical researchers in their own everyday practice. Sites of everyday clinical care— physicians’ offices, hospital clinics, inpatient wards, operating rooms, and emergency rooms—can become clinical research laboratories when data acquisition is efficient and standardized.
WHAT IS OUTCOMES RESEARCH?
“We are drowning in information but starved for knowledge,” noted the American business writer John Naisbitt. Similarly, Paul Ellwood noted in 1988 that our inability to measure and understand healthcare outcomes resulted in uninformed patients, skeptical payers, frustrated physicians, and besieged healthcare executives. In response, he proposed a technology for collaborative action called outcomes management. According to Ellwood, “Outcomes management is a technology of patient experience designed to help patients, payers, and providers make rational medical care related choices based on better insight into the effect of these choices on the patient’s life. Outcomes management consists of a common patient-understood language of health outcomes; a national data base containing information and analysis on clinical, financial, and health outcomes that estimates as best as we can the relation between medical interventions and health outcomes, as well as the relation between health outcomes and money; and an opportunity for each decision-maker to have access to the analyses that are relevant to the choices they must make.”
To achieve these goals, Ellwood proposed the following activities.
Systematic measurement of the functioning and well-being of patients, along with disease-specific clinical outcomes, at appropriate time intervals
Pooling of clinical and outcome data on a massive scale
Analysis and dissemination of results from the database segment most appropriate to the concerns of each decision maker
Greater reliance on practice guidelines by physicians when selecting appropriate interventions
Since 1989, the Agency for Health Care Policy and Research (now called the Agency for Healthcare Research and Quality) has funded special projects known as Patient Outcomes Research Teams. These large and complex 5-year undertakings, with average annual budgets of $1 million, represent the leading edge of outcomes research methodology. Emphasis is placed on outcomes that patients understand and care about, such as quality of life, functional capacity, symptom relief, and cost (in contrast to physiologic measures and parameters that focus more on organs than their owners). Because the vocabulary of outcomes studies may be unfamiliar to clinicians, definitions of commonly encountered terms are listed in Table 54-2.
When the American Academy of Otolaryngology-Head and Neck Surgery Foundation introduced its Health Services Research Grant in 1996, the purpose was to foster research that would improve the effectiveness and appropriateness of medical practice. Projects supported under the program would develop and disseminate scientific information on the effects of otolaryngology services and procedures on patient survival, health status, functional capacity, and quality of life. The three main categories of outcomes research identified—patient-based studies, record-based studies, and process assessment—illustrate the scope of outcomes research (Table 54-3).
PRIMER ON OUTCOMES RESEARCH
An uncanny aspect of outcomes research is the ease with which meaningless data are produced. This occurs because outcomes studies are observational, in contrast to the experimental method that underlies RCTs. From an evolutionary perspective, outcomes research is a giant step backward from RCTs, as are most other office-based observational studies. Nonetheless, bad RCTs are as feasible as bad outcomes studies. The latter, however, are much easier to produce.
TABLE 54-2. Definition of common terms used in clinical and outcomes research
Term
Definition
Outcomes
Measurable events and observations that are presumed to occur in part because of structure and process of medical care
Structure of care
Stable elements of the health system, including provider qualifications, administrative organization, and type of facility
Process of care
What happens in the medical interaction, including technical and interpersonal skills of physician and other providers
Quality of care
The difference between efficacy and effectiveness that can be attributed to health care providers
Efficacy
How a treatment works in ideal circumstances, when delivered to selected patients by skilled providers
Effectiveness
How a treatment works under ordinary conditions when delivered to a typical patient by an average practitioner
Health status
A measure that includes genetic characteristics, physiologic status, functional status, mental condition, and health potential (longevity and prognosis)
Functional status
An objective measure of the degree to which a person is able to function physically, emotionally, and socially in daily life
Quality of life
A subjective measure of the degree to which persons perceive themselves as able to function
TABLE 54-3. Scope of outcomes research
Patient-based outcomes research
Creation and validation of health-related quality of life measures
Creation and validation of disease-specific clinical severity scales
Observational studies of treatment effectiveness
Record-based outcomes research
Analysis of administrative and financial databases
Regional variations in practice patterns and outcomes
Meta-analysis, decision analysis, or cost-effectiveness studies
Analysis of national data sets or population-based surveys
Process assessment
Continuous quality improvement
Patient satisfaction with healthcare services
Development of clinical practice guidelines
Editorial peer review
Many outcomes studies begin with observation of a large group of patients. An attempt is then made to draw conclusions about the effectiveness of various treatments by measuring health status and quality of life. Although it is tempting to attribute any observed differences directly to treatment variations, a myriad of other explanations arise when randomization and control groups are absent. Some of the more common explanations are listed in Table 54-4, including chance, confounding, and a host of biases. The effect of these factors on patient-based, records-based, and process-based outcomes studies is described in Table 54-5. Whereas the average clinician need not memorize this information, he or she must at least be aware that accepting observational research results at face value often is unjustified.
Only gold members can continue reading. Log In or Register to continue