In a recent report, the Institute of Medicine discomfited health care professionals by placing the medical profession ten years behind other high-risk industries in protecting the population it serves. Although no one disputes the importance of reducing medical errors, there is little agreement on how to do it (1). Psychiatry's disinclination to provide evidence to justify its treatment effectiveness may have rendered it vulnerable to the forces of health care reform (2). The best-practices movement in medicine—a potential mechanism to protect excellence in practice—has much to offer practitioners who are striving to demonstrate their value in a commodity-driven market. In this article, we review the current status of the best-practices movement in general medicine and in psychiatry.
The ability to identify and implement best practices in industry has differentiated the successful organization from the unsuccessful organization. Because longevity and financial viability are determined by the quality of customer services and products, there has been a significant increase in comparative techniques that recognize the best and worst performers in the commercial world. Best practices in this context is defined as the measurement, benchmarking, and identification of processes that result in better outcomes (3). The steps include identifying the benchmarked activity or product; identifying potential benchmarking candidates—for example, other organizations similar in size or product offerings; comparing data; and establishing goals and activities to improve the benchmarked activity or product (4). In recent years, the best-practices movement has filtered into general medicine and to some extent mental health care, perhaps in part because of pressure from employers —payers—who have witnessed the value of this process.
Traditionally, best practices in the medical industry have been defined by a guild—for example, if a physician orders it, it is best; by individuals of a particular theoretical orientation —for example, psychoanalysts versus behaviorists; or by proponents of a particular program—for example, case management. More recently, consensus guidelines, such as those developed for the treatment of adult depression by the Agency for Healthcare Policy and Research (now the Agency for Healthcare Research and Quality) and the American Psychiatric Association, have delineated the boundaries for best practices specific to disease states such as depression or schizophrenia.
Best practices can also be identified through a continuum of clinical trials, multisite effectiveness studies, and testing of single-setting interventions. Unfortunately, only a few clinicians have been willing or able to apply this model to practice. Finally, identification of best practices can be achieved on-site by systematically measuring similar practices and carefully scrutinizing variability.
Outliers can be identified by measuring clinical processes and outcomes, resulting in a "drill down" that can determine the source of variation and pinpoint problems in service delivery or effectiveness. For example, one of five clinics may exhibit poor performance in treating patients who have major depression. Analysts may choose to drill down to the individual program or provider level or examine groups of patients—for example, adolescents—to discover the weak link in the care process, using indicators of success from the other four clinics as well as outside authoritative sources. Similarly, the drill-down process may facilitate identification of providers who excel in the treatment of a particular condition or patient.
In recent years, quality-focused programs in medical care have adopted industry's best-practices approach for improving process and outcome. These efforts have occurred nationally through accrediting organizations and cooperative partnerships among professional groups, and they have occurred across large clinical systems.
On the national front, Quality Compass 2000, a database of managed care information compiled by the National Committee for Quality Assurance (NCQA), includes performance information for 466 health plans (5). NCQA has found that health plans that consistently measure and publicly report their rates on adolescent immunizations, cholesterol screenings, and chicken pox vaccinations, to name a few, improve over time on these indicators of the Health Plan Employer Data and Information Set (HEDIS) (6). Furthermore, plans that score in the top 25 percent on most of the HEDIS indicators are rated higher on consumer measures of satisfaction, access, customer service, and overall performance. Similar performance measure projects by the Health Care Financing Administration—operating through state peer review organizations—and the Joint Commission on Accreditation of Healthcare Organizations, are currently under way.
Cooperative partnerships among professional organizations have also contributed to the identification of best practices in the medical arena. Studies have compared the performance of specialty providers in treating such conditions as low back pain, asthma, heart disease, and diabetes. By demonstrating widespread variation in practices, these efforts have created an environment of accountability and spurred improvement activities that will ultimately benefit the consumer.
Finally, numerous clinical systems have collaborated to design and employ best practices. These systems include hospitals with the same parent company, outpatient specialty clinics within a particular health plan, or a delivery system serving a specific population. For example, Burstin and colleagues (7) recently compared five Harvard-affiliated emergency departments on process-of-care guidelines for six chief complaints. They found that emergency department directors were able to improve patient care when they received comparative feedback on their department's performance and received assistance in designing quality improvement strategies.
The Department of Veterans Affairs (VA) Healthcare System recently evaluated the performance of its medical centers on 27 performance measures and, as part of its quality improvement activities, reported the results to clinical managers and each of the 22 VA integrated services network directors across the country (8).
Although psychiatry has lagged behind the medical industry in the search for best practices, there is evidence of progress. For example, last year NCQA introduced three indicators for the treatment of depression. After a year of data collection, it appears that at least one measure—follow-up with a provider during the first 12 weeks of acute treatment—shows considerable regional variability. Furthermore, larger health plans have higher compliance rates than smaller plans (9). It remains to be determined whether this indicator will differentiate providers with better-than-average care or whether improvement will occur over multiple reporting periods, as was the case in general medicine.
Collaborative partnerships in which providers receive comparative data on processes and outcomes of psychiatric care have also been forming. American Psychiatric Association- NetOutcomes Quality Care 2000 (10) is a joint project developed by the American Psychiatric Association (APA) and the Center for Outcomes Research and Effectiveness of the University of Arkansas for Medical Sciences. Through an Internet-based treatment outcomes system, participating APA members can benchmark the clinical outcomes of their patients with depression and determine whether their performance compares favorably with that of other providers.
A similar project, coordinated by the department of psychiatry at the University of Cincinnati College of Medicine (11), compares process and outcomes of care for behavioral health group practices across the country. Data from the project suggest that the proliferation of consensus and evidence-based guidelines may contribute to improved patient care. However, because of confidentiality constraints, the authors were unable to examine which factors distinguished the innovative from the average provider settings.
Finally, implementation of best practices from research is occurring at individual sites, such as the Center for Behavioral Health in Bloomington, Indiana, where clinicians are being trained in empirically supported treatments and patient outcomes are benchmarked against results of clinical trials (12).
Perhaps the most advanced capabilities in, and the most systematic approach to, benchmarking can be found in the VA. Its extensive service utilization and pharmacy databases allow comparisons across hospitals and specialty clinics, resulting in multiple reports of practice variations in lengths of stay for psychiatric inpatient admissions, prescriptions for antidepressant and antipsychotic medications, and extent and type of psychosocial interventions for veterans with psychiatric diagnoses in primary and specialty care clinics (13).
More recently, the VA has instituted the Quality Enhancement Research Initiative. The initiative addresses several areas of concern for veterans, including mental health. The project translates research findings into clinical practice through dissemination and implementation of findings and guidelines, development of toolkits and feedback mechanisms, and other intervention materials. The overarching goal is to reduce inappropriate variations in treatment.
Other initiatives include the benchmarking efforts of the National Association of State Mental Health Program Directors, which has released a comprehensive set of performance and outcomes indicators for public mental health programs.
Despite these developments, our knowledge of best practices in psychiatry remains limited. We have ventured into provider profiling, yet we have minimal information about what constitutes the best in our field. We have advocated outcomes measurement for years, yet we have published few studies linking effectiveness to specific care processes. We have documented variability in provider practices, but we have avoided the more intensive drill-down process to determine what works well for which patients.
Mental health professionals acknowledge that the best-practices movement in psychiatry faces multiple barriers. They include disagreement about the most appropriate process measures—for example, number of psychotherapy sessions and dosages of psychotropic medications—and lack of consensus about the most valid outcomes metrics—for example, improvement in symptoms, functioning, and quality of life and personal growth. A common concern among clinicians is that human behavior is complex and therefore difficult to measure or quantify. Consequently, many of them have resisted data collection and the development of critical information systems and analytic tools that would facilitate benchmarking and discovery of best practices.
Despite these obstructions, a few innovative organizations and delivery systems are developing unique approaches to quality measurement and improvement that have the potential to change the way care is administered. Psychiatry has much to offer the best-practices movement in preserving human uniqueness. We should not let our skepticism about the ability to capture the human experience through outcomes measurement deter our efforts to improve practice.
It is imperative that we keep pace with advances in the best-practices movement so as not to be left behind. Our ability to meet this challenge will influence our field's place in the new form of health care that is unfolding. It behooves us to recognize the strategies of successful business leaders in the corporate world and adopt a proactive approach to improving the quality and outcomes of treatment through measurement of our own practices and benchmarking with others, even if that process is conducted on a small scale. Otherwise, our quest for excellence will be limited.
Dr. Kramer is associate director of the Center for Outcomes Research and Effectiveness at the University of Arkansas for Medical Sciences, 5800 West 10th Street, Suite 605, Little Rock, Arkansas 72204 (e-mail, firstname.lastname@example.org). Dr. Glazer, who is editor of this column, is associate clinical professor of psychiatry at Harvard Medical School, Massachusetts General Hospital, in Boston.