Portal de Periódicos da CAPES

Presidential address: Sony, Porsche, and vascular surgery in the 21st century

1997; Elsevier BV; Volume: 25; Issue: 2 Linguagem: Inglês

10.1016/s0741-5214(97)70343-4

1097-6809

Jerry Goldstone,

Healthcare Policy and Management

“It was the best of times; it was the worst of times”1Dickens C. A tale of two cities.in: Dell Publishing, New York1963: 11Google Scholar A Tale of Two Cities Charles Dickens, 1859 To paraphrase Dickens, it is the best of times, it is the worst of times. Dickens, of course, was writing about the French Revolution of 1789. I am referring to the American, and to some extent worldwide, health care revolution of the 1990s. Never has the health care system been equipped with such extraordinary and exciting physical and intellectual resources and abilities to prevent and treat disease than it is now—surely the best of times. Yet, not in recent memory, and perhaps never has there been so much attention paid and action taken to control health care expenditures. The speed and magnitude of the changes have left most health care organizations and providers reeling, uncertain, and afraid—surely the worst of times in the opinion of many. It is beyond rational debate that health care costs have grown well beyond society's desire or ability to pay for them. U.S. health care costs were $949.4 billion in 1994 and have increased from 5% of the gross domestic product (GDP) in 1960 to nearly 14% of the GDP in 1994, and if the same rate of increase is sustained, it will reach 30% of the GDP by 2020. And yet, as profound as cost containment has been, health care expenditures have not declined. It is just the rate of increase in expenditures that has declined, from 12.9% in 1980, to 7% in 1993, to 6.4% in 1994. Nevertheless, Medicare spending increased 11.8% in 1994, compared with only a 4% increase for private health insurance plans. This is especially ominous for vascular surgeons who typically treat large numbers of Medicare beneficiaries. One way or another, sooner or later, health care costs will have to be stabilized and controlled and at some arbitrary level, for example, a constant percent of the GDP. Efforts to produce continuous savings of the magnitude desired will require logistic and administrative efficiencies, as well as control of the intensity and volume of health services. It is not yet clear whether health care costs have been controlled, but it can be accepted that they will be controlled, and we will all be affected. So far, most of the attention and effort in health care reform by administrators, payers, and government has focused on the issues of cost containment and access to care rather than on quality of care.2Chassin MR. The missing ingredient in health reform, quality of care.JAMA. 1993; 270: 377-378Crossref PubMed Google Scholar Chelimsky3Chemlimsky E. The political debate about health care: are we losing sight of quality?.Science. 1993; 262: 525-528Crossref PubMed Scopus (13) Google Scholar reviewed and analyzed the number of articles in the New York Times that primarily addressed the factors of medical cost, access to health services, and quality of care from 1989 to 1993. She found 302 articles on cost, 58 on access, and only 11 on quality. For physicians, who have traditionally borne the responsibility for clinical quality, an integral element of the health care revolution is a genuine concern about the quality of care. How can physicians and other clinicians constantly achieve high-quality care while meeting mandated and sometimes draconian cost restraints? To answer this question, it is first necessary to examine the relationship between quality and cost in health care. The relationship between quality and cost in health care is very complex and poorly understood. According to classic economic theory, there is a direct relationship between quality and cost—as quality goes up, costs go up. If it is assumed that this is true in health care, and I believe that most physicians do, then it follows that efforts to contain costs will threaten quality. Examples of cost-driven low-quality care are known to all of us. But if this assumption is wrong, if the relationship between cost and quality is negative or indirect, cost containment and quality improvement could be complementary, and it therefore might be possible to improve quality at the same time that costs are reduced. Quality is a term widely used in many facets of contemporary life. It implies excellence, fineness, or superiority. For example, Sony is a company universally known and recognized for innovation and manufacturing excellent, that is, high-quality electronic products. What exactly is this quality? Each person probably has a different idea, depending on what products he or she has, what they are used for, and how they are expected to work. In business jargon, if a product satisfies one's expectations, it is said to possess quality. Thus quality is meeting or exceeding customers' expectations. But what is quality as it applies to health care? It is a multidimensional construct that is extremely hard to define, measure, and observe.4Laffel GL Berwick DM. Quality health care.JAMA. 1993; 270: 254-255Crossref PubMed Scopus (16) Google Scholar Surgeons instinctively think they know what quality is: low stroke rates after carotid endarterectomy; high graft patency and limb salvage rates after lower extremity bypass; high survival rates after aortic aneurysm repair. But these are results or outcomes, important to be sure, and a part of quality but not synonymous with a contemporary definition of health care quality. Outcomes describe the application of medical science, knowledge, and technology to the diagnosis and treatment of disease. In short, this refers to the difference between a patient's current and future health status attributed to the medical care provided. In this construct, health care quality has three components: interpersonal, amenities, and technical. Briefly, the interpersonal component refers to the interactions between patient and provider and the perception by the patient and his or her family of the amount of responsiveness and attentiveness on the part of the provider. This is the “art” of medicine. Amenities of care include such things as the convenience and appeal of the health care facility. Managed care organizations are paying more and more attention to this aspect. The technical component is roughly equivalent to medical outcomes. Each of these three components can be assessed separately in terms of quality. Patients are not in a position to evaluate the technical aspects, but they are certainly capable of doing this for the other two components. Implicit in this concept is that of patient expectations. Just like the consumer expectations of Sony products discussed earlier, patients have expectations about the health care they receive and they make quality judgments about it. Quality in health care can be looked at in another way, by considering it to consist of three different components: structure, process, and outcome.5Donabedian A. The quality of care. How can it be assessed?.JAMA. 1988; 260: 1743-1748Crossref PubMed Scopus (4013) Google Scholar Structure involves the relatively fixed characteristics of the medical delivery system, such as number, types and qualifications of health care providers, and facilities, including equipment (CT and MRI scanners, for example). The process component is what is actually done to or for the patient—the application of drugs, medical procedures, etc. It too can further be regarded as consisting of inputs leading to actions resulting in outputs (outcomes). Outcomes, as defined earlier, are the changes in a patient's health status that can be attributed to the antecedent medical care. If one then considers the practice of medicine as a process, or rather an infinite series of processes, each one of which results in an output, one can then consider the quality of these processes as well as the quality of the results or outputs they produce. This model of quality emphasizes that quality is not the output (results) alone. Another way in which quality can be categorized and evaluated is in terms of content or delivery. Content quality describes the technical component of medical care and is roughly equivalent to medical outcomes, whereas delivery quality describes patients' satisfaction with their health care experiences. Quality then becomes an individual's personal judgment or evaluation of an output and the personal interactions that take place as the output(s) is (are) delivered to the individual. It is based on an individual's (or society's) value system. It is a relative term with no fixed unit of measurement, and therefore it cannot be described in static terms, such as good or bad, but only as better or worse over time. And because each individual's culture, past personal experiences, and prejudices create expectations, quality again can be seen to involve meeting expectations. Quality can be judged to be acceptable when positive expectations are met. When expectations are consistently exceeded, then quality is judged to be excellent. Because values and judgments are part of quality and because they can change over time (for example, through education), quality gets redefined each time there is an interaction between an individual and an item or process for which the term is evaluated. Service quality or delivery of care is evaluated primarily by patients, but as a learned profession, medicine has reserved the right to judge its own content quality. Our profession sets its own expectations regarding medical outcomes using a variety of methods, including informal consensus, clinical trials and hypotheses, individual, institutional, and multi-institutional experiences, etc., as expressed in the medical literature or less formally. When discussions about quality of care occur, most physicians believe that the care they are providing and have always provided is of the highest or best quality. Both consumers, i.e., the public, and third party payers have become skeptical. Consequently, the profession's exclusive hold over medical quality content is being challenged and eroded as payers, patients, and others become more informed and begin to take a more active role in setting medical outcomes' expectations. The preceding comments have demonstrated the complexity of the term quality as it is applied to health care. The current health care environment demands that we as physicians expand our concept of quality as it relates to both outcome and delivery. It is useful in this regard to accept the concept of patients as customers, even though this seems far too business-like to many physicians. Nevertheless, it is an undeniable fact that health care is becoming ever and ever more business-like and physicians must accept this and adapt to it. Quality, with its expanded definition, is related to costs in several ways. In manufacturing, process failures (low quality) result in error, rework, and waste. A bad outcome must either be fixed or thrown away. Health care differs from manufacturing in that there isn't the luxury of throwing bad outcomes away, for they represent loss of life or functional ability. When the output (results) of a health care process (intervention) fails and does not meet quality expectations, complications and death result and additional medical resources are expended in an attempt to correct the situation. But either way, throwing it out or fixing a bad outcome, increases costs. Thus diminished quality leads directly to higher costs. These additional costs and resources are referred to as quality waste and represent process failures in medical care just as they do in manufacturing.6James BC. Quality improvement in the hospital: managing clinical processes.Internist (Berl). 1993; 34 (March): 11-13Google Scholar When operations are performed for inappropriate indications, and there is abundant evidence that this is common, the resources used that do not contribute positively to patients' health status, plus those required to treat the complications that inevitably occur from any series of operations, both represent quality waste. Repeating laboratory tests and x-ray examinations whose results are lost are other common but simple examples of wasted resources in health care. The second way quality relates to costs is through productivity or its corollary, inefficiency. There is considerable variation amongst physicians in practice patterns in terms of resources used to achieve the same clinical results. For example, some surgeons operate faster than others. When two surgeons get the same clinical results from an operation but one takes twice as long to perform the operation as the other, the increased costs of the slower surgeon represent low productivity. In the provider-at-risk environment that now exists, inefficient practice patterns are no longer acceptable. This has led to efforts by some management groups to introduce economic profiling of physicians for credentialling purposes. When total health resources are limited, as they certainly are today, wasted resources are not available for the needs of other patients and therefore do positive quality harm. This concept has also been used by some managed care organizations as justification for denying certain treatments to high-risk or elderly patients. Anecdotally, quality waste and low productivity are very commonly present in both hospital management and patterns of care. It is estimated that quality waste accounts for 25% to 40% of all hospital operating costs!7James BC. Implementing practice guidelines through clinical quality improvement.Frontiers of Health Services Management. 1993; 10: 3-37PubMed Google Scholar Quality waste is a useful concept because it provides a framework for seeking and eliminating waste and rework, thereby improving a process, which then leads to improved quality and lowered costs. Quality and costs are also intertwined in other ways. One of the most widely discussed in health care is expressed in the concepts of cost effectiveness, which have been well described in recent publications and will not be dealt with herein.8Wagner EH. The cost-quality relationship: do we always get what we pay for?.JAMA. 1994; 272: 1951-1952Crossref PubMed Scopus (5) Google Scholar, 9Eddy DM. Clinical decision making: from theory to practice. Rationing resources while improving quality.JAMA. 1994; 272: 817-824Crossref PubMed Scopus (68) Google Scholar From the foregoing, it is obvious that quality is intimately related to cost. But costs do not control quality, instead, quality controls costs. This concept has enormous implications for health care providers because it provides a framework for making things better in health care in spite of the severe spending limitations that have recently been introduced. Quality has become a critical issue in almost all segments of business, industry, and organizations. It is now recognized that error, waste, and duplication of work lead to higher costs and less customer satisfaction, that is, decreased quality. This is not good for business and it's not good for health care. Although some say that we should strive for perfection, this is an unrealistic objective and goal in health care. It will always be possible to improve quality. Surgeons have led the way for the medical profession through the traditional morbidity and mortality conferences that are an ingrained and integral component of surgical training and practice. Over the past 20 years, layered on top of this, has been an enormous and expensive effort, mandated and supported by organizations such as the Joint Commission on Accreditation of Health Care Organizations (JCAHO), collectively known as quality assurance. The customary approach to hospital quality assurance (QA) is to set standards; identify individual providers, clinicians, or hospitals that demonstrate unacceptable performance relative to those standards; and then take action to eliminate (i.e., improve) that performance. It is not my purpose to debate the merits of QA programs, but only to point out that their focus has been on thresholds or standards, most of which are inherently arbitrary in nature and define acceptable or unacceptable levels of performance or compliance.4Laffel GL Berwick DM. Quality health care.JAMA. 1993; 270: 254-255Crossref PubMed Scopus (16) Google Scholar As such, they can become an artificial quality floor or ceiling. Hospital-based QA activities are designed to identify those activities (and practitioners) who are above or below these floors or ceilings, and then try to bring them into compliance with the standards. This approach uses thresholds to establish a statistical tail, then concentrates improvement efforts within that tail. It is analogous to industrial reliance on inspection as a means of improving manufactured products. It is what Berwick described as “the practice of finding the bad apple.”10Berwick DM. Controlling variation in health care: a consultation from Walter Shewhart.Med Care. 1991; 29: 1212-1225Crossref PubMed Scopus (221) Google Scholar It embraces the philosophy of quality that is “good enough” rather than “the best possible.” It is inherently a punitive system that elicits fear and resentment, and its effects on overall quality are far from optimum (Table I). Table IQuality Assurance versus Continuous Quality ImprovementQACQIBased onStandards, thresholdsSpecificationsObjectiveOutcomeProcess and outcomeFocusStatistical tailWhole groupEffectJudgmentalEducationalPhilosophyGood enoughBest possible Open table in a new tab Another and better approach for improving quality, also borrowed from industry, encompasses the theory of continuous improvement. The concepts of this theory were formulated and espoused in the 1930s, most notably by Joseph M. Juran, Walter Shewhart, and W. Edwards Deming. Shewhart, an engineer at the Bell Laboratory and the father of industrial process control, introduced Deming, a mathematical physicist, to the use of statistical methods for quality control.10Berwick DM. Controlling variation in health care: a consultation from Walter Shewhart.Med Care. 1991; 29: 1212-1225Crossref PubMed Scopus (221) Google Scholar, 11Berwick DM. A primer on leading the improvement of systems.BMJ. 1996; 312: 619-622Crossref PubMed Scopus (518) Google Scholar Shewhart was concerned with the application of his methods and techniques for controlling the quality of industrial production processes. He discovered that statistical information, i.e. data, was generated by all industrial processes and that this data could be used to describe the variation that exists in every industrial process. This became known as statistical process control. 12Deming WE. Out of the crisis. MIT Center for Advanced Engineering Study. MIT Center for Advanced Engineering Study, Cambridge, MA1982Google Scholar He also described the quality management cycle, which consists of planning a new or improved process that will ultimately cause an improved outcome, implementing (doing) the process, measuring (checking) to see whether the process is having the expected outcomes, and then acting on the information (data) to reassess the plan or continue its implementation10Berwick DM. Controlling variation in health care: a consultation from Walter Shewhart.Med Care. 1991; 29: 1212-1225Crossref PubMed Scopus (221) Google Scholar, 12Deming WE. Out of the crisis. MIT Center for Advanced Engineering Study. MIT Center for Advanced Engineering Study, Cambridge, MA1982Google Scholar, 13AHCPR Pub. 95-0045. Using clinical practice guidelines to evaluate quality of care. Vol. 1. 1995Google Scholar (Fig. 1). Deming realized that Shewhart's methods, combined with other statistical aids, had great potential for the continuous improvement of production processes and the delivery of quality products. He developed a theory for management transformation that involved the application of 14 points, which he demonstrated to be equally applicable to banks, department stores, railways, and other service industries.12Deming WE. Out of the crisis. MIT Center for Advanced Engineering Study. MIT Center for Advanced Engineering Study, Cambridge, MA1982Google Scholar Deming introduced his ideas to U.S. industry early in World War II, and they had a profound effect on the quality and volume of war material production. After the war ended, American industry lost interest in his ideas as a result of the post-war economic boom. But he was invited to Japan to teach statistical methods for industry as part of Japan's reconstruction efforts. He taught the Japanese how to use statistics to find out what any process would do, then design improvements to make the system yield the best results. His basic idea was that the more quality you build into anything, the less it costs. Deming's concepts and teachings contributed greatly to the post-war Japanese industrial revolution 14Mann NR. Why it happened in Japan and not in the U.S.Chance: New Directions for Statistics and Computing. 1988; 1: 8-15Google Scholar (Fig. 2). Fig. 2W. Edwards Deming.View Large Image Figure ViewerDownload Hi-res image Download (PPT)The incredible success of Japanese corporations over the past three decades is a well-known story—television sets, videocassette recorders, compact disc players, steel, automobiles—to name a few. Japanese products, once derided as merely cheap copies of American or European designs, are now the standards of quality, and Japanese companies are similarly recognized around the world: Sony, Honda, Toyota, Mitsubishi, NEC, to name just a few. Unfortunately, it was not until the 1980s that American industry started to adopt the theory and practice of continuous quality improvement (CQI), which is what Deming's concepts have come to be called. For example, Ford Motor Company began consulting with Deming in 1981. Their well-known slogan, “Quality is Job 1,” is more than just words. Ford is now building automobiles with fewer factory defects, and profits have increased. The Boeing company used similar methods, which they called “working together” and “design-build teams,” in the design of its newest commercial jetliner, the 777, with remarkable success.15Sabbagh K. Twenty first century jet: the making and marketing of the Boeing 777. Scribner, New York1996Google Scholar A more recently publicized example from Europe involves the elite German automobile manufacturer, Porsche A.G.16Cash NC. Putting Porshe in the pink: German craftsmanship gets Japanese fine-tuning.New York Times. 1996 Jan; 20: 17-19Google Scholar Porsche was on the brink of bankruptcy in 1992 as a result of out-of-control manufacturing costs and slumping sales. Annual sales had fallen from 50,000 to 14,000, with only 3000 sold in the United States. In a bold and un-Teutonic action, Japanese engineers—mostly Toyota alumni—were hired to lead the reengineering of Porsche using Japanese CQI manufacturing principles and techniques. The results were dramatic. Manufacturing flaws have decreased 50%, large parts inventories have decreased 81%, manufacturing time has decreased by 40%, the number of workers has decreased by 19%, the number of managers by 31%, and factory space by 30%. Porsche is now making more cars, faster, by fewer people, and without decreases in its renowned technical sophistication and road performance. In 1995, the company reported its first profit in 4 years after $300 million in losses. CQI, sometimes called industrial quality control, is an organized system that combines a set of methods and a management philosophy to continuously improve processes, outcomes, and services. According to CQI theory, real improvement in quality depends on understanding and continuously revising the production process on the basis of data generated by the process itself.6James BC. Quality improvement in the hospital: managing clinical processes.Internist (Berl). 1993; 34 (March): 11-13Google Scholar, 17Berwick DM. Continuous improvement as an ideal in health care.N Engl J Med. 1989; 320: 53-56Crossref PubMed Scopus (1193) Google Scholar, 18Berwick DM. Harvesting knowledge from improvement.JAMA. 1996; 275: 877-878Crossref PubMed Google Scholar The focus throughout an organization, be it a factory, hospital, practice group, or HMO, is on continuous improvement through constant effort to reduce waste, rework, and inefficient processes. Sony embraces CQI using incremental product improvement as a corporate value, to the extent that its goals are to work to make its own products continuously better until they become obsolete.19Branscomb LM Kodama F. Japanese innovation strategy. Harvard University, Center for Science and International Affairs. University Press of America, Lanham, Md1993Google Scholar The Japanese call this kaizen—doing things better, little by little, all the time.20Imai M. Kaizen: the key to Japanese competitive success. Random House, New York1986Google Scholar What have Sony, Porsche, Deming, and Shewhart got to do with American medicine in general and vascular surgery in particular? As noted earlier, a process can be defined as a series of linked, often (but not necessarily) sequential steps that are designed to cause some set of outcomes to occur, to transform inputs into outputs, add value, and generate useful information. The practice of medicine can then be defined as a complicated series of interrelated processes and subprocesses. Some are clinical procedures (i.e., operations), whereas others are nonclinical activities (admission to a hospital, billing, etc.). The use of this concept of process enables health care to be examined in a logical and insightful manner. As Deming and Shewhart demonstrated, there is variation in every process, which is of two types: random and nonrandom. Random variation is that which is inherent in and a part of any real process. It represents the sum of many small variations, that is, background noise, and follows the laws of probability, behaving as a random function. It cannot be traced back to a root cause and therefore is considered as “appropriate,” distinguishing it from “inappropriate” or nonrandom variation, which arises from a single cause that is not an inherent part of the process. Inappropriate variation can be traced, identified, and eliminated. Random variation is expected in any real care-delivery process as clinicians deal with differences in patient presentation, illness, and preferences. It is most easily seen, for example, in the ranges of normal values for standard laboratory tests or body temperature. One of the two fundamental principles of CQI is the elimination of inappropriate or nonrandom variation. This is accomplished by using statistical process control to separate the random from the nonrandom variation. Only then is a process considered to be under control or stabilized, and only then can it be analyzed and improved in a scientific manner. The process can then become a research system within which the scientific method can be applied to test innovations or new ideas about how the process can be further changed to improve quality or increase productivity. Here is the Shewhart cycle in action, and in this system a clinical practice can become a true clinical laboratory. Medical care is very complex, and variation is widespread in all areas of clinical practice. There are large variations in practice patterns even among physicians who work in the same health care system, as well as between physicians at different hospitals and in different communities. It is often assumed that because health care is a quintessentially human business, variation is inevitable, perhaps even desirable. Differences in patient attributes account for some differences in outcomes from similar treatment processes, but physicians vary in how they diagnose and treat similar patients beyond what can be explained by patient factors alone. Abundant data document that geographic variations in the use of medical and surgical procedures are common. In a study of carotid endarterectomy within 13 large geographic areas in the United States, the number of procedures varied from 5 per 10,000 to 23 per 10,000 Medicare enrollees, a ratio of 4.6.21Chassin MR Kosecoff J Park RE Winslow CM Kahn KL Merrick NJ et al.Does inappropriate use explain geographic variations in the use of health care services? A study of three procedures.JAMA. 1987; 258: 2533-2537Crossref PubMed Scopus (516) Google Scholar When examined in a smaller area—23 adjacent counties—the rate of carotid endarterectomy still varied, from 5 per 10,000 to 41 per 10,000 Medicare enrollees, a ratio of 8.2. No statistically significant differences in the appropriateness of use of carotid endarterectomy were found, and the reasons for such a wide variation in use were largely unexplained.22Leape LL Park RE Solomon DH Chassin MR Kosecoff J Brook RH. Does inappropriate use explain small-area variations in the use of health care services?.JAMA. 1990; 263: 669-672Crossref PubMed Scopus (198) Google Scholar Marked differences between institutions as well as between surgeons in incidence and mortality rates for coronary artery bypass grafting procedures (CABG) have been widely publicized in the lay press as well as the scientific press.23O'Connor GT Plume SK Olmstead EM Coffin LH Morton JR Maloney CJ et al.A regional prospective study of in-hospital mortality associated with coronary artery bypass grafting.JAMA. 1991; 266: 803-809Crossref PubMed Scopus (335) Google Scholar, 24Williams SV Nash DB. Differences in mortality from coronary artery bypass grafts surgery at five teaching hospitals.JAMA. 1991; 226: 810-815Crossref Scopus (86) Google Scholar These observed differences in mortality rates were found to be independent of patient case mix. If differences in ri