Radiology must regain initiative in research
ooking into the future always means interpreting mystical signs. We never know what will really happen because calculated reasoning plays only a small role in progress, and many medical procedures are founded on fallacies or financial interests. More than 10% of the population works in medical care, the pharmaceutical industry, or allied professions in some European countries .
Healthcare costs to society vary considerably. The United States spent 13% of its gross domestic product on healthcare in 2000, and Japan spent 7.8%. Yet Japanese life expectancy is the best in the world, while the U.S. is ranked 33rd . By comparison, Spain spent 7.7% of its GDP, Sweden 8.4%, and Germany/Switzerland 10.6%.
The contribution of medical imaging to healthcare costs is estimated to be less than 5% in the U.S. and from 1% to 3% elsewhere in the world [3,4]. The number of imaging procedures performed is rising by 10% each year, due mainly to increased utilization of x-ray angiography, CT, MRI, and PET. The technology with the largest expenditure is ultrasound because of the enormous number of ultrasound examinations.
More than a decade ago, in 1994, I posed the question: “Do radiologists have a future?” 
Radiology's future depends on more patient-oriented research and development of applications.
Some readers responded to the article, and together we concluded that while radiology does have a future, radiologists do not. Diagnostic imaging devices are no longer the exclusive province of radiologists. That is, if they ever were.
This verdict covers both routine clinical radiology and medical imaging research. Medical imaging has a rosy future ahead, and radiology will contribute to its success.
But radiologists do not own it. Radiologists depend on referrals from-and interaction with-other physicians. Meanwhile, other medical disciplines have established strongholds within medical imaging. In Germany, for instance, 70% of imaging examinations are performed by non-radiologists .
Radiologists must focus on clinical relevance and subspecialization to survive. They must become MR specialists or gastrointestinal radiologists, for example, if they are to be equal partners with clinical physicians and not simply their photographers.
Non-radiologists are more likely to use diagnostic imaging inappropriately and to select less suitable approaches than imaging specialists. Understanding constantly changing techniques and equipment is nearly impossible for somebody who is not completely dedicated to medical imaging.
Fighting the “amateur radiologists” is difficult and exhausting. Patients may be better served if radiologists educate their rivals instead. Careful preservation of existing radiological know-how for plain x-ray examinations, for example, might save future radiologists from having to learn these techniques from general practitioners.
Despite discussions of turf wars and threatened positions, there is actually a shortage of radiologists. Out of 400 positions for radiologists advertised in the U.K. in 2000, only half could be filled. The U.S. had 330 empty positions for radiologists during the same year, and Sweden had more than 50. The lack of staff is still a problem, and it is a vicious cycle. We have more medical imaging in hospitals and clinics, and fewer trained radiologists.
The situation in medical imaging research looks even bleaker. Research positions command lower salaries than clinical appointments and are less prestigious. Radiology research 30 years ago comprised mainly patient-centered investigations into the improvement of x-ray imaging techniques. The advent of computers and new modalities has broadened the scope of research considerably. Sometimes it seems as if the balance has shifted too far from intelligence and professional craft toward machines.
Today's discipline of “medical imaging” unites conventional radiography (including digital imaging), ultrasound, CT, MRI, interventional radiology, nuclear medicine, and optical imaging, as well as paramedical methods in biosciences, pharmacology, and computing. Boundaries between anatomic and/or metabolic data acquisition, the development of new tracers and/or contrast agents, and data storage and/or distribution have become blurred. Everything is part of medical imaging.
Budding researchers, group leaders, and department managers are faced with a wide range of topics to select from. The decision is not easy. Personal and financial interests may overlap and influence judgment. Driving forces in research are curiosity and ignorance, a hunger for power and money, and the question, “What is best for the patient?” Research by physicians should have a moral and ethical dimension.
Much of today's academic radiological research is solely technology-oriented or a combination of technology and application. There is hardly any “pure” or “basic” research. Sometimes young researchers perform “l'art pour l'art” and confuse method with result. Their results are changes or new versions of existing methods. Instead of keeping their eyes on the goal, they just play around.
Certain kinds of diagnostic screening that rely on indirect signs of malignancy will be replaced by non-imaging laboratory tests once these become available. Researchers should remember this when planning ahead. Screening methods that are morphologically and functionally nonspecific, such as x-ray mammography, will vanish over the next 20 years. It is useless to invest time and money in the development of techniques that are inferior to those that exist already.
A return from technology-focused R&D to simple patient-oriented research may still be possible. This kind of radiological research is not costly and requires little equipment. A change in mental attitude would be needed, however. Actual research requires tenacity and persistence. It is all too easy to become diverted. Not many people have the time, financial resources, or energy to investigate multiple scientific topics. Radiologists usually move into research while acquiring the fundamentals of their craft. Many part-time researchers regard this as a step in their career, not the beginning of a lifelong commitment. Yet this is not entirely bad news. These researchers will still have the opportunity to see and learn how academic life functions.
Anyone planning an academic research project should review his or her own-and the group's-competencies critically. These competencies should include the ability to organize, manage, and follow through on a project from inception to delivery of final results. Radiologists, in common with all other medical doctors, are not scientists per se.
The definition of what constitutes research is often in the eye of the beholder.
Most European countries have few, if any, dedicated research positions in radiology. Introductions to and basic training in research activities are scarce. A full-time clinical job cannot be combined with comprehensive research activities. If a department head suggests that research could be performed during evenings and weekends, young would-be researchers should consider moving elsewhere.
The definition of what constitutes research is often in the eye of the beholder. Few European countries have implemented a quality review system that involves visits from external assessors.
Europe is lacking a truly-scientific academy where young radiology researchers could be taught solid research skills. A “European Academy of Medical Imaging Sciences”, for instance, could teach the use of library databases, data analysis using inferential statistics, the design of complex experiments, preparation of formal laboratory reports, and presentation of results, both orally and in written form. A satellite network consisting of existing research laboratories with a proven track record could serve as the foundation for such an academy. The institution should be multidisciplinary but essentially medical rather than a mixture of computer science and industrial applications. Patients should not become oddities in the research scheme.
The nature of presentations given at major radiological meetings has changed over the past 20 years. The main focus is often not radiological skills but supporting technologies and financial management. Some 7,000 or so scientific papers on imaging topics are published every year. This includes 3,000 in North America and the same number in Europe .
In 1991, former British Medical Journal editor Dr. Richard Smith stated that fewer than 1% of these papers contain new scientific findings or relevant medical information that has an impact on medical diagnostics and therapy . This statement still holds true.
Radiology research is becoming increasingly competitive and aggressive, due mainly to the huge commercial market. Much of the explosive development in medical imaging is fueled by the enormous power of industrial players and their marketing departments. Most scientific groundwork in medical imaging is performed either by researchers from non-radiological disciplines, such as medical physics, biology, pharmacology, neuroscience, computer science, or the military or by x-ray technologists.
Many people still regard the U.S. as the best country to perform research in. Academic research does progress at the same level in certain countries on this side of the Atlantic, but U.S. institutions offer better working conditions in many instances. Approximately 20,000 German researchers, physicists, medical doctors, and molecular biologists are thought to be carrying out research in the U.S. at present.
Some researchers and research groups, mostly at U.S. universities, work specifically toward grand rewards such as the Nobel Prize. They choose their research topics with this in mind and structure their research teams accordingly. They no doubt perform excellent research. But their main aim is to gain money and power, not to help patients. They lobby widely, recruiting the assistance of their university's public relations department or external agencies. Their results are published in the daily press before they appear in scientific journals.
Europe has excellent facilities and competent researchers. Its main problems are money-sucking state bureaucracies, rigid hierarchical structures, and difficult access to multidisciplinary cooperation. Radiological research used to be the domain of European and North American scientists. Since the end of the 1980s, however, postdoctoral fellows from China and India have held powerful positions in U.S. academia. Some of these researchers have now returned to their home countries and are installing very competitive, low-cost, high-quality academic and commercial research facilities for the life sciences and medical technology.
Many European academics are not aware of this development. Multinational companies, on the other hand, have been watching this move and have created research centers in China and India to tap into the knowledge base of these well-trained and hard-working scientists.
Industry versus Individuals
Medical industry – not academic researchers – can take most of the credit for the explosive progress in medical imaging. Innovation from industry is driven by market needs. But the new developments companies promote are usually oriented toward short-term financial gain. They do not produce medical equipment or accessories for altruistic reasons.
At least 20% of all presentations at the annual RSNA meeting in Chicago have commercial links. It is not necessarily inappropriate for commercial interests to influence the topics of radiological research, but this kind of research might lack a direct impact on disease diagnosis and treatment. Efforts to improve PACS and RIS equipment, for example, will just produce better systems for patient administration. The impact is equivalent to that of a new color sticker code on x-ray envelopes 30 years ago. Working toward the paperless department is not radiological research. Novel computer applications in offices may be good for the economy, but they have no direct bearing on patient care.
Many universities and politicians have pushed third-party research over the past decades. Sponsoring agencies, such as state research foundations and the European Commission, manufacturers of equipment and accessories, and venture capitalists, have all contributed financially to research.
Academic researchers who have reached the “final goal” and cooperate with one of the big commercial companies usually have a fast, crude awakening. One should not harbor great expectations from such liaisons, which are as flexible as semi-democratic state administrations. Collaboration with small companies, however, can be different, and research by itself still counts.
The tremendous upsurge of medical imaging techniques has made it difficult to decide how best to choose between different examinations and how to interpret their results. An innovation may be deemed clinically valuable only after millions of euros have been spent or decades passed after its introduction. Healthcare payers and patients should demand hard evidence that a certain imaging examination is useful, cost-efficient, and beneficial. Outcomes research should govern such decisions if radiologists want to influence how and when these new technologies are used. Radiologists rarely perform this kind of research.
Change for the Better?
The radiologist's professional environment undergoes a slight metamorphosis every year. After 10 years, the entire environment has changed completely. Sometimes we would like to stop things, at least for a while. We would like to say, “That's it. No more change for the next 10 years.” Believing in progress is one engine of humankind. But uncritically believing in progress is stupid.
Try to answer these questions honestly: Do I understand the existing imaging technologies? Can they be used to answer the diagnostic (or therapeutic) questions asked by patients and/or referring physicians? If not, how can research improve my capabilities to answer them? If yes, do we need more diagnostic tools? Or do we need more standardization, better understanding and education, and continued professionalization?
My view is that we need basic research. We need patient-oriented research and development of applications. We do not need more consumer electronics turned into radiological toys.
1. Federal Statistical Office, Germany 2005.
2. Federal Statistical Office, Switzerland 2005.
3. Margulis AR, Sunshine JH. Radiology at the turn of the millenium. Radiology 2000; 214: 15-23.
4. Sunshine JH, Mabry MR, Bansal S. The volume and cost of radiologic services in the United States in 1990. AJR 1991; 157: 609-613.
5. Rinck PA. Rinckside. Do radiologists have a future? Rinckside 1994; 5,4: .
6. Kauffmann G. The status of diagnostic radiology in Germany (with special reference to MRI). Fortschr Röntgenstr 1999; 170: M7481. Paper in German.
7. Mela GS, Martinoli C, Poggi E, Derchi LE. Radiological research in Europe: a bibliometric study. Europ Radiol 2003; 13: 657-662.
8. Smith R. Where is the wisdom? BMJ 1991; 303 (6806): 798-799.