Of all the sciences, medicine has the most direct connection with human life. It encompasses everything from molecular mechanisms to the personal narratives, beliefs, and feelings of each individual. Sharing Medicine, starting with a series of 7 Viewpoints, tackles the challenge of this vast diversity head-on. The articles in this series summarize how physicians currently share knowledge, skills, and experiences among ourselves as a professional community and with the patients we serve; and in each article the authors suggest some ways in which we might do this better.
The challenges present themselves most immediately when we enter into dialog with patients and try to align our understanding with theirs. This process has long been called shared decision making, but within the series it will emerge not as a single act of choice but as something more wide-ranging and complex. Shared decision making implies that we have reached agreement about what kind of help people who come to us are seeking and what we have to offer. But the essential first part of the process—eliciting an understanding of the goals and preferences of the individual patient—is too often omitted or hurried through, so that clinicians can start offering their own views about the diagnosis and treatment options on offer. Effective encounters should open with a “preference diagnosis,”1 in which the patient and physician reach a common understanding of the reasons help is being sought and the outcomes that are most important. This can be followed by an exchange of knowledge made relevant and understandable to the patient with the help of structured dialog and, in some cases, a shared decision tool.
Most clinicians rightly regard tools as adjuncts and see themselves as repositories of knowledge and specialized skills, which are developed and augmented by their experience. This model is both affirmed and challenged in these Viewpoints. Knowledge not only should be available to be shared with patients, but also should be personally understood by clinicians to facilitate effective communication.
As the sources of knowledge for clinical decision making are examined, the shortcomings of evidence-based medicine as a foundation for personal treatment become evident. The springs of knowledge for shared decision making consist of the trials that produce evidence about treatments. But these trials may be poorly matched with clinical need2 and are often reported in ways that make it impossible to determine outcomes of importance to patients, to compare treatments, or to match the participants in a trial with populations of patients in real life.
A partial and retrospective solution would be to make all data available from the trials that determine the treatments that are currently standard. Many pharmaceutical companies have accepted this principle, but relatively few data sets have been made fully available, and of those that have, scarcely any have been subjected to primary reanalysis.3 Prospectively, data sharing in medicine should be on the scale that other sciences take for granted. It is anomalous that many in the academic medical community continue to regard data from clinical trials as personal or institutional property, to be hoarded and published only as those who hold the data decide. Other sciences, such as astronomy, particle physics, and genomics, flourish because researchers share their data in real time.
Compared with other sciences, the whole structure of knowledge generation and dissemination within medicine seems archaic, duplicative, and inefficient. A great cycle should exist connecting clinical encounters, the research agenda, the design of randomized clinical trials, the systematic synthesis of knowledge, and finally the tools and informational graphics that inform patients and physicians of treatment options. At present, this cycle operates inadequately at every point. Each clinical encounter is a potential learning opportunity, and an entry point for data that can be used to generate insights and hypotheses for research. But most research is not based on such assessment of clinical reality. Publication even of summary findings from interventional research is often absent or incomplete, and if complete data sets are ever shared, the process is slow and reactive. Meta-analysis is often an arbitrary process in which the spread of individual effects and their context can be lost in order to arrive at a plot summarizing effects across merged populations. To identify and transmit evidence relevant for individual decision-making, timelier and more sophisticated methods are needed, such as cumulative rapid meta-analysis and interactive graphics, and these are beginning to become available.4
Knowledge processing is vital for clinical practice because no clinician can master the entire evidence base. Skills, however, are of equal importance, especially those of attentive listening, rational diagnosis, and effective communication with patients. Without such skills, knowledge can be misunderstood and misapplied by both parties. At present, there is a lack of support for the structured development and sharing of skills among clinicians. Diagnosis is a fundamental skill that we practice from the moment we enter medicine, and which we pride ourselves in refining throughout a working lifetime. But continuing professional education does not currently include systems for sharing and improving diagnostic thinking, to make it more effective and understandable to each other and to patients. Diagnostic research too often fails to follow diagnostic pathways from inception to clinical outcomes. The predictive characteristics of tests cannot be divorced from context, and each offer of a diagnostic strategy is an interventional choice that should be capable of being explained and shared with patients.
Sharing Medicine requires the further skill of being able to communicate in ways that foster an open and trusting relationship with the people who seek help. In some simple clinical contexts, where there is a straightforward choice to be made, such as a decision about elective surgery, it could be argued that this is scarcely necessary: all that is needed is adequate information and a consent form. But most of medical practice is not like this: clinical care does not consist of a series of easily defined take-it-or-leave it choices but is a process of understanding developed and deepened over time. Sharing understanding with patients is a form of dialog and interaction which cumulatively develops and which effects changes in both parties: it lies at the heart of primary care, and it is essential for kind and effective clinical practice in all specialties.
A selection of books about personal experiences of illness can be found in any large bookstore. There may also be a selection of books by physicians describing their experience of practice. Very rarely, and generally only in a few articles in medical journals, have the 2 sets of experiences been aligned.5 Clinicians are insufficiently encouraged to learn from individual patient experiences and the ways in which these experiences affect them, as human beings who relate professionally with others.
Good medical practice carries with it a constant threat of overload: information overload, cognitive overload, expectation overload, and affective overload. Information overload from proliferating medical knowledge was noted in 1881 by John Shaw Billings, MD, the American librarian and military surgeon, in the second volume of the British Medical Journal,6 and it has increased exponentially ever since. Cognitive overload happens when clinicians try to match this wealth of knowledge with the demands of daily practice. Expectation overload can occur when motivated clinicians have excessive expectations of themselves, while at the same time struggling to meet the perceived expectations of others—not just patients but also administrators and colleagues. And affective overload can result from constant personal closeness to illness, suffering, and mortality. These burdens can and should be shared with colleagues to spread the load and to decrease the likelihood of burnout. The various forms of overload can also be alleviated by the way we interact with the people we serve. By achieving more open and mutually supportive relationships with patients (“shared mind”), we can help them more effectively and at the same time reduce our own feelings of overload and stress.7
The future of medicine cannot consist simply of big data processed by big machines, with clinicians standing back and accepting the printouts. Of course, the analysis of information in many databases, including electronic health records, may well produce important advances in knowledge and treatment, but these advances will invariably have to be made meaningful through a process of shared understanding. Such understanding begins with clinicians as individuals. It is shared between clinicians within our community of learning and practice. It is shared with patients in every clinical encounter. The future of medicine lies in sharing medicine.
Corresponding Author: Richard Lehman, BMBCh, MRCGP, Cochrane UK, Summertown Pavilion, Middle Way, Oxford OX2 7LG, England (firstname.lastname@example.org).
Published Online: July 3, 2017. doi:10.1001/jamainternmed.2017.2371
Conflict of Interest Disclosures: Dr Lehman reports that he is a Senior Advisory Fellow in Primary Care at Cochrane UK, and a paid consultant to the Yale University Open Data Access project (YODA). No other disclosures are reported.
Additional Contributions: I thank Harlan M. Krumholz, MD, SM, for his valued support, encouragement, and insights.
Lehman R. Sharing as the Future of Medicine. JAMA Intern Med. Published online July 03, 2017. doi:10.1001/jamainternmed.2017.2371