How do doctors decide?

Katherine D. van Schaik, MA, MD PhD Candidate (Harvard Department of the Classics/ Harvard Medical School)

 The following is a version of my conference talk, edited for the medium of a blog post.

I approach the patient-physician relationship both as an historian, and as a physician-in-training: as an MD candidate at Harvard Medical School, and as student in the PhD program in the Harvard Department of the Classics. I have pursued these two programs more or less concurrently, alternating between them every two years or so, and I will reach the end of this program of study in 2018.

Switching between programs has provoked a useful realignment of perspective every few years. I have had the opportunity to appreciate the concerns of the physician, the patient, and the historian, from all sides. My experiences working to appreciate perspective led me to a dissertation about decision making and definitions of health and disease in medical contexts. This research focus was born from the concern for evidence and for argument that, I believe, is a methodology shared by all those who would enquire, whether they ask questions in a library, a laboratory, a hospital, a classroom, or an examination room.

Modern medical narratives by physician-authors like Jerome Groopman, Atul Gawande, and Arthur Kleinman illustrate for a general audience the physician’s thought process during the various stages of an encounter with a patient, from the initial meeting to, in some cases, end-of-life care. These decision making processes increasingly involve diagnostic and therapeutic algorithms: heuristic tools that use meta-data to help physicians diagnose and treat, often by grouping patients into categories based on metrics (e.g., age, number of comorbidities, and levels of various biomarkers), then leading the physician through a series of if/then questions to reach the recommended treatment. This process of algorithmic decision making relies upon 1) defined disease categories and 2) the establishment of an organized system of factors determined to affect the diagnosis, treatment, and prognosis of the disease. Applied to the encounter between patient and physician, the algorithm shapes what a physician seeks to elicit from or observe in her patient.

In Greco-Roman antiquity, as the western world was beginning to explore and to debate concepts such as disciplinary boundaries, what constituted knowledge, and the workings of the human body and the natural world, individuals began to commit their observations of health and disease to writing – and then, to begin the millennia-long process of imposing sense and order on those observations.

Often considered ‘the father of western medicine,’ Hippocrates likely was a real person who lived in the Greek-speaking world of the fifth century BCE – but the works attributed to ‘Hippocrates’ are part of a corpus of texts that were written by many people, over the course of two hundred or more years. The ideas contained within the Hippocratic Corpus are diverse and often conflicting, products of communities of thinkers who sought new and different ways to explain their observations. Some texts present collections of symptoms and then suggest what treatments are effective for that particular combination; others discuss the symptoms experienced by particular patients, and over what time course; others present unifying theories of health and disease; still others focus on disease classification systems, or on the role of the environment in health and disease. There is great variation in the factors a physician is expected to take into account as he makes decisions about the diagnosis, prognosis, and treatment of disease.

In the 200s BCE in the city of Alexandria, in modern-day Egypt, two physician-scientists named Herophilus and Erasistratus carried out dissections and (sadly) vivisections of humans. Their writings have been preserved only in fragments, though their contributions to the development of medical knowledge were extensive – especially since, in the West, cadaver dissections were not carried out in similar ways again until the sixteenth century. Herophilus’ and Erasistratus’ contributions to the extant knowledge of human anatomy and physiology shifted debates about medical theory, as physicians began to incorporate this newly-acquired anatomical knowledge into their decision making processes.

Speculation about the workings of the human body, in part sparked by Herophilus’ and Erasistratus’ findings, led one group of physicians, the Empiricists, to emphasize the role of direct experience in medical practice. These physicians contended that, when it came to treating patients successfully, theoretical knowledge was less important than firsthand experience and personal knowledge of therapeutic outcomes: you should see as many patients as you can, so your decisions can be as well-informed as possible, they argued.

Creating a way to apply this experience practically and efficiently, the Methodists (named after the ‘method’ of healing they advocated) developed a system that grouped diseases into one of three categories: diseases of looseness, diseases of constriction, and diseases that were a combination of the two. Decisions regarding treatment were founded upon these groupings. Training for a would-be Methodist physician required only six months, one Methodist physician advertised, and this knowledge could be readily put into good use.

Galen (129-early 200s CE) condemned the Methodists for what he considered an excessively simplistic – even dangerous – approach. He wrote that, to train as a physician, one should learn (in our terms) philosophy, natural science, anatomy (through animal dissections), pharmacology, and the medical theories and practices of the physicians of generations prior, especially Hippocrates. Effective medical decision making involved a thorough knowledge of the patient, including psychological details, and was accomplished through the integration of many different pieces of information and recourse to theoretical frameworks.

Today, as we move toward medical decision making that involves concepts such as treatment algorithms, ‘big data,’ ‘crowd sourcing,’ and ‘precision medicine,’ patients are increasingly concerned that they are not being seen as individuals by their health care providers. A discerning view into past debates about the nature and organization of medical knowledge can help shed light on how modern physicians make decisions about diagnosis, treatment, and prognosis, and ensure that the patient remains at the heart of the decision making process.