Hospitals can be dangerous and inefficient; therefore it is easy to connect with Atul Gawande’s recent New Yorker essay “BigMed” suggesting that the streamlined, production processes found at the Cheesecake Factory can and likely will be applied to healthcare. Yet hospital care should not be confused with the full spectrum of healthcare. One must make the distinction between the cognitive process of medical diagnosis occurring in exam rooms, with the procedural basis of surgical care and hospital recovery. While Dr. Gawande has provided a wonderful revealing portrait of cost-effective, fast, food preparation and delivery at the Cheesecake Factory, he has focused on the process of creating the meal, not the process of deciding what meal to make. Successful surgery, for the wrong diagnosis, is a problem. If we are to solve some of healthcare’s largest failings we should focus on what happens as physicians try to address their patient’s problems, diagnose and make decisions, at the table of medicine called the exam room.
Consider the continuum of the patient encounter, from first symptoms, through diagnosis and therapy at a restaurant called Med. At Med I spend all of my shifts with my patrons at my tables. This is an unusual restaurant since the patrons are never sure of what they want to eat and appear every 20 minutes with ever changing lists of unique groups of ingredients to share with me. There are varying ingredients and thousands of meals that can be created. The patrons know the ingredients, but not the meal that they would like to eat. From memory I respond to the customers list of ingredients and ask many questions, take the pulse and other vital signs of the customer, order blood samples, radiographic studies and then decide for the patron which meal their ingredients add up to. All from memory. At Med, restaurant patrons also ask for foods and “food tests” they have seen on television all purported to be risk free. Further complicating the process is my customer is not out for a fun and relaxing evening, they are in small booths in skimpy, open at the back gowns, often anxious and uncertain if they will be harmed or poisoned by my foods, or simply receive a meal they do not want. Some are in pain and some are depressed, while other customers are totally unrealistic about the meal that is to be delivered. You see at Restaurant Med, where patrons only can speak to their wait staff about ingredients, and demand the modern but unhelpful ovens they heard about from friends and the media, it is really difficult to create meals that patrons thoroughly enjoy.
An appendectomy should be consistently performed and priced, but how do we consistently perform and price considering the ambiguity inherent in diagnosis itself? Unlike a restaurant, where customers choose a meal by ordering a meal, at restaurant Med some higher force gives an unfortunate person an undifferentiated and undiagnosed problem that needs and deserves an answer. As it turns out, none of the patrons really want to be eating at restaurant Med, as they always receive a meal they did not ask for.
Patients do not choose their diagnoses from menus; doctors must discover and diagnose them.
If your waiter tries to memorize all the orders at all the tables, you might get the wrong meal, and if your server is in a hurry, thai dipping sauce might be spilled on your new silk blouse. Likewise if physicians are in a rush, they might not take a thorough history, perform a complete physical exam, or have an accurate and thorough list of diagnostic possibilities, ultimately resulting in the wrong diagnosis. If your physician believes he or she can memorize all the questions, tied to all the possible diagnoses you also might receive the wrong diagnosis. With that wrong diagnosis you might end up in a hospital more efficient than the Cheesecake Factory with doctors efficiently ordering unnecessary tests, and performing wrong surgeries for the wrong diagnosis all with the ease and speed of the best assembly line on the planet.
Diagnostic and patient management error caused by cognitive mistakes in the exam room are all too often overlooked and unmentioned in the discussion of repairing our broken healthcare system. There are over a billion outpatient visits in the US each year, and numerous studies have shown 15-20% of these visits have an inaccurate diagnosis. Autopsy data proves this, malpractice insurers know this, and policy makers avoid it. Add diagnostic error in the emergency room and walk-in clinics to error in the out-patient offices of medicine and you have more than 200 million errors. If we are to resolve some of healthcare’s deepest woes we need to address diagnostic errors and the decision-making occurring at the restaurant table of medicine, the exam room. A bright light needs to be shined on the simple fact that there is too much to know, to ask and to apply during a 15 minute encounter unless the patient has the simplest of medical questions or problems. Medical informaticists, researchers and innovative companies are focusing on this essential limitation of medical decision-making by designing information systems to be used by physicians at the point of care, during the patient encounter. Problem oriented systems can also be designed for use by patients in advance of the visit, and the future holds home-based information coordinated with professional clinical decision support. These new information tools are beginning to take the guessing out of which ingredients (symptoms) relate to the meals that the patient ultimately receives (diagnosis and treatment). If medical care is truly to be driven back to primary care we need to arm the waiters of medicine with purposefully designed tools and training to resolve ambiguity, aid diagnosis and inform therapy in the exam room.
Art Papier MD
Art Papier MD is CEO of Logical Images the developer of www.visualdx.com a clinical decision support system, Associate Professor of Dermatology and Medical Informatics at the University of Rochester College of Medicine, and a Director of the Society To Improve Diagnosis In Medicine (SIDM) http://www.improvediagnosis.org/