Monday, May 21, 2012
Patterns tend to repeat themselves and regardless of what the mutual fund ads say, previous past results do tend to predict future performance. Baseball is all about statistics. Batting averages, RBIs, and ERAs are used by managers decide where to place his infielders, what batter to walk and whether a right hander should be allowed to pitch to a left handed batter. Percentages and odds try to predict what might happen next.
Decision making works much the same way in medicine. After gathering information, the doctor tries to come up with a differential diagnosis that will make sense of the symptoms, the patient’s complaints and signs, what is found on physical examination. The differential list varies in length and may begin with what is most likely and end with what is very rare. The list may also begin with what might kill the patient and finish with things that are less scary. In the scheme of things, making the right diagnosis may not be as important as not missing a killer diagnosis.
If a middle aged patient who smokes and has high blood pressure and diabetes presents with chest pain, the list begins with angina or pain from narrowed blood vessels to the heart. That list might also include aortic aneurysm or pulmonary embolus. If, though, the pain happens when the patient lies flat and is associated with heart burn and belching, perhaps the killer diagnoses aren’t there and the cause of the symptoms is gastro esophageal reflux, where acid backflows from the stomach into the esophagus. If only it were that easy. Women with angina may complain of fatigue instead of chest pain, but fatigue is such a non-specific complaint often there needs to be more to the story to push the diagnosis in that direction.
Many times, the diagnosis can be made after the first few minutes in talking with the patient and no tests are needed to confirm the presumptive diagnosis. Technology comes into play when the diagnosis is in doubt and usually, the more diagnoses that are being considered; the more tests are being ordered. It’s all about statistics and the comfort level of the doctor and patient living with uncertainty. The two ways the doctor can begin a discussion are “I know that…” and “I think that…” I know talks about certainty in diagnosis: “The x-ray show a broken bone” or the” CT scan show that there is no appendicitis”. I think means that the discussion is about percentages and the doctor taking past experience and predicting future outcomes
The history is often the key to making the diagnosis. If the doctor asks the right questions and listens hard enough, the patient should provide the answer. But patients can fool the doctor and unknowingly send the diagnosis hunt in the wrong direction. The doctor may take the information and misinterpret what is significant and what isn’t. There are many ways errors in judgment can occur, but a couple of the more common ones are anchoring and framing. These decision errors happen in everyday life and not just in medicine. Framing exists when the doctor makes a differential diagnosis list that is too narrow or too short and doesn’t keep an open mind about the patient’s situation. Anchoring talks about relying too gut feeling on one piece of information, and forgetting that patients with a known disease can always get sick with another illness.
In baseball, managers use statistics as a guide and add their gut instincts before making a decision. Regardless of past performance, the next at bat will be black or white. The batter who hits .300 will not get a third of a hit and the pitcher who will not get half a strikeout. In medicine, statistics may be based on studies involving thousands of patients and help narrow the differential diagnosis and the doctor needs to take those statistics, add experience and gut feelings to make a diagnosis and offer advice. At the end of the day, the patient can’t have half an appendicitis or a third of a stroke.
In medicine, when you here hoof beats, expect horses, not zebras…but never forget that there are zebras always lurking.