There are 3 algorithms in Harrison’s for evaluating a patient with transaminitis. At first glance, they seem comprehensive.
But are they?
Most algorithms and tables are NOT meant to be comprehensive. Some are. For example, the table on causes of FUO in HPIM has explicitly states that it includes all reported causes of FUO.
However, we should not assume that this is true for every table or algorithm.
Consider transaminitis, for example. When a patient with mild to moderate transaminitis comes to our ER, I will definitely think of diseases like leptospirosis. But Where is leptospirosis in these algorithms?
What about HLH? Histophagocytic lymphohistiocytosis almost invariably causes elevated AST and ALT. But it is not seen in algorithms and the diagnosis is usually missed.
It is because leptospirosis is geography-dependent and it is not seen in all areas. And HLH is kind of less familiar to physicians.
Here’s another example: the table for extremely elevated ESR levels. Subacute thyroiditis is also associated with very high levels of ESR. It is not in the table of HPIM (first chapter of infectious diseases section).
I’m not saying Harrison’s is a bad textbook. I love Harrison’s. My point is that we should not be misled by these algorithms and tables. They are not meant to be comprehensive. Rather, they serve as a guide, based on common epidemiology.
And this is one of the hard parts of the medicine. Every competent physician must adapt these widely-used algorithms to the local epidemiology of diseases in their practice area. And importanly, remain alert to diagnoses that may not be present in the algorithms and tables at all.
