29 Apr 2010

Diagnosis

Further to my eavesdropping, Ah Kap discussed whether we should rely on human memory or information technology in this era. He argued that doctors who looked up everything were more trustworthy than those who believed in their knowledge.

Frankly speaking, my textbook knowledge has decayed faster than uranium since I passed the membership exam. Apart from a few dozens of drugs that are commonly used, I always consult the British National Formulary before prescribing. Far from being ashamed, I believe treatment is the easy part. Anyone with average intelligence can copy from a drug book and determine anything from dosage to contraindications. With the help of computer programs, one can also easily check if there is dangerous interaction among the drugs being prescribed.

However, no matter how easy it is to obtain treatment information from various sources, the treatment cannot be applied if you do not know what is happening to your patient. You can easily find data showing that Drug A reduces mortality by 25% in a patient with Disease X, but what if you cannot diagnose Disease X?

For the latter point, people may have the romantic idea that new diagnostic tests and computer programs will one day help a doctor with little knowledge make the diagnosis of Disease X. To this, I refer to the discussion by my friend Szeto (see http://ccszeto.blogspot.com/2010/04/technology.html). If advanced technology takes over simple procedures, the inevitable result is junior doctors have few things for their practice, and, therefore, they would have more difficulty to master the difficult techniques. Similarly, if most of the bread-and-butter diagnoses are made by computers, future doctors will unlikely have medical sense. When a computer fails to make the diagnosis, what contribution can you expect from the doctor? Besides, even for bread-and-butter diseases, if a doctor has nothing in mind, he has no choice but to order numerous unnecessary tests before the diagnosis is made.

As Yuen Sir always says, 90% of the diagnosis should be made by history taking alone. I would add that the percentage depends on who and how the history is taken.

P.S. Is this discussion opposite to my original remarks described by Szeto? Yes, but people believe what they want to believe. My point is if robots cannot do not only simple procedures but also difficult ones, we are doomed by the time we retire.

P.P.S. Fine novels by Ah Kap can be found at http://wongsingk.blogspot.com/.

1 comment: