How helpful is the medical history and physical examination?

Hint: Very.

It is interesting to reflect that the advent of sophisticated laboratory testing and imaging studies ("blood-work and X-Rays") is recent in the history of medicine which is one of the world's oldest professions.  Just to put things in perspective, the stethoscope was first invented (by Courvoisier, if I remember correctly) in the late 1700's.  The widespread medical application of MRI did not occur until the 1980's.

Physicians up until recently had to be able to establish an accurate diagnosis purely on the basis of extended questions and answers (history taking) and a directed bediside exam (physical examination), as ultrasounds and scanners were not yet invented.

With sophisticated technology now at our disposal, it  becomes reasonable to wonder whether all the questions and answers and exams couldn't simply be foregone in favor of just jumping right to extensive lab testing and CT scans.

Well, someone did more than wonder.  They studied how over 400 patients were admitted through the emergency department at a large teaching hospital over the course of a little less than two months.  Specifically, they looked at whether the diagnoses were correct, and how correct diagnoses were made.

Reading about it so you don't have to, Dr. Gregory Rutecki reports that senior doctors arrived at a correct diagnosis in the ER over 84% of the time (residents, about 80% of the time).  As to the breakdown in how these correct diagnoses were made:

• History alone: 19.8% and 19.3%.
• Physical examination alone: 0.8% and 0.5%.
• Basic tests (complete blood cell count, chemistry panel, urinalysis, ECG, chest radiograph) alone: 1.1% and 1.3%.
• History and physical examination in combination: 39.5% and 38.6%.
• History plus basic tests: 14.7% and 14.7%.
• History, physical examination, and basic tests in combination: 16.9% and 18.5%.
• Imaging studies: 6.5% and 6.1%.

While physical examination all by itself is not helpful, history plus examination becomes the sharpest tool in the shed by double.  Note history taking alone is second.  Adding a few basic tests was in some cases helpful.  Note that "imaging studies" (here, scans and more sophisticated tests than just a basic X-Ray) was diagnostic only 6-6.5% of the time: less than half the accuracy of just history taking all by itself.

From what I was taught and what I continue to learn in practice, this finding is not surprising.  The history is telling you what's being experienced by the patient, and the examination can help to establish the basis for what's causing the problems being discussed.  Tests and studies can help to confirm your impression or exclude things that you would not like to delay in identifying.  However, you have to know where to look or what you're looking for to know what tests or studies to do.

When you think a bit about it, this sounds a lot like other areas in life.  A good mechanic wants to ask whether your car sounds like this all the time, or just on hills.  A technician will just hook your car up to the shop's computer, replace any/everything that doesn't green-light and bill you for it.  A good restaurant offers  a menu based on a thorough understanding of its clientele.  Mediocre ones offer lots of carbs/grease/salt for cheap.

Coming to a refined and accurate conclusion or diagnosis is based on a thorough (albeit time-consuming) comprehension of the problem or issue at hand.

h/t KevinMD