Dr John Wright

Gastroenterologist
MbChB MRCP(UK) PhD

slider

Extraordinarily unhelpful investigations

Dr John P Wright
Updated: 21 January 2009

Why do we continue to do investigations that take the diagnosis no further?

The point of clinical medicine is to make a diagnosis and treat it. A powerful pressure in everyday clinical medicine is to treat the symptom. The patient wants it, his/her mother wants it and the medical insurance want it because in the short term making a diagnosis can be expensive. This however is what we do and we should not be pressurized to take the shallow folklore route. This reduces us to no better than your average snake oil salesman.

After taking a directed history we normally consider the need for further investigation. From our side the diagnosis may be clear or not and investigations may be needed or not to confirm or make the diagnosis. From the patient’s side further investigation has a number of consequences. More cost is one of them. On the other hand it shows that their doctor is “concerned”. It may show their families that their symptoms are potentially serious and they deserve sympathy. Whatever the reason investigations carry more significance than simply making or confirming a clinical diagnosis.

So in gastroenterology when we have no real diagnosis what do we do. There are two tests that are top of the pops. The first is an abdominal ultrasound. This is the best test for gallstones and gallbladder pathology. It is also very useful in detecting ascites. If these conditions are not suspected the test is generally a waste of time and money. It is not a useful investigation for heartburn, peptic ulcer, irritable bowel, colon cancer, anaemia or non specific abdominal pain.

The second test of dubious usefulness is the blood test for Helicobacter pylori. If a patient has a duodenal ulcer a Helicobacter pylori infection may be of the variety that predisposes to duodenal ulcers. In this situation eradication therapy with clarithromycin and amoxicillin may reduce the ulcer recurrence rate. However in the case of reflux oesophagitis and gastric ulcer, particularly related to NSAIDs, the deleterious effect of Helicobacter pylori is a lot less clear. To confuse further some strains of Helicobacter pylori may be advantageous and protect against peptic ulceration. Bearing in mind that at least 40% of the population has an infection with Helicobacter pylori the merit of indiscriminate use of eradiation therapy is at best unclear.

A third test of limited use is a faecal occult blood. Previously used to detect early colon cancer, it is seriously limited by false positives and negatives. If the patient is over 50 a routine screening colonoscopy is indicated even without symptoms. If the patient is younger, symptoms are a better guide to the need for further investigation than a random faecal occult blood.

So to what purpose are investigations done in gastroenterology. The first reason is to separate the healthy from the potentially sick. The most useful modality is a normal C reactive protein. Not perfect but extremely useful. The second reason for further investigation is to confirm the clinical diagnosis. This may be necessary but not always. Persistent acid related symptoms with danger signs warrant gastroscopy. Change in bowel habits in the over 50 year old certainly warrants colonoscopy. Weight loss with a raised C reactive protein needs to be pursued in the direction of the symptoms. But just as often a clinical diagnosis with an eye on warning symptoms is enough to treat.

In the end the use of investigations needs to be goal orientated. The test must answer a significant clinical question and the results move the diagnosis along. A test with minimal discriminating value is simply a waste of resources.