Sciences de la Santé
then proceed to a gold standard examination, which can vary from the simple imaging to surgery or autopsy! If the results were inconclusive, the data analysis is often unorthodox and challenging. 2- A clinical diagnosis rarely resides in a single symptom or sign, but rather in the patterns emerging from many symptoms and signs combined. The data is thus subject to multivariate phenomena. 3- Clinical examination researches often don’ t grant any fame, prestige or reward. In result, researchers influenced by the exciting challenges and academic rewards, are much more interested in investigating laboratory or imaging tests. 4- Financial issues are probably the biggest reason behind this desert in research. Pharmaceutical and medical technology companies often promote and reward clinicians investigating in advanced technology. And one must not forget about political and financial purposes behind this: it is far more interesting for companies to have physicians buying their“ superadvanced” devices than taking their time with a thorough history taking and physical examination. 5- Finally, as mentioned earlier, many clinicians still consider bedside diagnosis as an“ art” more than a real science.
At this point, only one question remains: why is it worth it to invest in physical examination research? A part of the answer is that these researches could help make the physical examination so accurate that in many cases we would not need imaging or laboratory results at all. As an example, Crombie found out that more than ¾ of the diagnoses in primary care are established after a brief history and a routine clinical examination 8.
Steve McGee emphasizes on the importance of“ Stop rules” in daily practice. These rules result from combinations of bedside findings that argue so much against the diagnosis that the investigation should stop and no imaging or laboratory test is needed 9. Examples of these include the Ottawa rules for ankle injury and the Alvarado score for appendicitis. If such stop rules were applied in the US, an estimated 700 billion dollars a year spent in unnecessary tests and procedures would have been saved 10.
An Evidence-based approach 4, 9 As we mentioned above, one of the main reasons behind the delay which physical examination suffers from is the inability of physicians to rationalize their findings and fully benefit from Evidence-based researches. Here we will try to give examples of how to effectively use physical examination data. 1- Sensitivity and Specificity, classic but not very useful We perfectly know that biostatistics are a very powerful hypersomnic agent for most medical students, but we ask our beloved readers to forgive us, we have to
12 Printemps 2018 dive into some details... it will be worth it( or else refunded).
As a quick recap, let’ s review the definition of sensitivity and specificity. Sensitivity is the ability of a test to be positive in patients who have the disease. In other words, it is the proportion of patients with the disease who have the physical sign( i. e., have the positive result). Specificity is the ability of a test to be negative in patients who don’ t have the disease. In other words, it is the proportion of patients without the disease in whom the physical sign is absent.
The calculation is simple, here is an example: let’ s imagine a study of 100 patients to evaluate the power of a certain test to detect a disease. The following column summarizes the findings:
Test scores
Positive test
Negative test
Has the disease
True positives, ex: 22
False negatives, ex: 20
Reality Does not have the disease False positives, ex: 3
True negative, ex: 55
Remember that Sensitivity is the number of sick patients with a positive test compared with the total number sick patients. Thus, it is calculated by the following formula: Sensitivity = TP /( TP + FN). In this example: 22 /( 22 + 20) = 52 %. Specificity is the number non-sick patients with a negative test compared with the total number of non-sick patients. Thus, it is calculated by the following formula: Specificity = TN /( TN + FP). In this example: 55 /( 55 + 3) = 95 %.
Knowing how to calculate is good but dealing with the numbers’ meaning is what is most important. THAT is far more complex than seemingly thought. Consider a real example: a 40-year-old man with a history of chronic alcohol consumption comes to your office complaining of an increased abdomen volume. At a first glance, and given the history of the patient, you think of an ascites. Rather than directly requesting an abdominal ultrasound, you want to figure out how well physical examination alone can detect ascites. After a little research, you find 3 meta-analyses that reported the results of the different studies 11-13. The following table summarizes the results:
Specificity(%) Specificity(%) Inspection
Bulging flanks 11-13 72-93 44-70 Edema 12 87 77
Palpation et percussion Flank dullness 11, 12 80 – 94 29 – 69 Shifting dulness 11-13 60 – 88 56 – 90
Fluid wave 11-13 50 – 80 82 – 92