For those who work in science, a classic question is always asked “How accurate is this test?”. Usually, no test can give an absolute answer to the patient’s condition. There are always 2 possible false positives:

– Scenario 1: A person is perfectly healthy but the test results show “this person has the disease”. This situation is called “False Positive”.- Scenario 2: A person has the disease but the test results show “this person is healthy”. This situation is known as a “False Negative”.

Viewing: what is sensitivity

False positives will bring a lot of anxiety to patients, sometimes even having to undergo unnecessary treatment regimens. It should be remembered that not all treatment regimens have few side effects, wrong treatment will be both costly in terms of economy and not good for the health of the patient. Too many false-positive tests will cause huge economic losses in society, especially for diseases where the cost of a single course of treatment is high, such as cancer.

False negatives have equally serious consequences, a person who has an illness but tests negative again leaves them untreated and can cause them to miss the right times for treatment. For example, in cancer, early or late detection can lead to two completely different treatment outcomes.

For cancer, false negatives only bring harm to the patient itself. For infectious diseases, false negatives can be catastrophic for the community. Letting a patient with an infectious disease roam freely can sometimes lead to catastrophic collapse of the entire health system.


In any disease, society has only 2 groups of people: “sick people” and “healthy people”. In a test, the test is also designed to return only two types of results: “positive” and “negative”. The relationship between 2 groups of people in society and 2 test results is shown in Table 1.

READ MORE  Rpa Application In Life

Table 1. Possible Test Situations Truth Sick Person Healthy Test Positive A(correct diagnosis) B(false positive) C Negative(false negative) D(diagnostic) exactly)

There are quite a few statistical parameters that can be used to demonstrate the usefulness of a test, within the scope of this article will only refer to two values, “sensitivity” and “specificity”. (specificity).

– Sensitivity reflects how likely a person with a disease is to be correctly diagnosed, i.e. sensitivity = A / (A+C).- Specificity reflects how likely a healthy person is to be correctly diagnosed, i.e. specificity = D / (B+D).

Sensitivity and specificity represent likelihood only, so confidence intervals (CIs) can be calculated using standard methods. Of course, the method of calculating CI must be realistic, and it is clear that there is never a value in the > 100% confidence interval.

Another important issue to consider when assessing sensitivity and specificity accurately is the need to evaluate research methods. An inappropriate research method will not produce high-value data.


Of course it would be perfect if a test had both sensitivity and specificity at 100%, i.e. “false negative” and “false positive” would not occur. However, reality shows that when we try to increase the sensitivity of a method, the specificity will decrease and vice versa (Figure 1), so there is a need for choice.

Should a test with high sensitivity be chosen or should a test with high specificity be selected? Before we answer this question, let’s go back a little bit about the value that each type of test brings.

See also: What is Toeic – How to Practice Toeic Exam Effectively

READ MORE  What does it mean to run out of solution

Figure 1. Correlation graph between sensitivity and specificity

Meaning of High Sensitivity: Sensitivity = A / (A+C), the closer this value is to 100%, the closer the value of C is to 0. This means that “false negatives” are more unlikely, like So a test with 100% sensitivity can help patients feel secure if the test results are negative.

To put it simply, a highly sensitive test is a “better catch than miss” test. This is extremely important in preventing the spread of disease. Rapid screening tests by immune responses (antigens – antibodies) such as HIV, HBV, HCV, etc. usually work based on this principle.

For screening tests (rapid test), sensitivity is the first priority, a rapid test kit has a sensitivity that does not reach 100%, then the specificity value of that kit, even if it is high, is useless. because at this time using the kit will let the “carriers” go out into society.

Meaning of high specificity: Specificity = D / (B+D), the closer this value is to 100%, the closer the value of B is to 0. This means the more unlikely a “false positive” is, Thus, a test with 100% specificity that returns a positive result can be said to be “unquestionably”.

A test with high specificity is often appropriate for patients with clinical signs screening or have suspicious results during screening. This is considered a “reconfirmation” step before giving a treatment regimen and is extremely necessary for diseases where the treatment regimen is expensive or harmful to health such as cancer, …

For “reconfirmation” tests, high specificity is preferred, because the test results will be decisive for the treatment regimen. Inadequate specificity will risk causing doctors to offer unsuitable treatment regimens for patients that are economically costly but not highly effective, and may even miss treatment opportunities. patient treatment.

READ MORE  What's the difference between Imported Salmon and Farmed Salmon?

It is necessary to combine both types of tests: In fact, the combination of both types of tests is essential to both save time and costs during extensive screening, and ensure the right treatment of the right people, the right disease, avoid unnecessary effects on the health and economy of the patient.

Son Pham – Summary and Translation


1) Cook C, Cleland J, Huijbregts P. Creation and critique of studies of diagnostic accuracy: use of the STARD and QUADAS methodological quality assessment tools. Journal of Manual & Manipulative Therapy. 2007 Apr 1;15(2):93-102.

2) Leeflang MM, Moons KG, Reitsma JB, Zwinderman AH. Bias in sensitivity and specificity caused by data-driven selection of optimal cutoff values: mechanisms, magnitude, and solutions. Clinical chemistry. 2008 Apr 1;54(4):729-37.

See also: What is Tss – Total Floating Solids

3) Saunders LJ, Zhu H, Bunce C, Doré CJ, Freemantle N, Crabb DP. Ophthalmic statistics note 5: diagnostic tests—sensitivity and specificity. British Journal of Ophthalmology. 2015 Sep 1;99(9):1168-70.