# Why does SpIn SnOut exist?

Can somebody explain the logic behind "SpIn SnOut" ?

Let Sp = specificity, Sn = sensitivity, TP/TN = true positives/negatives, FP/FN = false positives/negatives, and PPV/NPV = positive/negative predictive value. SpIn = "Specific test when Positive test rules IN the disease", and SnOut = "Sensitive test when Negative test rules OUT the disease").

Let's take SpIn for a spin in the following examples.

Firstly, the phrase regarding specificity, "when Positive test rules IN the disease" implies that a positive test has a high probability to be true. Putting these words into formula leads to a high (TP/(TP+FP)) = high PPV. The phrase "when positive test rules IN" describes high PPV rather than high Sp.

Secondly, the formula for specificity is Sp = (TN/(TN+FP)) – it describes the probability of a healthy individual to be identified as healthy. So, while, the formula for specificity describes how good a test is at categorising healthy people (aka people who ideally should get a negative test result), the SpIn mnemonic refers to positive test results. This juxtaposition is confusing (formula referring to healthy people vs mnemonic referring to positive test results).

Thirdly, the formula for specificity Sp = (TN/(TN+FP)) doesn't directly address the probability of a positive test to be true (remember, the mnemonic states "specific is when positive test rules in disease), albeit it implies the existence of a low number of false positives (Sp = 1 if FP = 0), which itself implies a high PPV (PPV = 1 if FP = 0). However, for any value of FP ≠ 0, PPV can theoretically equal anything, so it doesn't make sense to say "specific test rules in the disease".

So, why does the SpIn SnOut mnemonic exist, and is it actually true?

• Are you asking to explain sensitivity and specificity? It is not clear for me what is your question.
– Tim
Commented May 30, 2021 at 20:01
• I'm asking why specificity is a feature of tests which can rule in (confirm) disease (in medicine).
– Tib
Commented May 31, 2021 at 5:22
• Not sure what you mean. It is just one of the metrics to assess how good is a diagnostic method or an predictive algorithm etc.
– Tim
Commented May 31, 2021 at 8:10
• An alternative way to ask my question is - why is specificity the most important metric for a diagnostic test, and why is sensitivity the most important metric for a screening test ? It's a very basic question about the very basics of specificity and sensitivity.
– Tib
Commented Jun 1, 2021 at 1:48

I think you are thinking way too deeply, and reading to much into those terms.

SpIn and SnOut are to help you not mix up sensitivity and specificity. Those acronyms remind you that sensitivity quantifies how well a test finds (rules in) disease in those who have the disease and that specificity quantifies how well a test rejects (rule out) that diagnosis in people without the disease.

That's it! A mnemonic to prevent mixing up two terms.

The SPIN rule assumes that your false positives are super low and the the prevalence is not skewed and the proportions with all the numbers are similar. In this instance when you have a high specificity (and assuming essentially a zero false positive rate) AND you have a positive result, then people assume that this positive result has ruled in a disease.

In my opinion, the assumptions are too great and the rule is not the best system to use. I would suggest knowing the a priori risk and then using the test as a conditional factor to give you a better posterior risk.

BTW.. you can use the same logic for SNOUT.

• Can you clarify what you mean by "prevalence is not skewed"? PPV depends on prevalence, even for extremely sensitive and extremely specific tests. However, prevalence is just a proportion which may be high, middling, or low... in what sense do you mean "skewed"? Commented Apr 22 at 18:19