3 doubts i have (sensitivity, Cruptococcus, Abuse)

[Kobebucsfan]

---

Members don't see this ad.

1) how to differentiate whether a child patient has been physically abused or its a defect. If a kid came in with bruises and history of broken bones. obviously if he has blue sclera aling with the symptoms, then he has a defect. what other clues would the question have for abuse vs. defect. I remember some questions stating that the patient has ribs that healed or something.


2) Cryptococcus vs. Toxoplasma. so they both cause ring lesions in the brain correct ? What clues in the question separates them ? I know Toxoplasma has to do with cats. what else tho


3) When it comes to Sensitivity and Specificity, i know how to calculate the formulas and such but what i dont get is the Sensitivity test negative - rules out a disease. and Specificity test positivity rules IN disease. what does that mean

 

[Jabbed]

1) how to differentiate whether a child patient has been physically abused or its a defect. If a kid came in with bruises and history of broken bones. obviously if he has blue sclera aling with the symptoms, then he has a defect. what other clues would the question have for abuse vs. defect. I remember some questions stating that the patient has ribs that healed or something.


2) Cryptococcus vs. Toxoplasma. so they both cause ring lesions in the brain correct ? What clues in the question separates them ? I know Toxoplasma has to do with cats. what else tho


3) When it comes to Sensitivity and Specificity, i know how to calculate the formulas and such but what i dont get is the Sensitivity test negative - rules out a disease. and Specificity test positivity rules IN disease. what does that mean

2) Pigeon exposure for crypto, cat feces/pork for toxo. When in doubt, it's toxo (more common than crypto in both HIV+/- patients). Honestly though ring-enhancing lesions on CT have a very broad differential and it's nigh impossible to differentiate the cause if you're only looking at a CT.

3) SPIN: If a high specificity test is positive, you can be reasonably sure that the patient has the disease. You rule the diagnosis in.
SNOUT: If a high sensitivity test is negative, you can be reasonably sure that the patient does not have the disease. You rule the diagnosis down/out.

Caveats: Both of these concepts need to be weighed against the idea of pre-test probability (i.e., the chance that the patient has the disease before testing). A high pre-test probability and a negative high sensitivity test still equals a high chance of disease. A low pre-test probability and a positive high specificity test still equals a low chance of disease. The core point here is that any sort of test has limitations to its overall accuracy.

Generally: A high sensitivity test is used to rule out disease when pre-test probability is low. A high specificity tst is used to rule in disease when pre-test probability is average/high.

 

[dempty]

While the general rule-of-thumb is O.K. to answer #3, it might be helpful to think a little deeper about the meaning of specificity and sensitivity.


Sensitivity is the ability of a test to classify patients as having the disease when they actually have the disease. In other words: the number of correctly positive test results over the number of truly sick patients who were tested: TP / (TP+FN). If you're one for probability, it's simply the probability of a positive test result given the patient is ill; P(+|Sick). If this value is close to 1 (100%), the test is very likely to classify a truly sick person as sick (positive test). So, we know beforehand, if you are sick, it's very likely the test will say you are sick. Once we administer the test, and if it doesn't say you're sick (yields a negative result), it's more probable that you don't have the disease (rule it out). Layman's terms: If I'm very good at calling a dog a dog, and I can't call this animal a dog, it's probably not a dog (rule out dog). Another way to look at this is from the false negative perspective; saying a patient is healthy, when the patient is actually sick. If the sensitivity is high, this means the false negative rate (probability) is low (1-sensitivity= false negative rate). If it's very unlikely to be called healthy when you're actually sick, and if that test shows you a healthy result, you can feel more comfortable that you don't have the disease (rule out).


Specificity is the ability of a test to classify patients as healthy (negative result) when they are actually healthy (no disease). Following the same process as before: the number of correctly negative test results over the number of truly healthy patients: TN / (TN+FP). In probability notation: it's the probability of a negative test result, given the patient is healthy; P(-|Healthy). If this value is close to 1 (100%), the test is very likely to classify a truly healthy person as healthy (show a negative test result). We have a test that is very good at calling truly healthy people healthy, and if it doesn't say you're healthy (it yields a positive result, indicating disease), it's more probable that you do have the disease (rule it in). Layman's terms: If I'm very good at saying what isn't a zebra, and I can't say this animal isn't a zebra, then it's more likely that this animal is a zebra (rule in zebra). And, to be complete, we can look at this from the false positive perspective: calling a healthy patient sick (getting a positive result). A high specificity implies a low false positive rate (probability) (1-specificity= false positive rate). If your probability of being called sick when you are truly healthy is low, and you receive a sick test result (positive), then it's more likely you're sick (rule in).

Realistically, though, you want a test to be highly specific and highly sensitive. Putting too much stock in a test with only high sensitivity or only high specificity isn't likely to lead to optimal decisions.

I hope this helps build a little more intuition behind these terms.

 

[worldbeater]

) how to differentiate whether a child patient has been physically abused or its a defect. If a kid came in with bruises and history of broken bones. obviously if he has blue sclera aling with the symptoms, then he has a defect. what other clues would the question have for abuse vs. defect. I remember some questions stating that the patient has ribs that healed or something.

Before giving a dx of child abuse, rule out:
-Osteogenesis imperfecta
-Bleeding disorders
-Fifth disease
-Mongolian spots

Child abuse will present with:
-Mulitple ecchymoses
-Retinal hemorrahage
-Epidural/subdural hemorrhage
-spiral fractures (twisted)
-Multiple fractures in different healing stages (there are your healed ribs)

 

[worldbeater]

2) Cryptococcus vs. Toxoplasma. so they both cause ring lesions in the brain correct ? What clues in the question separates them ? I know Toxoplasma has to do with cats. what else tho

Toxoplasmosis is present from cat urine as the poster above mentioned and presents in the parietal lobe. If I see ring enhances lesions I would go with Toxoplasma first and then Cryptococcus in that order.

3) When it comes to Sensitivity and Specificity, i know how to calculate the formulas and such but what i dont get is the Sensitivity test negative - rules out a disease. and Specificity test positivity rules IN disease. what does that mean

This has been explained pretty well above. One thing to remember is for biostat questions when your working it out is to make sure you line up everything in the correct order before filling out each box on your writing source (sensitivity and specificity on the horizontal axis & test+ and test - on the vertical axis), because sometimes they give the values in the question in the wrong order to make the question harder.