1/ Time for a Tweetorial on the BNP!
You are working up a patient with dyspnea on exertion, but you’re unsure whether it’s heart failure – in fact, your suspicion is 50/50! You order a BNP, which comes back elevated at 350 pg/mL.
This BNP level:
You are working up a patient with dyspnea on exertion, but you’re unsure whether it’s heart failure – in fact, your suspicion is 50/50! You order a BNP, which comes back elevated at 350 pg/mL.
This BNP level:
2/ Before we answer this question, let’s review what BNP can/cannot tell us by taking a look at this incredible table. It comes from a systematic review of tests used to diagnose acute heart failure in patients with dyspnea. source: onlinelibrary.wiley.com/doi/epdf/10.11… 
3/ If your brain exploded, you’re not alone – mine did too! But it’s fairly simple if we break it down. A positive likelihood ratio is the likelihood that a positive test appears in a patient WITH disease compared to the likelihood that it appears in a patient WITHOUT disease.
4/ LRs, then, are a great way to think about how clinically useful a diagnostic test is!
🔸LR>1: the test result supports the diagnosis
🔸LR<1: it makes it less likely
🔸LR = 1: the test has no diagnostic value.
🔸LR>1: the test result supports the diagnosis
🔸LR<1: it makes it less likely
🔸LR = 1: the test has no diagnostic value.
5/ This neat thing (called Fagan’s nomogram) allows us to compare pre- and post-test probabilities by drawing a line through the LR.
6/ Back to BNP: It turns out that a BNP level < 100 pg/mL is really useful in ruling OUT heart failure (LR = 0.14). Remember, your pre-test probability was 50%. With a BNP < 100, it is now below 10%! Nice!
7/ However, a BNP > 100 pg/mL doesn’t perform very well in ruling IN heart failure. The overall LR for BNP > 100 is only 2.2, meaning it only brings the pre-test probability up by about +15%. Why? B/c BNP is elevated in a bunch of other diseases, including kidney failure!
8/ Does that mean a high BNP level is always inconclusive? Not exactly. If your patient has a REALLY high BNP (1,000 and above), then your LR+ is closer to 8, which strongly supports the diagnosis of HF.
9/ Back to the original question: what does a BNP level of 350 pg/mL tell us in this patient? Basically… nothing. The LR for BNP = 350 is about 1!
(If your institution uses proBNP, the same principles apply, although very high proBNP levels are not as suggestive of HF)
(If your institution uses proBNP, the same principles apply, although very high proBNP levels are not as suggestive of HF)
10/ TL;DR:
👉BNP < 100 pg/mL = less likely to be heart failure
👉BNP > 1,000 pg/mL = more likely to be heart failure
👉100 < BNP < 1,000 = not very useful information!
That’s it for now. Thanks for reading!
👉BNP < 100 pg/mL = less likely to be heart failure
👉BNP > 1,000 pg/mL = more likely to be heart failure
👉100 < BNP < 1,000 = not very useful information!
That’s it for now. Thanks for reading!
