1/ I find the over reliance on the Adrogue-Madias (AM) formula to predict SNa changes after infusion of a specific IV fluid astounding. #hyponatremia tweetorial
2/ First of all, AM and other formulas (eg. Free Water Deficit formula) are simple mathematical manipulations of the simplified Edelman equation 
3/ A study showed that the AM formula underestimated the true rate of correction in up to 75% of patients with severe hyponatremia when given 3% saline. In many of them, by a factor of 5. pubmed.ncbi.nlm.nih.gov/17913972/
4/ Another study looking at the accuracy of 4 formulas (including AM) in predicting SNa changes using 5 lab data sets demonstrated that the root mean square error (RMSE), a measure of how spread out the residuals are, was 4.79-6.37 mEq/L for all sets. pubmed.ncbi.nlm.nih.gov/27478591/
5/ In the same study above, Bland-Altman analysis for AM formula showed high variability and discrepancies between the predicted and actual Na. pubmed.ncbi.nlm.nih.gov/27478591/
6/ Rick Sterns had a great editorial to this paper above. pubmed.ncbi.nlm.nih.gov/27478590/
7/ There are 2 main reasons why AM and other formulas fail:
1. Some formulas assume a close compartment and do not account for urinary loses of water and electrolytes (Imagine trying to estimate SNa change with AM after giving NS to a patient with hypovolemic hyponatremia)
1. Some formulas assume a close compartment and do not account for urinary loses of water and electrolytes (Imagine trying to estimate SNa change with AM after giving NS to a patient with hypovolemic hyponatremia)
8/ 2. Even for the formulas that account for urinary losses, they assume static conditions, when in reality urine composition can change on an hour-to-hour basis.
9/ AM in particular is designed to estimate SNa changes after 1 L of IV fluid is infused and then extrapolate this to calculate volume for a desired SNa change. However, recent data showed that this assumption is wrong and AM calculations are not scalable. kidney360.asnjournals.org/content/2/2/36…
10/ AM might be difficult to calculate for the not-math inclined people, so instead we suggest a simple weight based calculation for 3%: 1-1.5 mL/kg over 6h (for an 80 kg person, 80 x 1.5 = 120 mL over 6h for a rate of 120/6 = 20 mL/h pubmed.ncbi.nlm.nih.gov/34508830/
11/ Finally, there is no formula that can substitute frequent SNa checks (eg. every 2h) and then change rate of infusion as needed
• • •
Missing some Tweet in this thread? You can try to
force a refresh
