09-07-2025
Hyponatremia: Is Rapid Correction a Risk or Relief?
Sodium correction limits have been around since bed rest for back pain. The latter is gone but the former persists. While bed rest simply delays recovery, rapid sodium correction may cause osmotic demyelination syndrome (ODS). ODS can manifest as central pontine myelinolysis (CPM), otherwise known as the 'locked in' syndrome. The patient is cognitively aware but unable to move. It's equivalent to paralytics without sedation. CPM paranoia has slowed infusion rates for decades.
As an intern, I was warned about CPM and made to read a seminal hyponatremia review published the previous year. The authors cautioned against sodium correction rates greater than 8 mEq/L over 24 hours. The recommendation was juxtaposed with a frightening cartoon diagram of a swollen brain with the label 'osmotic demyelination.' The result of rapid correction of course. After a critical care fellowship and tour managing head trauma in Afghanistan, my aversion to rapid correction remained intact. I never bothered to check whether CPM was an apocryphal tale or an evidence-based phenomenon.
Several studies, all of them observational, show rapid correction is associated with ODS. Association means little without context, though. For example, in a large series of patients with admission sodium levels < 120 mEq/L (41% of whom had a correction rate > 8 mEq/L over the first 24 hours), the ODS incidence was only 0.5%. ODS is clearly conditional; rapid correction alone isn't enough. Additional ODS risk factors include the severity and chronicity of the hyponatremia (lower and longer equates to higher risk) along with alcoholism, malnutrition, hypokalemia, and liver disease.
Guidelines and reviews endorse conservative correction. While the definition of 'conservative' varies, most recommend 6-10 mEq/L over the first 24 hours. Short of engaging System 2 thinking, there's ample reason to adopt the practice. Besides, apocryphal or not, ODS gets blamed on you if the sodium goes up by more than 6 mEq/L overnight.
A study published last year flips the conventional wisdom. The authors conducted a systematic review and meta-analysis on sodium correction rates and found a dose-dependent relationship with hospital length of stay (LOS) and mortality. As opposed to what I'd read and been taught, the relationship was inverted. Rapid correction decreased mortality and LOS. Correction rates weren't associated with ODS.
What now? Is this a casus belli for war on correction limits? Probably not. The study is well done, but don't be fooled by the fancy title. They included 16 studies; all were observational, convenience samples and 14 were retrospective. Combining poor quality studies is like bundling faulty mortgages in a collateralized debt obligation circa 2008. The result can serve as a statistical weapon of mass destruction, where uncertainty is amplified under a veneer of respectability.
There's also type 2 error when examining ODS rates. Of the 11,811 patients included across all 16 studies, only 32 (0.3%) experienced ODS. This may reflect the true incidence, or it could be an underestimate due to a lack of systematic assessment. Either way, statistical power is a problem. It's also notable that the only subgroup analysis they were able to perform was for those with alcoholism — and the only outcome analyzed was ODS (no difference). So, for those with comorbid alcoholism, hypokalemia, malnutrition, or liver disease we're not sure how correction rates affect LOS and mortality.
The study highlights how little we know. The authors did an excellent job considering what they were working with. This is as evidence-based as it's going to get. Not sure how others will interpret this data, but I think it confirms ODS is rare and conservative sodium correction rates might lead to harm. In the absence of high-risk features, I'll be liberalizing my sodium infusions, a little, particularly when baseline sodium is > 120 mEq/L.
