For years, guidance for correcting sodium levels in patients hospitalized with severe hyponatremia — defined as serum sodium concentration < 120 mEq/L — has been to go slow, in part to avoid neurologic damage. But new research published in the Annals of Internal Medicine suggests that sodium correction could be kicked up a notch.

In a retrospective chart review, investigators with Kaiser Permanente’s CREST Network found faster sodium correction to be associated with a lower risk for 90-day mortality or delayed neurologic events.

Researchers reviewed records from 13,988 patients hospitalized with severe hyponatremia between 2008 and 2023 at 21 community hospitals that are part of Kaiser Permanente Northern California’s health system with a lot of practice variation in management of hyponatremia.

Compared with slow sodium correction, defined as < 8 mEq/L, both medium correction, defined as 8-12 mEq/L (risk difference [RD], -5.6 percentage points; 95% CI, -7.1 to -4.0 percentage points), and fast correction, defined as >12 mEq/L (RD, -9.0 percentage points; 95% CI, -11.1 to -6.9 percentage points) were associated with lower adjusted risk for their primary outcome of 90-day death or delayed neurologic events such as new demyelination, paralysis, epilepsy, or altered consciousness between 3 days and 90 days from admission.

There has been a longstanding debate regarding the optimal treatment of hyponatremia, said lead study author Dustin Mark, MD, a senior physician with Kaiser Permanente, in Oakland, California, and adjunct investigator in the health system’s research division.

Clinical practice guidelines from American and European physicians in 2013 and 2014 emphasized slow, controlled correction to avoid “potentially devastating neurologic damage from osmotic demyelination syndrome (ODS),” Mark said, with recommended sodium correction targets “around 8-12 mEq/L over 24 hours” but as slow as 4-6 mEq/L per 24 hours.

Some recent studies “have refuted the current thinking that slow is better,” he said, indicating that limiting sodium correction was associated with higher mortality and longer lengths of stay. ODS, “which is rare and somewhat idiosyncratic,” doesn’t always link to the rate of sodium correction, he added, sometimes happening regardless.

Mark and colleagues reviewed medical records from adults admitted from the emergency department with an initial serum sodium concentration ≤ 120 mEq/L. Order set guidelines emphasized treatment based on symptoms, he said. Patients had a median age of 74 years; 63% were women. Comorbidities included congestive heart failure (24%), liver disease (18%), alcohol dependence (14%), and metastatic cancer (10%).

Investigators calculated 24-hour sodium correction rates using one or two time-weighted sodium measurements obtained within 18-30 hours after a given reference sodium value. They determined the maximum 48-hour sodium correction rate using one or two sodium measurements obtained 42-54 hours after a given reference sodium value.

Delayed neurologic events were defined as new (not coded in the past 12 months) and delayed (entered between 3 days and 90 days from admission) International Classification of Diseases (ICD)-9 and ICD-10 diagnoses of demyelinating diseases of the central nervous system.

Slow sodium correction occurred in 5980 (43%) patients, medium in 3609 (26%) patients, and fast in 2551 (18%) patients. Sodium correction rate could not be calculated for 1848 (13%) patients. Patients with slow or missing correction rates had more chronic illness and physiologic derangements at admission than those with medium or fast correction rates.

Conversely, patients with fast correction compared with slow spent 1 less day in the hospital on average and were more likely to have initial serum sodium values ≤ 115 mEq/L, to be admitted to the ICU, to have a principal diagnosis of hyponatremia, and to receive intensive sodium-directed treatments.

The primary composite outcome occurred in 3000 patients (21%). Secondary outcomes of 90-day death occurred in 2554 patients (18%); 90-day delayed neurologic events occurred in 587 patients (4%), and in-hospital death occurred in 858 patients (5%). Coma or altered consciousness accounted for 60% of delayed neurologic events (355 of 587); there were 19 coded cases of demyelinating disease.

Investigators looked at neurologic sequelae because ODS “isn’t clearly always caught,” Mark said. “It’s defined based on MRI and clinical scenario, but there really is no gold standard test for it,” he said.

The key, Mark said, is that you note neurologic symptoms, and the net is cast broad for a new diagnosis of a neurologic symptom that might be consistent with ODS in case it might be missed. “[We] found that whether you looked at that alone, in combination with mortality, or mortality alone, all…were worse the slower you went.”

The study is one of several in an ongoing controversy positing that rapid correction of hyponatremia reduces mortality, said Richard Sterns, MD, professor emeritus of medicine at the University of Rochester School of Medicine and Dentistry, Rochester, New York. Sterns co-authored the 2013 US sodium correction guidelines, authored the book Hyponatremia and 40 Years of Controversy: Looking Back, and is part of an international study group called PRONATREUS that is looking at sodium correction in hyponatremia.

Studies such as this one tend to conflate death with death specifically from hyponatremia, he said. Patients in this and other studies had advanced heart failure, cancer, and liver disease, which can cause hyponatremia and death, Sterns said, “and they’re also associated with slow correction rates.”

Additionally, this study used ICD codes that have been shown in other previous studies to be “unreliable” for identifying ODS, he said. For example, one study of about 1500 patients found only nine patients with ODS, but none had an ICD-9 diagnosis. Instead, Sterns said, they were found by reviewing MRI scans, which the current study did not do.

“One of the things that I would agree on with these authors is that we need more data, and better data, but the [studies] need to be done in a way that will answer the question,” he said.

This study was funded by the Permanente Medical Group Rapid Analytics Unit Program. The interviewees reported no relevant financial disclosures. 

Karen Blum is a freelance medical/science writer in the Baltimore area.