Mandatory medical review for preoperative sodium below 130 did not change 30-day outcomes

Design

Retrospective registry study, regression discontinuity at sodium 130 mmol/L

Population

24,803 adult surgical patients with preoperative sodium 125–135 mmol/L, Singapore, 2013–2020

Primary outcome

30-day mortality

Effect

No threshold effect on 30-day mortality (p = 0.138); discontinuity in 1-year mortality (p = 0.010)

A Singapore institution required internist referral and treatment for any patient with serum sodium below 130 mmol/L before surgery. Using registry data from 2013 to 2020, investigators took 24,803 adults with preoperative sodium of 125 to 135 mmol/L and used a regression discontinuity design, comparing patients just either side of the 130 threshold to estimate the effect of the protocol itself.

Sodium below 130 was a marker of risk: unadjusted 30-day mortality was 3.32% against 1.33%, one-year mortality 21.25% against 11.03%, and ICU admission and length of stay were higher too. But at the threshold, the protocol showed no effect on 30-day mortality, ICU admission or length of stay. There was a discontinuity in one-year mortality, most marked in patients with cardiac disease, chronic kidney disease, moderate operative risk or emergency surgery. Lower starting sodium predicted rapid postoperative correction.

The interpretation matters. Hyponatraemia before surgery flags sicker patients, but correcting the number did not make the operation safer in the short term. The one-year signal may reflect what an internist found and treated — heart failure, kidney disease, drugs — rather than the sodium. This argues for using a low sodium as a reason to look for its cause, not as a reason to delay surgery until it rises.

• Treat preoperative sodium below 130 mmol/L as a marker of underlying illness and look for its cause.
• Review diuretics, SSRIs, carbamazepine and other common drug causes before surgery.
• Do not assume that raising the sodium number alone will reduce perioperative risk.
• Watch for rapid correction after surgery, especially when the starting sodium is lowest.
• Patients with heart failure or chronic kidney disease may gain most from a medical review.

Why it matters

It separates the sodium as a marker of risk from the sodium as a target for treatment.

Don't overread it

The one-year mortality signal is a secondary finding and may reflect treatment of comorbidity, not sodium correction.

The statistics, in plain English

Regression discontinuity compares patients just above and just below a cut-off, who should be otherwise similar, so a jump at the threshold suggests the intervention caused it. No jump appeared for 30-day outcomes. The one-year effect is a secondary outcome with several subgroups, which makes chance and confounding more likely.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.