
Premium content
Access to this content requires a subscription. You must be a premium user to view this content.

poster
Lactulose-induced hypernatremia in cirrhotic patients with hepatic encephalopathy
Background: Lactulose is an osmotic laxative used to treat hepatic encephalopathy (HE) in patients with cirrhosis. Here, we examined lactulose-induced hypernatremia (HN) in patients with cirrhosis being treated for HE.
Methods: We identified all patients admitted with cirrhosis and HE at our institution from 1/1/18 to 12/31/23. Treatment details and complete clinical data were abstracted. Matching, followed by propensity score calculation, was used to control for the severity of underlying liver disease using a case control study design (HN patients were matched to patients with HE, but without HN). The primary endpoint was hospital mortality, and secondary endpoints included total and rectal lactulose dose (g), time to mental status resolution, and hospital length of stay (LOS).
Results: Among 6,742 patients admitted with cirrhosis, 1,809 (27%) were admitted with HE as the primary diagnosis. Of these, 158 patients with liver transplantation were excluded. Of the remaining 1,651 patients, 347 patients (21%) had a sodium level > 145 mEq/L, and 106 patients (6%) had a sodium ≥ 150 mEq/L (hypernatremia (HN) group). Cirrhosis was most commonly caused by alcohol or MASH; the median MELD was 24. In HN patients, the median admission sodium was 137 and the peak was 152; all patients received lactulose. The median average daily lactulose dose to peak sodium was higher in the HN group than in controls (85 vs. 38 g, p<0.001). 67 (63%) patients in the HN group received rectal lactulose and had a median sodium level of 154 mEq/L. Rectal lactulose was the strongest independent predictor of hypernatremia (OR = 4.469, p<0.001). The time to mental status resolution and LOS were 2.4 and 2.4 times longer, respectively, in the HN than in the control group (both p<0.001). Overall, 64 patients (60%) in the HN group died compared to 37 (35%) in the group without HN (p<0.001). Hypernatremia was an independent predictor of mortality (OR = 3.596, 95% CI 1.712 – 7.552, p <0.05).
Conclusion: Lactulose-induced HN is associated with poor outcomes, and patients receiving rectal or high dose lactulose had a remarkably high mortality rate. The data lead us to urge careful monitoring of sodium levels in patients with HE who are treated with lactulose.