A Reassessment of Sodium Correction Rates and Hospital Length of Stay Accounting for Admission Diagnosis

Background Slow correction of severe hyponatremia has been historically recommended due to the risk of rare but catastrophic neurologic events with rapid correction. A recent study challenging this paradigm reported that rapid correction is associated with shorter hospital length of stay, but that study did not control for admission diagnosis. The objective of this study was to determine whether rapid correction is associated with shorter length of stay when controlling for admission diagnosis. Methods This retrospective cohort study is based on the fourth edition of the Medical Information Mart for Intensive Care, MIMIC-IV, a deidentified, publicly available clinical research database which includes admissions from 2008-2019. Patients were identified who presented to the hospital with initial sodium <120 mEq/L and were categorized according to total sodium correction achieved in the first day (<6 mEq/L; 6-10 mEq/L; >10 mEq/L). Linear regression was used to assess for an association between correction rate and hospital length of stay, and to determine if this association was significant when controlling for admission diagnosis classifications based on diagnosis related groups (DRGs). Results There were 419 patients with severe hyponatremia (<120 mEq/L) included in this study, of whom 374 survived to discharge. Median [IQR] hospital length of stay was 6 [4, 11] days. In a univariable linear regression, there was a trend towards a significant association between the highest rate of correction (>10 mEq/L) and shorter length of stay, as compared with a moderate rate of correction (coef. -2.764, 95% CI [-5.791, 0.263], p=0.073), but the association was not significant when controlling for admission diagnosis group (coef. -1.561, 95% CI [-4.398, 1.276], p=0.280). There was a significant association in the survivor subset (coef. -3.455, 95% CI [-6.668, -0.242], p=0.035), but it was also not significant when controlling for admission diagnosis group (coef. -2.200, 95% CI [-5.144, 0.743], p=0.142). Conclusions Rapid correction is not associated with shorter length of stay when controlling for admission diagnosis, suggesting that the disease state confounds this association. Findings from prior and future studies reporting this association should not drive clinical decision making if the confounding effect of hospital admission diagnosis and competing risk of death are not fully accounted for.