(344) Is It Worth Its Salt? Maintenance Fluid Selection in DKA

Introduction: The American Diabetes Association recommends that either normal (NS) or half-normal (1/2NS) saline be administered continuously until resolution of DKA. Recently, there has been increasing interest in optimal intravenous (IV) fluid selection for resuscitative purposes, sparked by an association between NS and acute kidney injury (AKI). However, there is little data pertaining to fluid selection and its effect on clinical outcomes in DKA. Protocols at our institution require selection of either lactated ringers (LR) or 1/2NS as maintenance fluid for DKA management. It is unclear if one is superior to the other.

Methods: We retrospectively reviewed all medical intensive care unit (MICU) patients with DKA between 2008-2021 who received at least 0.5 liters (L) of either fluid. Patients who received similar amounts of both fluids were excluded.

Results: 515 patients were included. 447 (72%) predominantly received 1/2NS. Baseline characteristics were similar between the two groups except gender (52% male in 1/2NS group, 63% in LR group, p< 0.02) and lactate levels (2.5 ± 2.1 in 1/2NS group, 3.5 ± 3.6 in LR group, p=0.005). At 48 hours, patients in the 1/2NS group received 6.0L (±3.7L) of 1/2NS and 1.7L (±1.1L) of LR compared to 1.4L (±1.3L, p< 0.00001) and 5.2L (±2.6L, p< 0.00001), respectively, in the LR group. Both groups received similar volumes of normal saline (1.2L ± 0.9L in 1/2NS group, 1.1L ± 0.8L in LR group, p = 0.44). Mean time to DKA resolution, insulin drip duration, and LOS for MICU and hospital in the 1/2NS group were 0.6 (±0.3) days, 0.8 (±0.7) days, 1.8 (±1.6) days, and 4.7 (±4.2) days, respectively, compared to 0.5 (±0.3, p=0.44), 0.6 (±0.5, p=0.08), 2.2 (±3.4, p=0.11), and 6.6 (±11.8, p=0.006) in the LR group. Serum creatinine and anion gap levels were similar between the two groups.

Conclusions: Hospital LOS was 29% shorter for patients who primarily received 1/2NS compared to LR for DKA management. No difference in time to DKA resolution, duration of insulin drip, or MICU LOS was found between the two groups. Creatinine and anion gap levels were similar regardless of fluid choice. Randomized control trials could further delineate causation between fluid choice and hospital LOS.