Fluid Restriction Reduces Delayed Hyponatremia and Hospital Readmissions After Transsphenoidal Surgery.

Postoperative hyponatremia leads to prolonged hospital length of stay and readmission within 30 days. To assess 3 strategies for reducing rates of postoperative hyponatremia and analyze risk factors for hyponatremia. Two retrospective analyses and 1 prospective study. Tertiary referral hospital. Patients undergoing transsphenoidal surgery for pituitary adenomas and other sellar and parasellar pathologies. Phase 1: no intervention. Phase 2: postoperative day (POD) 7 sodium testing and patient education. Phase 3: fluid restriction to 1 L/day on discharge in addition to phase 2 interventions. Rates of early and delayed hyponatremia and readmissions. Secondary outcomes were risk factors for hyponatremia and readmission costs. In phase 1, 296 patients underwent transsphenoidal surgery. Twenty percent developed early and 28% delayed hyponatremia. Thirty-eight percent underwent POD 7 sodium testing. Readmission rates were 15% overall and 4.3% for hyponatremia. In phase 2 (n = 316), 22% developed early and 25% delayed hyponatremia. Eighty-nine percent complied with POD 7 sodium testing. Readmissions were unchanged although severity of hyponatremia was reduced by 60%. In phase 3 (n = 110), delayed hyponatremia was reduced 2-fold [12.7%, relative risk (RR) = 0.52] and readmissions 3-fold [4.6%, RR = 0.30 (0.12-0.73)]; readmissions for hyponatremia were markedly reduced. Hyponatremia readmission increased costs by 30%. Restricting fluid to 1 L/day on discharge decreases rates of delayed hyponatremia and readmissions by 50%. Standardized patient education and POD 7 sodium testing decreases severity of hyponatremia but does not impact readmission rates. These protocols should be considered standard practice for patients undergoing transsphenoidal surgery.

Informations de copyright

© The Author(s) 2023. Published by Oxford University Press on behalf of the Endocrine Society. All rights reserved. For permissions, please e-mail: [e-mail].