Changes in clinical practice patterns and outcomes of hyponatremia after implementation of electronic alerts for electrolyte disturbances

Abstract Background Electrolyte disturbances are common but frequently overlooked. We investigated if clinical practice and outcomes changed after implementation of electronic alerts (e-alerts) for electrolyte disturbances. Methods This single-center, retrospective cohort study was conducted at a university hospital that implemented the e-alert system, ‘Electrolyte Busters’, in May 2021. E-alert criteria were serum sodium < 125 or >160 mEq/L, potassium < 2.5 or > 6.0 mEq/L, corrected calcium < 7.5 or > 11.5 mg/dL, and magnesium < 1.0 or > 4.0 mg/dL. Outcomes among patients with hyponatremia were compared between the years before and after e-alert implementation. Urinary chemistry measurements, treatment, and overcorrection were analyzed using logistic regression. Referral and fall rates were analyzed using incidence rate ratios (IRRs) and negative binomial regression, respectively. Length of stay (LOS) was compared using the Mann-Whitney U test. Results Data from 36 379 pre-intervention and 37 395 post-intervention patients were analyzed. Median age was 62 years, and 48% were male. After e-alert implementation, nephrology referrals for hyponatremia increased from 1.1 to 2.3 per month (IRR, 95% confidence interval [CI]: 2.07 [1.04–4.36]). Urinary chemistry measurements increased (odds ratio [OR], 95% CI: 1.62 [1.08–2.44]), whereas untreated cases decreased (OR: 0.31 [0.19–0.53]). Median (interquartile range) LOS decreased from 27 (15–47) to 20 (12–36) days (P = 0.01). Hyponatremia overcorrection (OR: 0.64 [0.36–1.14]) and fall rates (IRR: 0.82 [0.36–1.69]) showed downward trends. Conclusion E-alerts were significantly associated with improved diagnostic evaluation, therapeutic interventions, and shorter LOS among patients with hyponatremia.