All CS patients treated in the cardiac intensive care unit (ICU) of the Ludwig-Maximilians-University hospital between January 2010 and August 2022 were included in the LMUshock registry (WHO International Clinical Trials Registry Platform Number DRKS00015860). Data collection and analysis was in accordance with the Declaration of Helsinki and German data protection laws. The study was approved by the local ethics committee (IRB number: 18–001). CS was defined by ESC guidelines, the IABP-SHOCK II trial and CULPRIT SHOCK trial [13,14,15]. Telephone follow-up was performed for patients dismissed from the hospital before day 30.

From 2018 on, the default maintenance and resuscitation fluid in our institution was switched from saline (0,9% sodium chloride, B. Braun SE, Melsungen, Germany) to a balanced resuscitation fluid (Jonosteril®, Fresenius Kabi, Bad Homburg, Germany). Jonosteril® is a balanced crystalloid buffered with acetate that has electrolyte concentrations closer to those of blood plasma than saline (Jonosteril®: Na + 137 mmol/l, Cl − 110 mmol/l, K + 4 mmol/l, Ca2 + 1.65 mmol/l, Mg2 + 1.25 mmol/l, acetate 36.8 mmol; Saline: Na + 154 mmol/l, Cl− 154 mmol/l). This decision was driven by the results of the SMART trial and was upheld since then [8]. However, a different choice of resuscitation fluid was allowed at the discretion of the attending intensive care physician. Patients receiving Jonosteril® at the day of admission to ICU were assigned to the balanced group, while patients receiving exclusively unbalanced fluid as resuscitation fluid were assigned to the saline group. Fluids required for dilution of parenteral medication were not analyzed due to missing data. Output data was available as total fluid output including urine, feces and iatrogenic output such as pleuracentesis or dialysis.

The primary endpoint was defined as all-cause 30-day mortality rate. Secondary endpoints included time on ICU, serum creatinine, new onset of renal-replacement therapy, time-weighted catecholamine doses (using the formula: dobutamine (mg/h) + 100 × epinephrine (mg/h) + 100 × norepinephrine (mg/h) as previously described), blood pressure, arterial blood pH, bicarbonate, serum chloride concentration [16]. In patients after OHCA, peak neuron specific enolase (NSE) within 3 days after resuscitation was assessed as a predictor of neurologic outcome in resuscitated patients [17].

Statistical analysis was performed in accordance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) statement using the software R (version 4.2.2, The R foundation, Vienna, Austria) [18]. Normally distributed continuous variables were depicted as mean with standard deviation and non-normally distributed continuous variables as median with interquartile ranges. Students t-test and Mann–Whitney-U test were used to compare groups as appropriate. Categorical variables were reported as absolute numbers and percentages and Chi-squared test was utilized for comparison. All tests were 2-tailed, and p-values < 0.05 were considered significant. Mortality rates were calculated using the Kaplan–Meier method and comparisons were made by log-rank tests. Missing data for body mass index and Simplified Acute Physiology Score 2 was imputed by R package “mice” (version 3.13.0) with predictive mean matching. Univariate and multivariate cox regression analysis was performed to identify predictors for the primary endpoint of 30-day all-cause mortality. Parameters for multivariate analysis were stepwise selected by Akaike information criterion (AIC) with backward direction and 1000 bootstrap iterations using the stepAIC function of the R-package MASS (version 7.3–58.1).

R-package MatchIt (version 4.5.0; Ho, Imai, King, and Stuart) was used for propensity score matching with a 1:1 nearest neighbour algorithm, no replacement, a 0.1 caliper and the following variables:[19]. age, sex, estimated glomerular filtration rate, OHCA, duration of out-of-hospital cardiopulmonary resuscitation, myocardial infarction (STEMI, NSTEMI) as well as mechanical ventilation, VA-ECMO, percutaneous transvalvular microaxial flow pump (Impella®), catecholamine doses and serum lactate at admission. Subgroup analyses were conducted for patients after OHCA as well as for patients receiving more than the median hourly volume of resuscitation fluids during the first day. The subgroup of OHCA patients was then divided into two groups based on the median serum lactate at admission to assess OHCA patients with severe shock.