Mortality, Renal Replacement Therapy and Renal Impairment Markers in Critically Ill Patients in a Multinational Point-Prevalence ICU Cohort over 2007 to 2024: A nutritionDay Descriptive Multivariable Analysis
Arabella Fischer-Hammerschmied, Peter Bauer, Fiona Trumshi, Silvia Tarantino, Emir Softic, Christian Schuh, Sylvia Ryz, Martin H. Bernardi, Andrea Lassnigg, Johanna Tichy, Sarah Yadavalli, Anna Schranz, Michael Hiesmayr · 2026
Read Mortality, Renal Replacement Therapy and Renal Impairment Markers in Critically Ill Patients in a Multinational Point-Prevalence ICU Cohort over 2007 to 2024: A nutritionDay Descriptive Multivariable Analysis by Arabella Fischer-Hammerschmied, Peter Bauer, Fiona Trumshi, Silvia Tarantino, Emir Softic, Christian Schuh, Sylvia Ryz, Martin H. Bernardi, Andrea Lassnigg, Johanna Tichy, Sarah Yadavalli, Anna Schranz, Michael Hiesmayr (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13607511.txt.
Introduction
Arabella Fischer-Hammerschmied and co-authors study in 2026 how kidney harm links to death in intensive care patients around the world. Their data come from a project in which care units in many lands record all patients on one set day each year. The study spans the years from 2007 to 2024.
Kidney signs in the study were a waste marker in the blood, urine output at admission, and whether a patient got a blood cleaning treatment. The team weighed how each sign links to death, while they allowed for traits of the patient and of the unit. Long runs of shared data like these show trends that one hospital cannot see. They can also show where kidney care in intensive units has improved, and where it has stalled, across many health systems.
Catalog record
| CREATOR | Arabella Fischer-Hammerschmied, Peter Bauer, Fiona Trumshi, Silvia Tarantino, Emir Softic, Christian Schuh, Sylvia Ryz, Martin H. Bernardi, Andrea Lassnigg, Johanna Tichy, Sarah Yadavalli, Anna Schranz, Michael Hiesmayr |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 37 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Mortality, Renal Replacement Therapy and Renal Impairment Markers in Critically Ill Patients in a Multinational Point-Prevalence ICU Cohort over 2007 to 2024: A nutritionDay Descriptive Multivariable Analysis Authors: Arabella Fischer-Hammerschmied, Peter Bauer, Fiona Trumshi, Silvia Tarantino, Emir Softic, Christian Schuh, Sylvia Ryz, Martin H. Bernardi, Andrea Lassnigg, Johanna Tichy, Sarah Yadavalli, Anna Schranz, Michael Hiesmayr Published: 2026 ABSTRACT Background: The study goal was to assess the association between mortality and renal impairment markers in the nutritionDay ICU cohort, incorporating creatinine, urine volume, and renal replacement therapy (RRT) while adjusting for ICU- and patient-level factors. Methods: During nutritionDay, participating ICUs worldwide prospectively documented all patients treated on a predefined study day in November. Renal impairment markers included creatinine and urine volume at ICU admission and RRT status on nutritionDay. ICU-level factors included world region, calendar year and time of assessment. Patient-level factors included sex, age, BMI, disease severity, and underlying disease leading to ICU admission. Sixty-day hospital mortality and use of RRT on nutritionDay were assessed in multivariable logistic regressions including all ICU- and patient-level factors. Results: The post hoc analysis included 23,485 adult ICU patients from 1855 ICU-year units in 60 countries from 2007 to 2024, with analytic sample sizes of 23,485 for RRT, 18,245 for mortality, 21,054 for creatinine, and 20,753 for urine volume. Mean age was 61.4 (SD 17.0) years. Mean SAPS 2 at ICU admission was 39.8 (SD 17.7) points. Renal replacement therapy was associated with approximately twice the observed mortality. Severity of illness remained a major determinant of mortality after removing age and urine volume from SAPS 2. RRT use was similar up to 70 years but declined thereafter. Creatinine increased with age and BMI, while urine volume progressively declined with age. The reason for ICU dependency was associated with mortality, RRT use, creatinine, and urine volume. Conclusions: RRT, age and severity of illness were strongly associated with mortality in this large ICU cohort. The reason for ICU dependency, reflecting the underlying disease, was substantially associated with mortality, RRT use, creatinine, and urine volume. 1. Introduction Serum creatinine, urine volume, and renal replacement therapy (RRT) are markers of renal impairment in critically ill patients [1]. In critically ill patients, RRT was associated with higher mortality. Yet RRT is a life-saving therapy for renal failure [2]. The association between mortality and RRT may therefore reflect both renal impairment and its treatment. Large studies have described prevalence of RRT and associated mortality without including ICU- and patient-level factors [2,3] or the underlying disease leading to ICU admission [4]. Smaller studies focused on BMI and did not include ICU- and patient-level factors [5,6,7]. Previous studies investigating the association between renal impairment and mortality have lacked information on creatinine and urine volume [2,4,5,6]. Previous studies were often restricted to selected renal subgroups, introducing post-selection and limiting insight into how the observed range of renal impairment markers relates to mortality [4,5,6,7,8]. A comprehensive assessment of kidney function in the ICU requires more than RRT status. Creatinine levels and urine volume provide complementary information on renal impairment, as RRT may also be initiated for refractory fluid overload only [9]. This is the first multicentre ICU cohort analysis that jointly considers creatinine, urine volume and RRT and integrates ICU- and patient-level factors to assess the associations of RRT and other covariables with mortality within the same multivariable model [2,3,4,5,6,7,8]. Therefore, our goal was to quantify the association between mortality and renal impairment markers in a large worldwide cohort from the nutritionDay ICU project over 2007 to 2024, incorporating creatinine levels, urine volume, and RRT, and accounting for ICU- and patient-level factors. We hypothesised that mortality would be related to both renal impairment markers and a range of ICU- and patient-level factors. 2. Methods 2.1. Study Design and Population Briefly, nutritionDay in the ICU is an annual worldwide audit in which participating ICU health care providers prospectively record information on all ICU patients on a designated study day, which is called nutritionDay [10,11,12]. The nutritionDay project is a yearly repeated cross-sectional point-prevalence cohort study with 60-day outcome follow-up. This cross-sectional study design on a maximum of three possible inclusion days (first week: Thursday, second week: Wednesday, third week: Tuesday) in November captures the ICU patient population on that day. Each patient is recruited only once. nutritionDay therefore served as the study landmark and inclusion time. On nutritionDay, we assessed reasons for ICU dependency at admission, laboratory values obtained at ICU admission and on nutritionDay, and RRT use on nutritionDay. The nutritionDay questionnaire does not distinguish between intermittent and continuous RRT, or between acute and chronic RRT. Furthermore, the questionnaire does not assess indication, initiation time, duration, dose, net ultrafiltration, previous discontinuation, or subsequent initiation of RRT. Sixty days after nutritionDay, mortality, ICU and hospital length of stay were assessed. The ICU nutritionDay questionnaire used between 2007 and 2024, including all assessed variables, is provided in the Supplementary Materials. The participating ICUs were in Europe (50%), North America (6%), Latin America (18%) and Asia (25%) (Table S1, Supplementary Materials). In the present post hoc analysis, all adult ICU patients who participated between 2007 and 2024 were included. We excluded patients younger than 18 years (n = 313) or with missing ICU- or patient-level factors essential to the analysis, such as age (n = 49), sex (n = 398) or region (n = 88). Exclusions were applied sequentially, with each patient assigned to only one exclusion category; thus, the exclusion categories were mutually exclusive. Missing values denote variables not documented on the study day. African regions (Ethiopia (n = 26), Egypt (n = 1)) were excluded because of the small African sample size (Figure 1). Missing values for outcomes are indicated in Figure 1. The nutritionDay project was approved by the Ethics Committee of the Medical University of Vienna (EK1080/2020). Additional ethical clearance was obtained from local or national committees in the participating countries as required. The nutritionDay project is registered at ClinicalTrials.gov (NCT02820246). During the preparation of this work, the authors used ChatGPT version 1.2026.048 to assist with looking up specific R (version 4.3.3) codes, troubleshooting R errors and improving the formulation of selected sentences. 2.2. Outcomes The primary outcomes were 60-day hospital mortality after nutritionDay and renal replacement therapy (RRT) on nutritionDay. Because mortality was analysed as a binary 60-day hospital outcome rather than as a time-to-event outcome, no additional censoring rules were applied. RRT use included both continuous and intermittent RRT on nutritionDay. Additional outcomes were serum creatinine and urine volume as renal impairment markers at ICU admission. These additional outcomes were chosen to identify ICU- and patient-level factors associated with renal impairment markers already at ICU admission, rather than only on nutritionDay or 60 days later. 2.3. Covariables Considered for Analyses Covariables included ICU- and patient-level factors. The covariables included in the models were selected a priori as the clinically most relevant factors with a plausible temporal relationship to mortality and RRT. Missing values for independent covariables were retained and reported as a category for each covariable, because missingness may be informative [11]. Independent continuous covariables were categorised to provide clinically easily interpretable estimates without assuming a linear association and to allow inclusion of missing-value categories. Categories were specified before fitting the