Volume removal intolerance during net ultrafiltration in patients with acute kidney injury in the intensive care unit: protocol for a prospective multicentre observational study (VINKO) Authors: Gonzalo Ramírez-Guerrero, Cristian Pedreros-Rosales, Pablo Galindo, Rodrigo Ulloa, Jonathan Alarcón, Sandra Ávila, Eduardo Kattan, Paulo Melo, David Ballesteros, Raúl Valenzuela, Gabriela Guimaraes, Ricardo Ferrada, Charlotte Pfeifer, Vivian Villalobos, Gianpiero Mas, Matteo Marcello, Gerd Klinkmann, Claudio Ronco Published: 2026 ABSTRACT AbstractIntroductionFluid overload in critically ill patients with acute kidney injury (AKI) is independently associated with increased mortality, prolonged mechanical ventilation and impaired renal recovery. Net ultrafiltration (UFNET) during continuous renal replacement therapy (CRRT) is a cornerstone de-resuscitation strategy; however, tolerance to fluid removal is highly variable and difficult to predict. Ultrafiltration intolerance, encompassing haemodynamic instability, peripheral perfusion deterioration and tissue hypoperfusion during UFNET, lacks a standardised definition and has not been prospectively evaluated using a comprehensive multiparametric monitoring strategy. The Volume removal Intolerance during Net ultrafiltration in acute Kidney injury patients, an Observational study (VINKO) aims to identify clinical, haemodynamic, ultrasound-based, perfusion-related and biochemical factors associated with the development of ultrafiltration intolerance in critically ill patients with AKI undergoing CRRT.Methods and analysisVINKO is a prospective, multicentre, observational analytical case–control study enrolling adult intensive care unit patients with AKI requiring CRRT with prescribed UFNET. Cases are defined as patients who develop ultrafiltration intolerance during the 24-hour observation period, whereas controls are patients who tolerate UFNET without meeting intolerance criteria. Serial assessments include macrohaemodynamic variables, dynamic functional tests (passive leg raising and sitting manoeuvre), focused echocardiography, venous congestion assessment using the Venous Excess Ultrasound protocol, peripheral perfusion markers (capillary refill time, mottling score and peripheral perfusion index) and biochemical markers including lactate, albumin, ionised calcium and B-type natriuretic peptide/N-terminal pro-B-type natriuretic peptide when available. Measurements are obtained at predefined time points from UFNET initiation through 24 hours, with additional off-schedule assessments triggered by intolerance events. The primary objective is to identify independent predictors of ultrafiltration intolerance. A total of 128 patients (64 cases and 64 controls) will be enrolled based on a predefined sample size calculation.Ethics and disseminationThe study has been approved by the Scientific Ethics Committee of the Talcahuano Health Service (Act No. 15, 14 April 2026) and will be conducted in accordance with the Declaration of Helsinki, Council for International Organizations of Medical Sciences (CIOMS) guidelines and current Chilean regulations. Results will be disseminated through presentations at scientific meetings and publication in peer-reviewed journals.Trial registration numberNCT07643597. STRENGTHS AND LIMITATIONS OF THIS STUDY The prospective, multicentre design allows standardised data collection from before net ultrafiltration initiation and throughout the predefined 24-hour observation period. The prospective, multicentre design allows standardised data collection from before net ultrafiltration initiation and throughout the predefined 24-hour observation period. The study integrates complementary physiological domains, including macrohaemodynamics, focused echocardiography, venous congestion assessment (Venous Excess Ultrasound), dynamic functional tests, peripheral perfusion markers and biochemical biomarkers. The study integrates complementary physiological domains, including macrohaemodynamics, focused echocardiography, venous congestion assessment (Venous Excess Ultrasound), dynamic functional tests, peripheral perfusion markers and biochemical biomarkers. Standardised prospective assessments at predefined time points, together with event-triggered evaluations, will enable detailed characterisation of the temporal evolution of ultrafiltration intolerance. Standardised prospective assessments at predefined time points, together with event-triggered evaluations, will enable detailed characterisation of the temporal evolution of ultrafiltration intolerance. The multicentre pragmatic design enhances the generalisability of the findings while reflecting routine clinical practice in intensive care units. The multicentre pragmatic design enhances the generalisability of the findings while reflecting routine clinical practice in intensive care units. As an observational study, it is designed to identify associations and predictors of ultrafiltration intolerance rather than establish causal relationships. As an observational study, it is designed to identify associations and predictors of ultrafiltration intolerance rather than establish causal relationships. Introduction Fluid balance management in critically ill patients with acute kidney injury (AKI) is a fundamental therapeutic challenge. Fluid overload is independently associated with worse outcomes across multiple organ systems, including increased mortality, prolonged mechanical ventilation and impaired renal recovery.1 Net ultrafiltration (UFNET) through continuous renal replacement therapy (CRRT) has emerged as a cornerstone de-resuscitation strategy for patients who cannot achieve negative fluid balance through diuresis alone. Debate on UFNET has traditionally centred on the rate of fluid removal, since early studies suggested that rates exceeding 1.7 mL/kg/hour were associated with haemodynamic instability and worse outcomes.2 More recent evidence indicates that failure to achieve prescribed fluid removal targets, whether due to intradialytic hypotension or repeated therapy interruptions, is itself consistently associated with persistent positive fluid balance, organ congestion and increased mortality.3–7 This suggests that the primary risk may lie not in rapid fluid removal per se, but in poor tolerance that prevents effective de-resuscitation. Haemodynamic instability related to renal replacement therapy is the main clinical barrier to effective ultrafiltration and an independent predictor of mortality,8 with a reported incidence of 40%–60% during prolonged intermittent therapy and 19%–43% during CRRT.9–15 Conventional monitoring, based primarily on mean arterial pressure (MAP), is a relatively insensitive marker of early circulatory deterioration: significant reductions in cardiac output may occur before overt hypotension, as blood pressure is maintained through compensatory vasoconstriction.16–18 Hypotension therefore represents a late event, occurring only once vasoconstrictive reserve is exhausted. ‘Ultrafiltration intolerance’ currently lacks a standardised definition. By analogy with fluid tolerance, it can be conceptualised as limited physiological reserve and/or an ultrafiltration prescription inadequately adapted to the patient’s haemodynamic and perfusion status.17 18 Proposed assessment tools span ultrasound-based techniques (Venous Excess Ultrasound (VExUS), focused echocardiography), dynamic functional tests (passive leg raising (PLR), postural manoeuvres), peripheral perfusion parameters, circulating biomarkers and dynamic haemodynamic indices17–38; however, no prospective study has integrated these domains into a single, comprehensive evaluation. The VINKO study was designed to address this gap. Using a prospective, analytical case–control design, it aims to characterise the incidence, timing and haemodynamic phenotypes of ultrafiltration intolerance in critically ill patients with AKI undergoing CRRT and to identify independent predictors using a multiparametric monitoring strategy that integrates all relevant circulatory domains. Objectives The primary objective of VINKO is to identify factors associated with the development of ultrafiltration intolerance during the 24-hour observation period in critically ill patients with AKI undergoing CRRT with prescribed UFNET. Clinical, haemodynamic, ultrasound-based, perfusion-related and biochemical variables will be evaluated as candidate predictors of this outcome. Secondary objectives include (1) describing the incidence and timing of ultrafiltration intolerance during the 24-hour observation window; (2) comparing macrohaemodynamic, echocardiographic, venous ultrasound, peripheral perfusion and biochemical parameters between cases and controls; (3) characterising the pathophysiological profiles of ultrafiltration intolerance according to the predominant circulatory domain (haemodynamic compromise, peripheral perfusion deterioration, increased vasoactive support or metabolic alteration); (4) evaluating the association between UFNET rate and the development of intolerance; (5) assessing the agreement between intolerance defined by hypotension criteria alone versus intolerance defined by a composite perfusion-based definition; (6) describing clinical outcomes including cumulative fluid balance, achieved ultrafiltration, therapy interruptions, renal recovery, ventilator-free days and intensive care unit (ICU) and hospital mortality; and (7) characterising timing of onset (early vs late; eg, <6 hours vs ≥6 hours). Exploratory objectives include characterisation of intolerance severity according to a prespecified three-level grading system and assessment of the dose–response relationship between cumulative haemodynamic perturbation and clinical outcomes. Study design and setting VINKO is a prospective, multicentre, observational study conducted in adult patients admitted to the ICU with AKI who are receiving CRRT with prescribed UFNET. Importantly, all participants are enrolled before the occurrence of the primary outcome, and case or control status is not determined at enrolment. The observation period begins at T0, immediately before the initiation of prescribed UFNET, and continues prospectively for 24 hours, with standardised assessments at T0, T2, T4, T6, T12, T18 and T24, as well as mandatory off-schedule recordings triggered by any prespecified intolerance criterion. Patients are subsequently classified according to events occurring during this prospective observation period. Cases are patients who develop at least one prespecified criterion of ultrafiltration intolerance during the 24-hour observation window, whereas controls are patients who complete the observation period without meeting any ultrafiltration intolerance criterion. Thus, classification as a case or control occurs only after prospective observation for the primary outcome has begun. Recruitment will continue until the predefined sample size of 64 cases and 64 controls is reached. The study does not interfere with therapeutic prescriptions or modify routine clinical decision-making. Participating centres are selected to represent a range of high-complexity ICU settings with established experience in CRRT and multiparametric haemodynamic monitoring. The study is coordinated by the Nephrology Service, Hospital Las Higueras, Talcahuano, Chile, in collaboration with an international multidisciplinary team encompassing critical care nephrology, intensive care medicine and clinical research expertise from Chile, Perú, Mexico, Ecuador, Germany, Colombia, Italy and Brazil. Study population Eligible participants are adult patients (>18 years) admitted to the ICU with a diagnosis of AKI according to Kidney Diseases: Improving Global Outcomes (KDIGO) criteria, with an indication for CRRT and for whom the treating team has made the clinical decision to initiate UFNET as part of routine care. CRRT modality (CVVH: Continuous Venovenous Hemofiltration; CVVHD: Continuous Venovenous Hemodialysis; CVVHDF: Continuous Venovenous Hemodiafiltration), anticoagulation strategy and dialysis prescription will follow standard institutional protocols at each centre and will not be modified by the study. Both mechanically ventilated and non-mechanically ventilated patients are eligible for inclusion. Mechanical ventilation status will be recorded at baseline. Exclusion criteria include patients receiving chronic renal replacement therapy, pregnancy, active limitation of therapeutic effort or goals-of-care decisions, inability to perform haemodynamic or perfusion assessment, patients supported by extracorporeal membrane oxygenation and refusal to participate. Consecutive eligible patients will be screened during routine ICU care to minimise selection bias (table 1). Table 1Inclusion and exclusion criteriaInclusion criteriaExclusion criteriaAdult patients (>18 years)Patients on chronic renal replacement therapyAdmitted to the intensive care unitPregnancyDiagnosis of acute kidney injury according to KDIGO criteriaLimitation of therapeutic effort or goals-of-care decisions at admission or during the observation periodIndication for CRRT with a prescription for net ultrafiltrationInability to perform haemodynamic or perfusion assessmentClinical decision by the treating team to initiate net ultrafiltration as part of routine carePatients on extracorporeal membrane oxygenationWritten informed consent from the participant or legally authorised representativeRefusal to participate in the studyEligibility criteria for enrolment in the VINKO prospective observational case–control study.AKI, acute kidney injury; CRRT, continuous renal replacement therapy; KDIGO, Kidney Disease: Improving Global Outcomes; VINKO, Volume removal Intolerance during Net ultrafiltration in acute Kidney injury patients, an Observational study. Inclusion and exclusion criteria Eligibility criteria for enrolment in the VINKO prospective observational case–control study. AKI, acute kidney injury; CRRT, continuous renal replacement therapy; KDIGO, Kidney Disease: Improving Global Outcomes; VINKO, Volume removal Intolerance during Net ultrafiltration in acute Kidney injury patients, an Observational study. Continuous renal replacement therapy and net ultrafiltration CRRT will be performed according to current institutional protocols. The study does not impose any protocol-driven changes to standard therapeutic management. UFNET is defined as the volume of fluid removed by the CRRT machine through ultrafiltration, calculated as the difference between total effluent and the volume of fluids administered by the machine itself (dialysate and/or replacement solutions).36 UFNET does not include external fluid inputs (nutrition, medications, boluses, blood products), residual urine output or other non-CRRT fluid losses (drains, gastrointestinal losses). The prescribed and achieved UFNET rate (mL/kg/hour) will be recorded at each time point throughout the observation window. External fluid inputs, urine output and other non-CRRT fluid losses will be recorded separately and incorporated into the calculation of cumulative overall fluid balance. Overall fluid balance will therefore be analysed separately from machine-derived UFNET, and its association with clinical outcomes will be explored. Haemodynamic and perfusion assessment All enrolled patients will undergo a comprehensive, multiparametric assessment before the initiation of UFNET (T0) and at predefined intervals throughout the observation window. The assessment strategy integrates six complementary domains (table 2): Table 2Schedule of assessmentsAssessmentT0 (baseline)T2 (2 hours)T4 (4 hours)T6–T24 (extended)Eligibility screening (inclusion and exclusion criteria)✓–––Written informed consent✓–––Demographic and clinical data (age, sex, comorbidities)✓–––SOFA score and severity assessment✓–––CRRT modality and prescription parameters✓✓✓✓Prescribed and achieved UFNET rate (mL/kg/hour)✓✓✓✓Cumulative fluid balance✓––✓*Macrohaemodynamic variables (MAP, SAP, HR, PP, DSI)✓✓✓✓*Vasopressor/inotrope doses and VIS score✓✓✓✓*Dynamic haemodynamic variables (SVV, PPV)✓✓✓✓*Peripheral perfusion assessment (CRT, mottling, PPI)✓✓✓✓*Focused echocardiography (VTI, cardiac index, LVEF)✓✓✓✓*VExUS protocol (venous congestion assessment)✓✓✓✓Lung ultrasound✓–––Passive leg raising test✓–––Sitting/postural manoeuvre✓–––Serum lactate✓✓✓✓Albumin✓–––Ionised calcium✓––✓BNP/NT-proBNP (when available)✓–––Safety assessment and adverse event recording–✓✓✓Ultrafiltration intolerance event (off-schedule recording)–***Clinical outcomes (ICU/hospital mortality, renal recovery, MV-free days)––––Timing of haemodynamic, perfusion, ultrasound and biochemical assessments performed at T0 (baseline, prior to UFNET initiation) and during the observation window in the VINKO prospective observational study. The T6–T24 column reflects assessments performed at T6, T12, T18 and T24 (extended phase, every 6 hours).*Off-schedule abbreviated recording triggered by any ultrafiltration intolerance criterion. BNP, B-type natriuretic peptide; CRRT, continuous renal replacement therapy; CRT, capillary refill time; DSI, diastolic shock index; HR, heart rate; ICU, intensive care unit; LVEF, left ventricular ejection fraction; MAP, mean arterial pressure; MV, Mechanical Ventilation; NT-proBNP, N-terminal pro-B-type natriuretic peptide; PP, pulse pressure; PPI, peripheral perfusion index; PPV, pulse pressure variation; SAP, systolic arterial pressure; SOFA, Sequential Organ Failure Assessment; SVV, stroke volume variation; UFNET, net ultrafiltration; VExUS, Venous Excess Ultrasound; VINKO, Volume removal Intolerance during Net ultrafiltration in acute Kidney injury patients, an Observational study; VIS, Vasoactive-Inotropic Score; VTI, velocity-time integral. Schedule of assessments Timing of haemodynamic, perfusion, ultrasound and biochemical assessments performed at T0 (baseline, prior to UFNET initiation) and during the observation window in the VINKO prospective observational study. The T6–T24 column reflects assessments performed at T6, T12, T18 and T24 (extended phase, every 6 hours). Off-schedule abbreviated recording triggered by any ultrafiltration intolerance criterion. BNP, B-type natriuretic peptide; CRRT, continuous renal replacement therapy; CRT, capillary refill time; DSI, diastolic shock index; HR, heart rate; ICU, intensive care unit; LVEF, left ventricular ejection fraction; MAP, mean arterial pressure; MV, Mechanical Ventilation; NT-proBNP, N-terminal pro-B-type natriuretic peptide; PP, pulse pressure; PPI, peripheral perfusion index; PPV, pulse pressure variation; SAP, systolic arterial pressure; SOFA, Sequential Organ Failure Assessment; SVV, stroke volume variation; UFNET, net ultrafiltration; VExUS, Venous Excess Ultrasound; VINKO, Volume removal Intolerance during Net ultrafiltration in acute Kidney injury patients, an Observational study; VIS, Vasoactive-Inotropic Score; VTI, velocity-time integral. Macrohaemodynamic variables: MAP, systolic and diastolic arterial pressure, heart rate, pulse pressure, diastolic shock index, vasopressor and inotrope doses and Vasoactive-Inotropic Score. Dynamic haemodynamic indices: stroke volume variation and pulse pressure variation, when available under controlled mechanical ventilation; dynamic arterial elastance, when feasible. Ultrasound-based assessment: focused echocardiography for evaluation of left ventricular systolic and diastolic function, velocity-time integral, cardiac index and preload status; inferior vena cava assessment; and the VExUS protocol for venous congestion grading, including hepatic, portal and intrarenal vein Doppler patterns. Lung ultrasound will be performed at T0 and T24. Peripheral perfusion markers: capillary refill time (CRT), mottling score (0–5) and peripheral perfusion index, when available. Circulating biomarkers: serum lactate, albumin, ionised calcium and BNP or NT-proBNP (when available). Dynamic functional tests (performed at T0 only, prior to UFNET initiation): PLR, when feasible, to assess preload dependence; and a standardised sitting or postural manoeuvre, when feasible, as a functional test simulating an acute reduction in central blood volume. Standardised procedures for these manoeuvres are detailed in the online supplemental table S1. Definition of ultrafiltration intolerance Ultrafiltration intolerance is defined as a composite outcome because clinically relevant intolerance to fluid removal may manifest across different but complementary domains of circulatory dysfunction and may occur before overt hypotension. The components were prespecified to capture arterial pressure deterioration, escalation of pharmacological cardiovascular support, worsening peripheral perfusion and biochemical evidence of tissue hypoperfusion. These domains were selected to provide a clinically meaningful and multiparametric assessment of the physiological response to fluid removal rather than relying exclusively on arterial hypotension. Ultrafiltration intolerance will be defined as the occurrence of one or more of the following criteria during the observation period: Hypotension associated with ultrafiltration, defined as systolic blood pressure below 90 mm Hg or a decrease of ≥10 mm Hg from baseline in patients already hypotensive at T0. Hypotension associated with ultrafiltration, defined as systolic blood pressure below 90 mm Hg or a decrease of ≥10 mm Hg from baseline in patients already hypotensive at T0. Escalation of vasoactive support, defined as an increase of ≥25% in vasopressor or inotrope dose compared with the dose immediately preceding the event. Escalation of vasoactive support, defined as an increase of ≥25% in vasopressor or inotrope dose compared with the dose immediately preceding the event. Worsening of peripheral perfusion, defined as CRT exceeding 3 s or new onset or increase in mottling score. Worsening of peripheral perfusion, defined as CRT exceeding 3 s or new onset or increase in mottling score. Evidence of tissue hypoperfusion, defined as an increase in serum lactate ≥0.5 mmol/L from T0 with an absolute lactate exceeding 2 mmol/L, without an alternative explanation. Evidence of tissue hypoperfusion, defined as an increase in serum lactate ≥0.5 mmol/L from T0 with an absolute lactate exceeding 2 mmol/L, without an alternative explanation. In all cases, alternative causes of haemodynamic deterioration unrelated to ultrafiltration (eg, bleeding, changes in sedation) will be prospectively assessed and recorded. Attribution of alternative causes will be determined by the treating investigator in the case report form. Reduction, temporary interruption or discontinuation of UFNET due to haemodynamic instability or hypoperfusion will be recorded separately as a clinically relevant consequence of ultrafiltration intolerance and will contribute to the assessment of intolerance severity, but will not independently define the primary outcome. Among patients meeting the primary outcome definition, intolerance severity will be classified prospectively according to three levels: mild (signs of hypoperfusion without need for major intervention), moderate (need for UFNET reduction, fluid bolus or vasopressor adjustment without complete interruption) and severe (suspension or interruption of UFNET due to haemodynamic instability or hypoperfusion). Outcome measures Primary outcome: the primary outcome is the development of ultrafiltration intolerance during the 24-hour observation period, treated as a binary outcome (yes/no), according to the prespecified operational definition described above. Secondary outcomes include (1) the incidence of ultrafiltration intolerance during the observation window; (2) between-group comparisons of all assessed haemodynamic, perfusion and biochemical parameters; (3) characterisation of pathophysiological phenotypes of intolerance according to the predominant circulatory domain; (4) association between UFNET rate and the development of intolerance; (5) agreement between hypotension-based and perfusion-based intolerance definitions; (6) timing of onset of intolerance (early, ≤6 hours, vs late, >6 hours); and (7) describing clinical outcomes including 24-hour cumulative overall fluid balance, achieved ultrafiltration, therapy interruptions, renal recovery, ventilator-free days and ICU and hospital mortality, and exploring the association between cumulative overall fluid balance and these clinical outcomes. Statistical analysis Statistical analysis will follow Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) recommendations for observational studies. Continuous variables will be reported as median with IQR; categorical variables will be expressed as frequency and percentage. Baseline characteristics, haemodynamic parameters and perfusion variables will be summarised by group (cases vs controls). Normality of continuous variables will be assessed using the Shapiro–Wilk test. Given the anticipated non-normal distribution of haemodynamic variables in critically ill patients, inferential analyses will employ nonparametric methods. Continuous variables will be compared using the Mann–Whitney U test; categorical variables using the χ2 test or Fisher’s exact test, as appropriate. Patients will also be classified by timing of intolerance onset (early, ≤6 hours, vs late, >6 hours) for sensitivity analyses. Analyses will be performed using IBM SPSS Statistics and Python, employing standard libraries including pandas and numpy for data handling, scipy for non-parametric statistics, statsmodels for regression modelling and matplotlib for visualisation. Identification of predictors Associations between baseline demographic, clinical, haemodynamic, ultrasound, perfusion and biochemical variables, as well as their early changes during the first 6 hours of ultrafiltration, and the development of ultrafiltration intolerance will first be assessed through bivariate analysis. Variables identified as clinically relevant or statistically associated in bivariate analyses will subsequently be included in a multivariable logistic regression model to identify independent predictors of intolerance. Covariate selection will be guided by clinical relevance and the number of events to avoid overfitting. Results will be expressed as adjusted ORs with 95% CIs. A two-sided p value below 0.05 will be considered statistically significant. No formal adjustment for multiplicity will be applied to secondary and exploratory analyses, given their hypothesis-generating nature; findings will be interpreted with appropriate caution. For selected variables, diagnostic accuracy to identify UF intolerance will be assessed using receiver operating characteristic curves and summarised as the area under the receiver operating characteristic curve. (AUROC) with 95% CIs. Diagnostic performance will be described including sensitivity, specificity, positive predictive value and negative predictive value at specific cut-offs for each variable derived by the Youden Index. Data quality and missing data Data consistency, physiological plausibility and completeness will be verified for all variables. Variables with more than 25% missing data will be excluded from inferential analyses; those with 25% or fewer missing observations will be analysed using available-case analysis. Sample size Sample size was calculated using an analytical case–control approach, with preload dependence, defined by a positive PLR test prior to UFNET initiation, as the primary predefined exposure of interest. Following standard methodology for analytical studies, the proportion of exposure in controls (P2) was assumed to be 25%, with a minimum clinically relevant OR of 3.0, a two-sided significance level of 5% and 80% statistical power, at a 1:1 case–control ratio. Under these assumptions, the estimated sample size is 58 patients per group (116 total). Accounting for an approximately 10% over-recruitment allowance for potential losses, incomplete data or inability to definitively adjudicate outcomes, the target enrolment is 128 patients. Informed consent Written informed consent will be obtained from all participants or, when applicable, from their legally authorised representatives prior to enrolment. Oral consent and waiver of informed consent will not be used in this study. Ethics and dissemination The study was reviewed and approved by the Scientific Ethics Committee of the Talcahuano Health Service (Act No. 15, 14 April 2026). The study is being conducted in accordance with the ethical principles outlined in the Declaration of Helsinki, the CIOMS International Ethical Guidelines for Health-Related Research Involving Humans and current Chilean regulations, including Ministry of Health Technical Standard No. 0151/2013 governing research involving human participants. Given the observational nature of the study, all clinical management decisions are made according to routine clinical practice and are independent of the research protocol. Patient and public involvement Patients and members of the public were not involved in the design, conduct, reporting or dissemination plans of this research. Funding and budget The present study does not have external funding. All planned activities, including haemodynamic assessment, ultrasound monitoring and clinical data collection, are performed as part of the routine care of critically ill patients in the ICU, using equipment, supplies and human resources already available at the participating institutions. The execution of the study, therefore, does not generate additional costs for the institutions or for enrolled patients. Discussion The VINKO study addresses an important evidence gap in critical care nephrology by providing the first prospective, multiparametric, case–control evaluation of ultrafiltration intolerance in critically ill patients with AKI receiving CRRT. The study is anchored in a growing body of evidence highlighting that failure to achieve effective de-resuscitation, rather than rapid fluid removal per se, is a primary determinant of adverse outcomes in this population.3–5 By extending the definition of ultrafiltration intolerance beyond hypotension to encompass peripheral perfusion deterioration and biochemical evidence of tissue hypoperfusion, VINKO operationalises a pathophysiologically coherent framework for characterising the individual circulatory response to fluid removal. A key strength of this protocol is the integration of complementary haemodynamic domains into a single, standardised assessment strategy. Dynamic functional tests, particularly PLR, which has demonstrated robust predictive performance for fluid responsiveness and tolerance across multiple settings,19 25 are combined with echocardiographic evaluation of cardiac function and preload, venous congestion assessment via the VExUS protocol and real-time peripheral perfusion monitoring.21 24 This multiparametric approach is grounded in the well-established observation that blood pressure is an insensitive marker of early circulatory compromise, and that reductions in cardiac output and microcirculatory flow may precede overt hypotension.17 Several limitations should be acknowledged when interpreting future findings. As an observational study, VINKO is not designed to support causal inference; identified associations between candidate predictors and ultrafiltration intolerance will be hypothesis-generating. Variability in CRRT prescription, haemodynamic monitoring capabilities and clinical management across participating centres reflects real-world heterogeneity and may introduce clinical noise that attenuates detectable associations. Some advanced monitoring modalities, particularly continuous cardiac output estimation and biomarker assays, may not be universally available across sites, potentially limiting measurement completeness for certain variables; this is explicitly accounted for in the analytic plan through available-case analysis. The feasibility-based sample size provides adequate power for the primary case–control comparison but limits sensitivity for detecting weak or heterogeneous effects in subgroup analyses. Notwithstanding these limitations, the VINKO study provides a rigorous methodological foundation for characterising the pathophysiology of ultrafiltration intolerance in critically ill patients with AKI. Systematic data collection will enable exploratory phenotyping analyses, inform the selection of candidate predictors for future interventional platforms and help define appropriate monitoring standards for safe and effective CRRT-based de-resuscitation. More broadly, this protocol may serve as a methodological template for prospective, multiparametric studies evaluating haemodynamic tolerance to therapeutic interventions in the critical care setting. STROBE compliance statement This protocol was developed with reference to the STROBE statement for observational studies in epidemiology. Although STROBE was conceived as a reporting guideline for completed studies, its principles were applied where relevant to this prospective protocol, particularly regarding the transparent description of study design, population, exposures, outcomes, potential sources of bias and analytical strategy. A summary mapping of STROBE items to the VINKO protocol is provided in online supplemental table S2. Supplementary material