Predicting mortality risk in dialysis: Assessment of risk factors using traditional and advanced modeling techniques within the Monitoring Dialysis Outcomes initiative.

INTRODUCTION: Several factors affect the survival of End Stage Kidney Disease (ESKD) patients on dialysis. Machine learning (ML) models may help tackle multivariable and complex, often non-linear predictors of adverse clinical events in ESKD patients. In this study, we used advanced ML method as well as a traditional statistical method to develop and compare the risk factors for mortality prediction model in hemodialysis (HD) patients. MATERIALS AND METHODS: We included data HD patients who had data across a baseline period of at least 1 year and 1 day in the internationally representative Monitoring Dialysis Outcomes (MONDO) Initiative dataset. Twenty-three input parameters considered in the model were chosen in an a priori manner. The prediction model used 1 year baseline data to predict death in the following 3 years. The dataset was randomly split into 80% training data and 20% testing data for model development. Two different modeling techniques were used to build the mortality prediction model. FINDINGS: A total of 95,142 patients were included in the analysis sample. The area under the receiver operating curve (AUROC) of the model on the test data with XGBoost ML model was 0.84 on the training data and 0.80 on the test data. AUROC of the logistic regression model was 0.73 on training data and 0.75 on test data. Four out of the top five predictors were common to both modeling strategies. DISCUSSION: In the internationally representative MONDO data for HD patients, we describe the development of a ML model and a traditional statistical model that was suitable for classification of a prevalent HD patient's 3-year risk of death. While both models had a reasonably high AUROC, the ML model was able to identify levels of hematocrit (HCT) as an important risk factor in mortality. If implemented in clinical practice, such proof-of-concept models could be used to provide pre-emptive care for HD patients.

Estimation of Muscle Mass in the Integrated Assessment of Patients on Hemodialysis.

Assessment of muscle mass (MM) or its proxies, lean tissue mass (LTM) or fat-free mass (FFM), is an integral part of the diagnosis of protein-energy wasting (PEW) and sarcopenia in patients on hemodialysis (HD). Both sarcopenia and PEW are related to a loss of functionality and also increased morbidity and mortality in this patient population. However, loss of MM is a part of a wider spectrum, including inflammation and fluid overload. As both sarcopenia and PEW are amendable to treatment, estimation of MM regularly is therefore of major clinical relevance. Whereas, computer-assisted tomography (CT) or dual-energy X-ray absorptiometry (DXA) is considered a reference method, it is unsuitable as a method for routine clinical monitoring. In this review, different bedside methods to estimate MM or its proxies in patients on HD will be discussed, with emphasis on biochemical methods, simplified creatinine index (SCI), bioimpedance spectroscopy (BIS), and muscle ultrasound (US). Body composition parameters of all methods are related to the outcome and appear relevant in clinical practice. The US is the only parameter by which muscle dimensions are measured. BIS and SCI are also dependent on either theoretical assumptions or the use of population-specific regression equations. Potential caveats of the methods are that SCI can be influenced by residual renal function, BIS can be influenced by fluid overload, although the latter may be circumvented by the use of a three-compartment model, and that muscle US reflects regional and not whole body MM. In conclusion, both SCI and BIS as well as muscle US are all valuable methods that can be applied for bedside nutritional assessment in patients on HD and appear suitable for routine follow-up. The choice for either method depends on local preferences. However, estimation of MM or its proxies should always be part of a multidimensional assessment of the patient followed by a personalized treatment strategy.

Three compartment bioimpedance spectroscopy in the nutritional assessment and the outcome of patients with advanced or end stage kidney disease: What have we learned so far?

Bioimpedance spectroscopy (BIS) is an easily applicable tool to assess body composition. The three compartment model BIS (3C BIS) conventionally expresses body composition as lean tissue index (LTI) (lean tissue mass [LTM]/height in meters squared) and fat tissue index (FTI) (adipose tissue mass/height in meters squared), and a virtual compartment reflecting fluid overload (FO). It has been studied extensively in relation to diagnosis and treatment guidance of fluid status disorders in patients with advanced-stage or end-stage renal disease. It is the aim of this article to provide a narrative review on the relevance of 3C BIS in the nutritional assessment in this population. At a population level, LTI decreases after the start of hemodialysis, whereas FTI increases. LTI below the 10th percentile is a consistent predictor of outcome whereas a low FTI is predominantly associated with outcome when combined with a low LTI. Recent research also showed the connection between low LTI, inflammation, and FO, which are cumulatively associated with an increased mortality risk. However, studies toward nutritional interventions based on BIS data are still lacking in this population. In conclusion, 3C BIS, by disentangling the components of body mass index, has contributed to our understanding of the relevance of abnormalities in different body compartments in chronic kidney disease patients, and appears to be a valuable prognostic tool, at least at a population level. Studies assessing the effect of BIS guided nutritional intervention could further support its use in the daily clinical care for renal patients.

Fluid status assessment in hemodialysis patients and the association with outcome: review of recent literature.

PURPOSE OF REVIEW: In this review, we will discuss the most recent literature regarding fluids status assessment in hemodialysis patients, and the associations with outcome. RECENT FINDINGS: Research toward technique-assisted assessment of fluid status in hemodialysis patients has been going on for many years. However, there is no absolute agreement between techniques, such as bioimpedance, lung ultrasound, biochemical markers or vena caval diameter, likely because they reflect different fluid compartments with potentially altered distribution in hemodialysis patients. Recent studies, mostly based on bioimpedance, have shown not only an association of severe, but also of moderate predialysis fluid overload with overall survival. Also predialysis fluid depletion has been found to associate with and increased mortality risk. Interventional studies with fluid-guided management are scarce and outline the difficulties of achieving dry weight is the dialysis population. SUMMARY: Optimal estimation of predialysis fluid status remains challenging and may require a combination of clinical and technical derived parameters. There appears to be a narrow window of optimal predialysis fluid status. Further clinical studies are necessary to identify strategies to improve survival in hemodialysis patients with abnormalities in fluid status.

ISPD Cardiovascular and Metabolic Guidelines in Adult Peritoneal Dialysis Patients Part I - Assessment and Management of Various Cardiovascular Risk Factors.

Cardiovascular disease contributes significantly to the adverse clinical outcomes of peritoneal dialysis (PD) patients. Numerous cardiovascular risk factors play important roles in the development of various cardiovascular complications. Of these, loss of residual renal function is regarded as one of the key cardiovascular risk factors and is associated with an increased mortality and cardiovascular death. It is also recognized that PD solutions may incur significant adverse metabolic effects in PD patients. The International Society for Peritoneal Dialysis (ISPD) commissioned a global workgroup in 2012 to formulate a series of recommendations regarding lifestyle modification, assessment and management of various cardiovascular risk factors, as well as management of the various cardiovascular complications including coronary artery disease, heart failure, arrhythmia (specifically atrial fibrillation), cerebrovascular disease, peripheral arterial disease and sudden cardiac death, to be published in 2 guideline documents. This publication forms the first part of the guideline documents and includes recommendations on assessment and management of various cardiovascular risk factors. The documents are intended to serve as a global clinical practice guideline for clinicians who look after PD patients. The ISPD workgroup also identifies areas where evidence is lacking and further research is needed.

Body composition in dialysis patients: a functional assessment of bioimpedance using different prediction models.

OBJECTIVES: The assessment of body composition (BC) in dialysis patients is of clinical importance given its role in the diagnosis of malnutrition and sarcopenia. Bioimpedance techniques routinely express BC as a 2-compartment (2-C) model distinguishing fat mass (FM) and fat-free mass (FFM), which may be influenced by the hydration of adipose tissue and fluid overload (OH). Recently, the BC monitor was introduced which applies a 3-compartment (3-C) model, distinguishing OH, adipose tissue mass, and lean tissue mass. The aim of this study was to compare BC between the 2-C and 3-C models and assess their relation with markers of functional performance (handgrip strength [HGS] and 4-m walking test), as well as with biochemical markers of nutrition. METHODS: Forty-seven dialysis patients (30 males and 17 females) (35 hemodialysis, 12 peritoneal dialysis) with a mean age of 64.8 ± 16.5 years were studied. 3-C BC was assessed by BC monitor, whereas the obtained resistivity values were used to calculate FM and FFM according to the Xitron Hydra 4200 formulas, which are based on a 2-C model. RESULTS: FFM (3-C) was 0.99 kg (95% confidence interval [CI], 0.27 to 1.71, P = .008) higher than FFM (2-C). FM (3-C) was 2.43 kg (95% CI, 1.70-3.15, P < .001) lower than FM (2-C). OH was 1.4 ± 1.8 L. OH correlated significantly with ΔFFM (FFM 3-C - FFM 2-C) (r = 0.361; P < .05) and ΔFM (FM 3-C - FM 2-C) (r = 0.387; P = .009). HGS correlated significantly with FFM (2-C) (r = 0.713; P < .001), FFM (3-C) (r = 0.711; P < .001), body cell mass (2-C) (r = 0.733; P < .001), and body cell mass (3-C) (r = 0.767; P < .001). Both physical activity (r = 0.456; P = .004) and HGS (r = 0.488; P = .002), but not BC, were significantly related to walking speed. CONCLUSIONS: Significant differences between 2-C and 3-C models were observed, which are partly explained by the presence of OH. OH, which was related to ΔFFM and ΔFM of the 2-C and 3-C models, is therefore an important parameter for the differences in estimation of BC parameters of the 2-C and 3-C models. Both FFM (3-C) and FFM (2-C) were significantly related to HGS. Bioimpedance, HGS, and the 4-m walking test may all be valuable tools in the multidimensional nutritional assessment of both hemodialysis and peritoneal dialysis patients.

Magnetic resonance angiographic assessment of upper extremity vessels prior to vascular access surgery: feasibility and accuracy.

A contrast-enhanced magnetic resonance angiography (CE-MRA) protocol for selective imaging of the entire upper extremity arterial and venous tree in a single exam has been developed. Twenty-five end-stage renal disease (ESRD) patients underwent CE-MRA and duplex ultrasonography (DUS) of the upper extremity prior to hemodialysis vascular access creation. Accuracy of CE-MRA arterial and venous diameter measurements were compared with DUS and intraoperative (IO) diameter measurements, the standard of reference. Upper extremity vasculature depiction was feasible with CE-MRA. CE-MRA forearm and upper arm arterial diameters were 2.94 +/- 0.67 mm and 4.05 +/- 0.84 mm, respectively. DUS arterial diameters were 2.80 +/- 0.48 mm and 4.38 +/- 1.24 mm; IO diameters were 3.00 +/- 0.35 mm and 3.55 +/- 0.51 mm. Forearm arterial diameters were accurately determined with both techniques. Both techniques overestimated upper arm arterial diameters significantly. Venous diameters were accurately determined with CE-MRA but not with DUS (forearm: CE-MRA: 2.64 +/- 0.61 mm; DUS: 2.50 +/- 0.44 mm, and IO: 3.40 +/- 0.22 mm; upper arm: CE-MRA: 4.09 +/- 0.71 mm; DUS: 3.02 +/- 1.65 mm, and IO: 4.30 +/- 0.78 mm). CE-MRA enables selective imaging of upper extremity vasculature in patients requiring hemodialysis access. Forearm arterial diameters can be assessed accurately by CE-MRA. Both CE-MRA and DUS slightly overestimate upper arm arterial diameters. In comparison to DUS, CE-MRA enables a more accurate determination of upper extremity venous diameters.

Ionic dialysance and the assessment of Kt/V: the influence of different estimates of V on method agreement.

BACKGROUND: Ionic dialysance was recently introduced as a means to assess Kt/V (K(ID)t/V). With this method, urea distribution volume (V) has to be estimated. The primary aim of the present study was to assess the agreement between equilibrated Kt/V assessed by urea kinetic modelling (eKt/V) with K(ID)t/V taking into account different estimates of V, and to assess the monthly variation in V. Secondly, the mechanisms behind the intra-treatment changes in ionic dialysance and inter-treatment variability of K(ID)t/V were assessed. METHODS: Sixty-six patients were included. eKt/V was estimated using 30 min post-treatment sampling in the second generation Daugirdas equation. V was assessed by the formulae of Watson and Chertow (V(Watson); V(Chertow)), double-pool urea kinetic modelling (V(UKM)) and by ionic dialysance (V(IOD)) [Diascan; Hospal(R)]. RESULTS: The use of V(UKM) or V(IOD) instead of V(Watson) or V(Chertow) improved the relation between eKt/V and K(ID)t/V (both r = 0.93; P < 0.001 vs r = 0.84 and r = 0.81; P < 0.001). Mean values of eKt/V (1.19 +/- 0.21), K(ID)t/V(UKM) (1.19 +/- 0.30) and K(ID)t/V(IOD) (1.21 +/- 0.25) were comparable. Intra-class correlation coefficient of V(IOD) was 0.87 with a 1-month interval and <0.75 after 2 and 3 months. Intra-class correlation coefficient of V(DP) was 0.79 with a 1-month interval and <0.75 after 2 and 3 months. Inter-treatment variation in K(ID)t/V during six consecutive dialysis sessions was 6.1% +/- 0.6%. Changes in blood flow were the main determinant of variations in K(ID)t/V (P < 0.05). During treatment, ionic dialysance decreased by 12 +/- 13 ml/min (P < 0.001). The decline in blood volume was the major determinant of the intra-dialytic change in ionic dialysance (P < 0.05). CONCLUSION: The use of V(IOD) and V(UKM) results in better agreement between eKt/V and K(ID)t/V compared with anthropometric formulae. K(ID)t/V was comparable with eKt/V and thus lower than expected for a single-pool method. V(IOD) and V(UKM), should be assessed at least monthly. K(ID)t/V varies widely between consecutive dialysis sessions, mainly due to differences in blood flow. During treatment, ionic dialysance decreases, which is related to the relative decline in blood volume.

An evaluation of blood volume changes during ultrafiltration pulses and natriuretic peptides in the assessment of dry weight in hemodialysis patients.

Changes in blood volume (BV) during dialysis as well as plasma levels of brain natriuretic peptide (BNP) and N-terminal (NT) pro-BNP levels are possible tools to assess dry weight in hemodialysis (HD) patients. The aim of the study was to compare these parameters with other non-invasive techniques used to assess dry weight in HD patients, and to study their relation with intradialytic hypotension (IDH) and the presence of cardiovascular disease BV changes during HD, both during regular dialysis and during an ultrafiltration pulse, plasma levels of NT pro-BNP and BNP, and vena cava diameter index (VCDI) were assessed in a cohort of 66 HD patients, which was subdivided according to tertiles of total body water (TBW) corrected for body weight, assessed by bioimpedance analysis. Parameters were also related to the presence of IDH and history of cardiovascular disease. The decline in BV during regular dialysis and during an ultrafiltration pulse, as well as VCDI and BNP were significantly different between the tertiles of normalized TBW, but refill after the ultrafiltration pulse and NT pro-BNP were not. Only VCDI and the decline in BV during regular dialysis were significantly different between patients with or without IDH. Vena cava diameter index, BNP, and NT pro-BNP were significantly higher in patients with cardiovascular disease. Using bioimpedance as the reference method, changes in BV, either during regular dialysis or during an ultrafiltration pulse, as well as VCDI and BNP are all indicative of hydration state in dialysis patients, but refill after an ultrafiltration pulse is not. Only VCDI and BV changes were related to IDH. The presence of cardiovascular disease appears to influence both VCDI as well as BNP.

Impact of a quality improvement programme based on vascular access flow monitoring on costs, access occlusion and access failure.

BACKGROUND: Vascular access thrombosis is a substantial source of morbidity in chronic haemodialysis patients. Periodical access flow measurements can predict the presence of vascular access stenosis and provide an opportunity for early intervention to prevent subsequent vascular access thrombosis. By this system of quality improvement, vascular access-related costs might be reduced. The aim of this study was to analyse the cost impact of a quality improvement programme based on periodic access flow measurements. METHODS: The number and costs of vascular access interventions (summary of angiography, percutaneous transluminal angioplasty, catheter placement, hospitalization days and costs for surgery) in the period 2001-2003 (quality improvement period; QIP, 218.6 patient-years observed) were retrospectively compared with a reference period (RP, 1996-1998, 214.4 patient-years observed) during which no access flow was measured. All access flow measurements were done on a regular base and interventions were performed according to the Kidney Disease Outcome Quality Initiative. RESULTS: Surgical thrombectomy procedures were significantly less during the QIP (0.25 +/- 0.57 events/patient-year) compared with RP (0.63 +/- 1.06 events/patient-year; P = 0.000), whereas access loss was not significantly different. During the QIP, 205 radiological interventions were performed (0.88 +/- 1.16 events/patient-year), and in the RP around 48 (0.33 +/- 0.65 events/patient-year; P = 0.000). Access-related costs tended to be lower during the QIP compared with the RP. The cost reduction appeared to be limited to patients with arteriovenous graft (AVG), in which access-related costs were significantly lower during the QIP (2360.95 euro +/- 2838.17 euro patient-year) compared with the RP (4003.96 euro +/- 3810.92 euro patient-year; P = 0.012), but not in patients with arteriovenous fistula (AVF). CONCLUSION: A quality improvement programme based on periodical access flow measurement reduced the number of acute vascular access failures due to thrombotic events and also significantly reduced health care costs in patients with AVG, but not in patients with AVF. The quality improvement programme had no effect on access survival.

Effect of ultrafiltration on thermal variables, skin temperature, skin blood flow, and energy expenditure during ultrapure hemodialysis.

The cause of the increase in core temperature (CT) during hemodialysis (HD) is still under debate. It has been suggested that peripheral vasoconstriction as a result of hypovolemia, leading to a reduced dissipation of heat from the skin, is the main cause of this increase in CT. If so, then it would be expected that extracorporeal heat flow (Jex) needed to maintain a stable CT (isothermic; T-control = 0, no change in CT) is largely different between body temperature control HD combined with ultrafiltration (UF) and body temperature control HD without UF (isovolemic). Consequently, significant differences in DeltaCT would be expected between isovolemic HD and HD combined with UF at zero Jex (thermoneutral; E-control = 0, no supply or removal of thermal energy to and from the extracorporeal circulation). During the latter treatment, the CT is expected to increase. In this study, changes in thermal variables (CT and Jex), skin blood flow, energy expenditure, and cytokines (TNF-alpha, IL-1 receptor antagonist, and IL-6) were compared in 13 patients, each undergoing body temperature control (T-control = 0) HD without and with UF and energy-neutral (E-control = 0) HD without and with UF. CT increased equally during energy-neutral treatments, with (0.32 +/- 0.16 degrees C; P = 0.000) and without (0.27 +/- 0.29 degrees C; P = 0.006) UF. In body temperature control treatments, the relationship between Jex and UF tended to be significant (r = -0.51; P = 0.07); however, there was no significant difference in cooling requirements regardless of whether treatments were done without (-17.9 +/- 9.3W) or with UF (-17.8 +/- 13.27W). Changes in energy expenditure did not differ among the four treatment modes. There were no significant differences in pre- and postdialysis levels of cytokines within or between treatments. Although fluid removal has an effect on thermal variables, no single mechanism seems to be responsible for the increased heat accumulation during HD.

Assessment of fluid status in peritoneal dialysis patients.

OBJECTIVES: To assess the influence of abnormalities in fluid status and body composition on agreement between multifrequency bioimpedance analysis (MF-BIA), segmental BIA (sigmaBIA), the Watson formula, and tracer dilution techniques. DESIGN: Cross-sectional. SETTING: Multicenter. PATIENTS: 40 patients (29 males, 11 females) on peritoneal dialysis (PD). MAIN OUTCOME MEASURES: Agreement between the various techniques used to assess total body water (TBW) [MF-BIA, deuterium oxide (D2O), and the Watson formula] and extracellular water (ECW) [MF-BIA, bromide dilution (NaBr), and sigmaBIA], also in relation to the relative magnitude of the body water compartments [ECW (NaBr):body weight (BW) and TBW (D2O):BW] and body composition (DEXA). Second, the relation between body water compartments with echocardiographic parameters. RESULTS: Wide limits of agreement were observed between tracer dilution techniques and MF-BIA [TBW (D2O - MF-BIA) 2.0 +/- 3.9 L; ECW (NaBr - MF-BIA) -2.8 +/- 3.9 L], which were related to the relative magnitude of the body water compartments: r = 0.70 for ECW and r = 0.40 for TBW. sigmaBIA did not improve the agreement [ECW (NaBr-sigmaBIA): 3.7 +/- 2.9 L]. Also, wide limits of agreement were observed between D2O and the Watson formula (-2.3 +/- 3.3 L). The difference between D2O and Watson was related to hydration state and to percentage of fat mass (r = 0.70 and r = -0.53, p < 0.05). Both ECW and TBW as assessed by BIA and tracer dilution were related to echocardiographic parameters. CONCLUSION: Wide limits of agreement were found between MF-BIA and sigmaBIA with dilution methods in PD patients, which were related to hydration state itself. The disagreement between the Watson formula and dilution methods was related to both hydration state and body composition.