Proposal of a new equation for estimating resting energy expenditure of acute kidney injury patients on dialysis: a machine learning approach

Nutrition support for patients with renal dysfunction in the intensive care unit: A narrative review

Providing optimal nutrition support in the intensive care unit (ICU) is a challenging and dynamic process. Energy, protein, fluid, electrolyte, and micronutrient requirements all can be altered in patients with acute, chronic, and acute-on-chronic kidney disease. Given that renal dysfunction occurs in up to one-half of ICU patients, it is imperative that nutrition support providers understand how renal dysfunction, its metabolic consequences, and its treatments, including renal replacement therapy (RRT), affect patients' nutrition needs. Data on nutrient requirements in critically ill patients with renal dysfunction are sparse. This article provides an overview of renal dysfunction in the ICU and identifies and addresses the unique nutrition challenges present among these patients, including those receiving RRT, as supported by the available literature and guidelines.

Ensemble of machine learning techniques to predict survival in kidney transplant recipients

Kidney transplant recipients face a high cardiovascular risk, which is a leading cause of death in this patient group. This article proposes the application of clustering techniques and feature selection to predict the survival outcomes of kidney transplant recipients based on machine learning techniques and mainstream statistical methods. First, feature selection techniques (Boruta, Random Survival Forest and Elastic Net) are used to detect the most relevant variables. Subsequently, each set of variables obtained by each feature selection technique is used as input for the clustering algorithms used (Consensus Clustering, Self-Organizing Map and Agglomerative Clustering) to determine which combination of feature selection, clustering algorithm and number of clusters maximizes intercluster variability. Next, the mechanism called False Clustering Discovery Reduction is applied to obtain the minimum number of statistically differentiable populations after applying a control metric. This metric is based on a variance test to confirm that reducing the number of clusters does not generate significant losses in the heterogeneity obtained. This approach was applied to the Organ Procurement and Transplantation Network medical dataset (n = 11,332). The combination of Random Survival Forest and consensus clustering yielded the optimal result of 4 clusters starting from 8 initial ones. Finally, for each population, Kaplan-Meier survival curves are generated to predict the survival of new patients based on the predictions of the XGBoost classifier, with an overall multi-class AUC of 98.11%.

A Proposed Predictive Equation for Energy Expenditure Estimation Among Noncritically Ill Patients With Acute Kidney Injury

OBJECTIVE

The incidence of acute kidney injury (AKI) is identified more frequently in noncritical compared with intensive care settings. The prognosis of malnourished AKI patients is far worse than those with normal nutritional status. However, a method for estimating the optimal amount of energy required to guide nutritional support among noncritically ill AKI patients is yet to be determined.

METHODS

We evaluated the performance of weight-based formulas (20-30 kcal/kg/day) with the reference values of energy expenditure (EE) measured by indirect calorimetry (IC) among noncritically ill AKI patients during hospitalization. The statistics for assessing agreement, including total deviation index and accuracy within 10% represent the percentage of estimations falling within the IC value range of ±10%, were tested. Parameters for predicting the EE equation were also developed using a regression analysis model.

RESULTS

A total of 40 noncritically ill AKI patients were recruited. The mean age of participants was 62.5 ± 16.5 years with 50% being male. The average IC-derived EE was 1,124.6 ± 278.9 kcal/day with respiratory quotients 0.8-1.3, indicating good validity of the IC test. Receiving dialysis, protein catabolic rate, and age was not significantly associated with measured EE. Nearly all weight-based formulas overestimated measured EE. The magnitude of total deviation index values was broad with the proportion of patients achieving an accuracy of 10% being as low as 20%. The proposed equation to predict EE derived from this study was EE (kcal/day) = 618.27 + (8.98 x weight in kg) + 137.0 if diabetes - 199.7 if female (r2 = 0.68, P < .001). In the validation study with an independent group of noncritically ill AKI patients, predicted EE using the newly derived equation was also significantly correlated with measured EE by IC (r = 0.69, P = .004).

CONCLUSION

Estimation of EE by weight-based formulas usually overestimated measured EE among noncritically ill AKI patients. In the absence of IC, the proposed predictive equation, specifically for noncritically ill AKI patients might be useful, in addition to weight-based formulas, for guiding caloric dosing in clinical practice.

Machine learning models using non-linear techniques improve the prediction of resting energy expenditure in individuals receiving hemodialysis

PURPOSE

Approximately 700,000 people in the USA have chronic kidney disease requiring dialysis. Protein-energy wasting (PEW), a condition of advanced catabolism, contributes to three-year survival rates of 50%. PEW occurs at all levels of Body Mass Index (BMI) but is devastating for those people at the extremes. Treatment for PEW depends on an accurate understanding of energy expenditure. Previous research established that current methods of identifying PEW and assessing adequate treatments are imprecise. This includes disease-specific equations for estimated resting energy expenditure (eREE). In this study, we applied machine learning (ML) modelling techniques to a clinical database of dialysis patients. We assessed the precision of the ML algorithms relative to the best-performing traditional equation, the MHDE.

METHODS

This was a secondary analysis of the Rutgers Nutrition and Kidney Database. To build the ML models we divided the population into test and validation sets. Eleven ML models were run and optimized, with the best three selected by the lowest root mean squared error (RMSE) from measured REE. Values for eREE were generated for each ML model and for the MHDE. We compared precision using Bland-Altman plots.

RESULTS

Individuals were 41.4% female and 82.0% African American. The mean age was 56.4 ± 11.1 years, and the median BMI was 28.8 (IQR = 24.8 - 34.0) kg/m2. The best ML models were SVR, Linear Regression and Elastic net with RMSE of 103.6 kcal, 119.0 kcal and 121.1 kcal respectively. The SVR demonstrated the greatest precision, with 91.2% of values falling within acceptable limits. This compared to 47.1% for the MHDE. The models using non-linear techniques were precise across extremes of BMI.

CONCLUSION

ML improves precision in calculating eREE for dialysis patients, including those most vulnerable for PEW. Further development for clinical use is a priority.

The Validity of the Energy Expenditure Criteria Based on Open Source Code through two Inertial Sensors

Through this study, we developed and validated a system for energy expenditure calculation, which only requires low-cost inertial sensors and open source R software. Five healthy subjects ran at ten different speeds while their kinematic variables were recorded on the thigh and wrist. Two ActiGraph wireless inertial sensors and a low-cost Bluetooth-based inertial sensor (Lis2DH12), assembled by SensorID, were used. Ten energy expenditure equations were automatically calculated in a developed open source R software (our own creation). A correlation analysis was used to compare the results of the energy expenditure equations. A high interclass correlation coefficient of estimated energy expenditure on the thigh and wrist was observed with an Actigraph and Sensor ID accelerometer; the corrected Freedson equation showed the highest values, and the Santos-Lozano vector magnitude equation and Sasaki equation demonstrated the lowest one. Energy expenditure was compared between the wrist and thigh and showed low correlation values. Despite the positive results obtained, it was necessary to design specific equations for the estimation of energy expenditure measured with inertial sensors on the thigh. The use of the same formula equation in two different placements did not report a positive interclass correlation coefficient.

Nutritional assessment and support during continuous renal replacement therapy

Malnutrition is highly prevalent in patients with acute kidney injury, especially in those receiving renal replacement therapy (RRT). For the assessment of nutritional status, a combination of screening tools, anthropometry, and laboratory parameters is recommended rather than a single test. To avoid underfeeding and overfeeding during RRT, energy expenditure should be measured by indirect calorimetry or calculated using predictive equations. Nitrogen balance should be periodically measured to assess the degree of catabolism and to evaluate protein intake. However, there is limited data for nutritional targets specifically for patients on RRT, such as protein intake. The composition of commercial solutions for continuous renal replacement therapy (CRRT) varies. CRRT itself can be associated with both, nutrient losses into the effluent fluid and caloric gain from dextrose, lactate, and citrate. The role of micronutrient supplementation, and potential use of micronutrient enriched CRRT solutions in this setting is unknown, too. This review provides an overview of existing knowledge and uncertainties related to nutritional aspects in patients on CRRT and emphasizes the need for more research in this area.