Executive Summary of the Korean Society of Nephrology 2021 Clinical Practice Guideline for Optimal Hemodialysis Treatment

Predicting early mortality in hemodialysis patients: a deep learning approach using a nationwide prospective cohort in South Korea

Early mortality after hemodialysis (HD) initiation significantly impacts the longevity of HD patients. This study aimed to quantify the effect sizes of risk factors on mortality using various machine learning approaches. A cohort of 3284 HD patients from the CRC-ESRD (2008-2014) was analyzed. Mortality risk models were validated using logistic regression, ridge regression, lasso regression, and decision trees, as well as ensemble methods like bagging and random forest. To better handle missing data and time-series variables, a recurrent neural network (RNN) with an autoencoder was also developed. Additionally, survival models predicting hazard ratios were employed using survival analysis techniques. The analysis included 1750 prevalent and 1534 incident HD patients (mean age 58.4 ± 13.6 years, 59.3% male). Over a median follow-up of 66.2 months, the overall mortality rate was 19.3%. Random forest models achieved an AUC of 0.8321 for first-year mortality prediction, which was further improved by the RNN with autoencoder (AUC 0.8357). The survival bagging model had the highest hazard ratio predictability (C-index 0.7756). A shorter dialysis duration (< 14.9 months) and high modified Charlson comorbidity index scores (7-9) were associated with hazard ratios up to 7.76 (C-index 0.7693). Comorbidities were more influential than age in predicting early mortality. Monitoring dialysis adequacy (KT/V), RAAS inhibitor use, and urine output is crucial for assessing early prognosis.

Monitoring and maintaining quality in the paediatric haemodialysis unit

Chronic kidney disease in children is being increasingly recognised and reported worldwide, and the focus of paediatric dialysis planning has changed from acute care alone to encompass chronic care. In many parts of the world, haemodialysis for children is performed in adult units and is based on standards established for adults. This review proposes standards for paediatric haemodialysis, incorporating special requirements for children while simultaneously drawing from the adult experience. We discuss the optimum requirements, including space utilisation, equipment needed, water treatment facilities, disposables, safety standards, staffing needs, monitoring and maintenance, infection prevention, waste disposal and quality indicators. We also review recent advancements in the field that should be incorporated into future dialysis units and the steps required for achieving carbon neutrality and protecting the environment.

Which blood pressure metrics should be used in patients on dialysis?

Remarkable progress has recently been achieved in blood pressure (BP) control based on key research findings in the general population. It has been observed that maintaining BP slightly lower than previously recommended goals leads to better clinical outcomes, provided that patients can tolerate it. Previously, BP control targets for dialysis patients were extrapolated from studies conducted on the general population. However, dialysis patients are considered a distinct group with unique characteristics, which makes defining appropriate BP targets a matter of debate. Several observational studies measuring BP in hemodialysis (HD) patients within dialysis units have shown that lower peridialysis BP (pre-, post-, and interdialytic BP) is associated with worse clinical outcomes. However, this association is likely confounded by factors specific to dialysis patients. The relationship between BP and mortality appears to be more linear in patients with fewer underlying cardiovascular diseases and longer survival. Recent studies have indicated that BP measurements taken outside of dialysis sessions, such as standardized BP on nondialysis days, home BP, and ambulatory BP monitoring between HD sessions, are more predictive of clinical outcomes. Due to the varied effects of dialysis-related treatment practices on BP, there is a lack of data from large-scale clinical trials. As a result, it is challenging to provide strong recommendations for BP targets directly applicable to dialysis patients. This review addresses various factors influencing BP in dialysis patients, including the establishment of individualized target BP levels and discussions on maintenance strategies, while incorporating a recent literature review.

Glycated hemoglobin levels and risk of all-cause and cause-specific mortality in hemodialysis patients with diabetes

AIM

Adequate glycemic control is fundamental for improving clinical outcomes in hemodialysis patients with diabetes. However, the target for glycated hemoglobin (HbA1c) level and whether cause-specific mortality differs based on HbA1c levels remain unclear.

METHODS

A total of 24,243 HD patients with diabetes were enrolled from a multicenter, nationwide registry. We examined the association between HbA1c levels and the risk of all-cause and cause-specific mortality.

RESULTS

Compared to patients with HbA1c 6.5%-7.5%, patients with HbA_1c 8.5-9.5% and ≥9.5% were associated with a 1.26-fold (95% CI, 1.12-1.42) and 1.56-fold (95% CI, 1.37-1.77) risk for all-cause mortality. The risk of all-cause mortality did not increase in patients with HbA1c < 5.5%. In cause-specific mortality, the risk of cardiovascular deaths significantly increased from small increase of HbA1c levels. However, the risk of other causes of death increased only in patients with HbA1c > 9.5%. The slope of HR increase with increasing HbA1c levels was significantly faster for cardiovascular causes than for other causes.

CONCLUSIONS

There was a linear relationship between HbA1c levels and risk of all-cause mortality in hemodialysis patients, and the risk of cardiovascular death increased earlier and more rapidly, with increasing HbA1c levels, compared with other causes of death.