Predictive value of machine learning for the risk of acute kidney injury (AKI) in hospital intensive care units (ICU) patients: a systematic review and meta-analysis.

Background: Recent studies suggest machine learning represents a promising predictive option for patients in intensive care units (ICU). However, the machine learning performance regarding its actual predictive value for early detection in acute kidney injury (AKI) patients remains uncertain. Objective: This study represents the inaugural meta-analysis aiming to investigate the predictive value of machine learning for assessing the risk of AKI among ICU patients. Methods: PubMed, Web of Science, Embase, and the Cochrane Library were all thoroughly searched from inception to June 25, 2022. Eligible studies for inclusion were those concentrating on the predictive value and the development, validation, or enhancement of a prediction model for AKI patients in the ICU. Measures of effects, including c-index, sensitivity, specificity, and their corresponding 95% confidence intervals (CIs), were employed for analysis. The risk of bias in the included original studies was assessed using Probst. The meta-analysis in our study was carried out using R version 4.2.0. Results: The systematic search yielded 29 articles describing 13 machine-learning models, including 86 models in the training set and 57 in the validation set. The overall c-index was 0.767 (95% CI [0.746, 0.788]) in the training set and 0.773 (95% CI [0.741, 0.804]) in the validation set. The sensitivity and specificity of included studies are as follows: sensitivity [train: 0.66 (95% CI [0.59, 0.73]), validation: 0.73 (95% CI [0.68, 0.77])]; and specificity [train: 0.83 (95% CI [0.78, 0.87])], validation: 0.75 (95% CI [0.71, 0.79])]. Conclusion: The machine learning-based method for predicting the risk of AKI in hospital ICU patients has excellent predictive value and could potentially serve as a prospective application strategy for early identification. PROSPERO Registration number ID: CRD42022362838.

Clinical predictors for nondiabetic kidney diseases in patients with type 2 diabetes mellitus: a retrospective study from 2017 to 2021.

BACKGROUND: Nondiabetic kidney disease (NDKD), which is prevalent among patients with diabetes mellitus (DM), is considerably different from diabetic kidney disease (DKD) in terms of the pathological features, treatment strategy and prognosis. Although renal biopsy is the current gold-standard diagnostic method, it cannot be routinely performed due to a range of risks. The aim of this study was to explore the predictors for differentiating NDKD from DKD to meet the urgent medical needs of patients who cannot afford kidney biopsy. METHODS: This is a retrospective study conducted by reviewing the medical records of patients with type 2 DM who underwent percutaneous renal biopsy at the Affiliated Hospital of Guizhou Medical University between January 2017 and May 2021. The demographic data, clinical data, blood test results, and pathological examination results of the patients were obtained from their medical records. Multivariate regression analysis was performed to evaluate the predictive factors for NDKD. RESULTS: A total of 244 patients were analyzed. The median age at biopsy was 55 (46, 62) years. Patients diagnosed with true DKD, those diagnosed with NDKD and those diagnosed with NDKD superimposed DKD represented 48.36% (118/244), 45.9% (112/244) and 5.74% (14/244), respectively, of the patient population. Immunoglobulin A nephropathy was the most common type of lesion in those with NDKD (59, 52.68%) and NDKD superimposed DKD (10, 71.43%). Independent predictive indicators for diagnosing NDKD included a DM duration of less than 5 years (odds ratio [OR] = 4.476; 95% confidence interval [CI]: 2.257-8.877; P < 0.001), an absence of diabetic retinopathy (OR = 4.174; 95% CI: 2.049-8.502; P < 0.001), a high RBC count (OR = 1.901; 95% CI: 1.251-2.889; P = 0.003), and a negative of urinary glucose excretion test result (OR = 2.985; 95% CI: 1.474-6.044; P = 0.002).. CONCLUSIONS: A DM duration less than 5 years, an absence of retinopathy, a high RBC count and an absence of urinary glucose excretion were independent indicators for the diagnosis of NDKD, suggesting that patients with NDKD may require a different treatment regimen than those with DKD.

Efficacy and safety of angiotensin converting enzyme inhibitors and angiotensin II receptor blockers in diabetic nephropathy: a systematic review and meta-analysis.

BACKGROUND: Diabetic nephropathy (DN) is one of the most typical microangiopathies caused by diabetes. It often leads enormous physiological and psychological burdens for patients and seriously affects their quality of life. Therefore, effective combination therapy is necessary for these patients. In this study, we performed a meta-analysis to systematically evaluate and discuss the efficacy and safety of angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin II receptor blockers (ARBs) in the treatment of DN. METHODS: The PubMed, Embase, Web of Science, and Medline databases were selected as the sources of the literature search, and the search was limited to studies published in English. Studies related to ACEIs and ARBs in the treatment of DN published from January 2001 to January 2021 were included in this analysis. Meta-analysis was performed to calculate the reinforcement mean difference. RESULTS: In total, eight articles involving 1,893 cases with DN were included in this study. The results of this systematic review and meta-analysis showed that for patients with diabetic nephropathy，there were significant differences in 24-hour proteinuria [mean difference (MD) =-78.46, 95% confidence interval (95% CI): -80.25 to -76.66, P<0.00001], systolic blood pressure (MD =-9.11, 95% CI: -13.44 to -4.78, P<0.0001), and diastolic blood pressure (MD =-3.39, 95% CI: -5.68 to -1.11, P=0.004) between the combined ACEI and ARB group and the single ACEI or ARB group (P<0.05). In terms of safety, in addition to the significant difference in serum potassium (MD =0.1, 95% CI: 0.05 to 0.15, P=0.0001) between the combined ACEI and ARB group and the single drug group (P<0.1), there were no notable differences in serum creatinine (MD =0.66, 95% CI: -8.0 to 2.12, P=0.37), creatinine clearance (MD =-0.25, 95% CI: -0.62, 0.11, P=0.17), or the incidence of adverse reactions [odds ratio (OR) =1.19, 95% CI: 0.81 to 1.75, P=0.37]. DISCUSSION: A total of eight studies were included in this meta-analysis. The results showed that for patients with diabetic nephropathy, the combination of ACEI and ARB was more effective than ACEI or ARB alone, and also had higher safety.

Non-linear relationship between basal serum albumin concentration and cardiac arrest in critically ill patients with end-stage renal disease: a cross-sectional study.

OBJECTIVES: The aim of our study was to investigate the association between serum albumin concentration and the risk of cardiac arrest in critically ill patients with end-stage renal disease in the intensive care unit (ICU). DESIGN: This was a secondary analysis. SETTING: The Phillip electronic-ICU collaborative database from 2014 to 2015. PARTICIPANTS: This study included 4990 critically ill patients diagnosed with end-stage renal disease. PRIMARY AND SECONDARY OUTCOME MEASURES: The exposure of interest was serum albumin concentration. The outcome variable was cardiac arrest. RESULTS: A non-linear relationship was observed between serum albumin concentration and risk of cardiac arrest, with an inflection point of 3.26 g/dL after adjusting for potential confounders. The effect sizes and the CIs on the left and right sides of the inflection point were 0.88 (0.65 to 1.19) and 0.32 (0.16 to 0.64), respectively. CONCLUSIONS: Within an albumin range of 3.26-5.6 g/dL, each 1 g/dL increase in serum levels is associated with a 68% decrease of the risk of cardiac arrest in critically ill patients with end-stage renal disease.