Machine Learning Prediction Models for Mortality in Intensive Care Unit Patients with Lactic Acidosis.

BACKGROUND: Lactic acidosis is the most common cause of anion gap metabolic acidosis in the intensive care unit (ICU), associated with poor outcomes including mortality. We sought to compare machine learning (ML) approaches versus logistic regression analysis for prediction of mortality in lactic acidosis patients admitted to the ICU. METHODS: We used the Medical Information Mart for Intensive Care (MIMIC-III) database to identify ICU adult patients with lactic acidosis (serum lactate ≥4 mmol/L). The outcome of interest was hospital mortality. We developed prediction models using four ML approaches consisting of random forest (RF), decision tree (DT), extreme gradient boosting (XGBoost), artificial neural network (ANN), and statistical modeling with forward stepwise logistic regression using the testing dataset. We then assessed model performance using area under the receiver operating characteristic curve (AUROC), accuracy, precision, error rate, Matthews correlation coefficient (MCC), F1 score, and assessed model calibration using the Brier score, in the independent testing dataset. RESULTS: Of 1919 lactic acidosis ICU patients, 1535 and 384 were included in the training and testing dataset, respectively. Hospital mortality was 30%. RF had the highest AUROC at 0.83, followed by logistic regression 0.81, XGBoost 0.81, ANN 0.79, and DT 0.71. In addition, RF also had the highest accuracy (0.79), MCC (0.45), F1 score (0.56), and lowest error rate (21.4%). The RF model was the most well-calibrated. The Brier score for RF, DT, XGBoost, ANN, and multivariable logistic regression was 0.15, 0.19, 0.18, 0.19, and 0.16, respectively. The RF model outperformed multivariable logistic regression model, SOFA score (AUROC 0.74), SAP II score (AUROC 0.77), and Charlson score (AUROC 0.69). CONCLUSION: The ML prediction model using RF algorithm provided the highest predictive performance for hospital mortality among ICU patient with lactic acidosis.

Machine Learning Consensus Clustering Approach for Patients with Lactic Acidosis in Intensive Care Units.

BACKGROUND: Lactic acidosis is a heterogeneous condition with multiple underlying causes and associated outcomes. The use of multi-dimensional patient data to subtype lactic acidosis can personalize patient care. Machine learning consensus clustering may identify lactic acidosis subgroups with unique clinical profiles and outcomes. METHODS: We used the Medical Information Mart for Intensive Care III database to abstract electronic medical record data from patients admitted to intensive care units (ICU) in a tertiary care hospital in the United States. We included patients who developed lactic acidosis (defined as serum lactate ≥ 4 mmol/L) within 48 h of ICU admission. We performed consensus clustering analysis based on patient characteristics, comorbidities, vital signs, organ supports, and laboratory data to identify clinically distinct lactic acidosis subgroups. We calculated standardized mean differences to show key subgroup features. We compared outcomes among subgroups. RESULTS: We identified 1919 patients with lactic acidosis. The algorithm revealed three best unique lactic acidosis subgroups based on patient variables. Cluster 1 (n = 554) was characterized by old age, elective admission to cardiac surgery ICU, vasopressor use, mechanical ventilation use, and higher pH and serum bicarbonate. Cluster 2 (n = 815) was characterized by young age, admission to trauma/surgical ICU with higher blood pressure, lower comorbidity burden, lower severity index, and less vasopressor use. Cluster 3 (n = 550) was characterized by admission to medical ICU, history of liver disease and coagulopathy, acute kidney injury, lower blood pressure, higher comorbidity burden, higher severity index, higher serum lactate, and lower pH and serum bicarbonate. Cluster 3 had the worst outcomes, while cluster 1 had the most favorable outcomes in terms of persistent lactic acidosis and mortality. CONCLUSIONS: Consensus clustering analysis synthesized the pattern of clinical and laboratory data to reveal clinically distinct lactic acidosis subgroups with different outcomes.

Impact of Palliative Care Services on Treatment and Resource Utilization for Hepatorenal Syndrome in the United States.

Background: This study aimed to determine the rates of inpatient palliative care service use and assess the impact of palliative care service use on in-hospital treatments and resource utilization in hospital admissions for hepatorenal syndrome. Methods: Using the National Inpatient Sample, hospital admissions with a primary diagnosis of hepatorenal syndrome were identified from 2003 through 2014. The primary outcome of interest was the temporal trend and predictors of inpatient palliative care service use. Logistic and linear regression was performed to assess the impact of inpatient palliative care service on in-hospital treatments and resource use. Results: Of 5571 hospital admissions for hepatorenal syndrome, palliative care services were used in 748 (13.4%) admissions. There was an increasing trend in the rate of palliative care service use, from 3.3% in 2003 to 21.1% in 2014 (p < 0.001). Older age, more recent year of hospitalization, acute liver failure, alcoholic cirrhosis, and hepatocellular carcinoma were predictive of increased palliative care service use, whereas race other than Caucasian, African American, and Hispanic and chronic kidney disease were predictive of decreased palliative care service use. Although hospital admission with palliative care service use had higher mortality, palliative care service was associated with lower use of invasive mechanical ventilation, blood product transfusion, paracentesis, renal replacement, vasopressor but higher DNR status. Palliative care services reduced mean length of hospital stay and hospitalization cost. Conclusion: Although there was a substantial increase in the use of palliative care service in hospitalizations for hepatorenal syndrome, inpatient palliative care service was still underutilized. The use of palliative care service was associated with reduced resource use.

Acute Kidney Injury after Lung Transplantation: A Systematic Review and Meta-Analysis.

BACKGROUND: Lung transplantation has been increasingly performed worldwide and is considered an effective therapy for patients with various causes of end-stage lung diseases. We performed a systematic review to assess the incidence and impact of acute kidney injury (AKI) and severe AKI requiring renal replacement therapy (RRT) in patients after lung transplantation. METHODS: A literature search was conducted utilizing Ovid MEDLINE, EMBASE, and Cochrane Database from inception through June 2019. We included studies that evaluated the incidence of AKI, severe AKI requiring RRT, and mortality risk of AKI among patients after lung transplantation. Pooled incidence and odds ratios (ORs) with 95% confidence interval (CI) were obtained using random-effects meta-analysis. The protocol for this meta-analysis is registered with PROSPERO (International Prospective Register of Systematic Reviews; no. CRD42019134095). RESULTS: A total of 26 cohort studies with a total of 40,592 patients after lung transplantation were enrolled. Overall, the pooled estimated incidence rates of AKI (by standard AKI definitions) and severe AKI requiring RRT following lung transplantation were 52.5% (95% CI: 45.8-59.1%) and 9.3% (95% CI: 7.6-11.4%). Meta-regression analysis demonstrated that the year of study did not significantly affect the incidence of AKI (p = 0.22) and severe AKI requiring RRT (p = 0.68). The pooled ORs of in-hospital mortality in patients after lung transplantation with AKI and severe AKI requiring RRT were 2.75 (95% CI, 1.18-6.41) and 10.89 (95% CI, 5.03-23.58). At five years, the pooled ORs of mortality among patients after lung transplantation with AKI and severe AKI requiring RRT were 1.47 (95% CI, 1.11-1.94) and 4.79 (95% CI, 3.58-6.40), respectively. CONCLUSION: The overall estimated incidence rates of AKI and severe AKI requiring RRT in patients after lung transplantation are 52.5% and 9.3%, respectively. Despite advances in therapy, the incidence of AKI in patients after lung transplantation does not seem to have decreased. In addition, AKI after lung transplantation is significantly associated with reduced short-term and long-term survival.

Incidence of atrial fibrillation in pregnancy and clinical significance: A meta-analysis.

PURPOSE: Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia worldwide, and is associated with increased morbidity and mortality. However, the incidence and maternal/fetal outcomes of AF in pregnancy remain unclear. This study's aims were to investigate the pooled incidence of AF in pregnant women and to assess maternal/fetal outcomes of AF in pregnancy. MATERIAL AND METHODS: A literature search for studies that reported incidence of AF in pregnancy, was conducted using MEDLINE, EMBASE and Cochrane Database from inception through May 2018. Pooled incidence with 95%CI were calculated using a random-effect model. The protocol for this meta-analysis is registered with PROSPERO (International Prospective Register of Systematic Reviews; no. CRD42018095955). RESULTS: We identified 7 cohort studies including 301,638 pregnancies. The pooled estimated incidence of AF in pregnancy among women with no known heart disease, and those with structural heart disease was 0.3% (95%CI: 0.01%-40.6%) and 2.2% (95%CI: 0.96%-5.01%), respectively. Among women with known AF, the pooled estimated incidence of recurrent AF in pregnancy was 39.2% (95%CI: 16.9%-67.2%). The pooled estimated incidence of pre-eclampsia and congestive heart failure among pregnant patients with AF was 4.1% (95%CI: 2.1%-7.8%) and 9.6% (95%CI: 5.7%-15.9%), respectively. The pooled estimated incidence of fetal events including premature birth, small for gestational age, respiratory distress syndrome, intraventricular hemorrhage, death was 26.6% (95%CI: 20.4%-34.0%). CONCLUSION: The overall estimated incidence of AF and recurrent AF during pregnancy is as high as 2.2% and 39.2%, respectively. AF during pregnancy may result in poor maternal and fetal outcomes.

Epidemiology of atrial fibrillation in patients with cirrhosis and clinical significance: a meta-analysis.

OBJECTIVE: The epidemiology of atrial fibrillation (AF) in patients with cirrhosis and its clinical significance remain unclear. This study aimed (i) to investigate the pooled prevalence and/or incidence of AF in patients with cirrhosis and (ii) to assess the mortality risk of AF in patients with cirrhosis. PATIENTS AND METHODS: A literature search for studies that reported incidence of AF in patients with cirrhosis was carried out using Medline, Embase, and Cochrane Database from inception through July 2018. Pooled incidence with 95% confidence interval (CI) was calculated using a random-effect model. The protocol for this meta-analysis is registered with PROSPERO (International Prospective Register of Systematic Reviews; no. CRD42018102664). RESULTS: Seven cohort studies including 385 866 patients with cirrhosis were identified. The pooled estimated prevalence of AF in patients with cirrhosis was 5.0% (95% CI: 2.8-8.6%). When studies that solely assessed patients undergoing transplant evaluation or on transplant waiting list were excluded, the pooled estimated prevalence of AF in patients with cirrhosis was 7.4% (95% CI: 3.5-15.2%). There was a significant association between AF and increased mortality risk in cirrhotic patients with a pooled odds ratio of 1.44 (95% CI: 1.36-1.53). CONCLUSION: The overall estimated prevalence of AF among patients with cirrhosis is 5.0%. Our study demonstrates a statistically significant increased mortality risk in cirrhotic patients with AF.

Efficacy and safety of anticoagulation for atrial fibrillation in patients with cirrhosis: A systematic review and meta-analysis.

OBJECTIVE: The atrial fibrillation-related stroke is clearly prevented by anticoagulation treatment, however, management of anticoagulation for AF in patients with cirrhosis represents a challenge due to bleeding concerns. To address this issue, a systematic review and meta-analysis of the literature was performed. METHODS: A literature search for studies reporting the incidence of AF in patients with cirrhosis was conducted using MEDLINE, EMBASE and Cochrane Database, from inception through July 2018. RESULTS: 7 cohort studies including 19,798 patients with AF and cirrhosis were identified. The use of anticoagulation (%) among included studies ranged from 8.3% to 53.9%. Anticoagulation use for AF in patients with cirrhosis was significantly associated with a reduced risk of stroke, with a pooled HR of 0.58 (95%CI: 0.35-0.96). When compared with no anticoagulation, the use of anticoagulation was not significantly associated with a higher risk of bleeding, with a pooled HR of 1.45 (95%CI: 0.96-2.17). Compared to warfarin, the use of direct oral anticoagulants (DOACs) was associated with a lower risk of bleeding among AF patients with cirrhosis. CONCLUSION: Our study demonstrates that anticoagulation use for AF in patients with cirrhosis is associated with a reduced risk of stroke, without increasing significantly the risk of bleeding, when compared to those without anticoagulation.

Atrial fibrillation following heart transplantation: A systematic review and meta-analysis of observational studies.

OBJECTIVE: Previous studies have suggested a high incidence of atrial fibrillation (AF) in heart transplant recipients. However, incidence trends of AF in heart transplant recipients remain unclear. The study's aims were (1) to investigate the pooled incidence/incidence trends of AF following heart transplantation and (2) to assess the mortality risk of heart transplant recipients with AF. METHODS: A literature search for studies that reported the incidence of AF following heart transplantation was conducted using MEDLINE, EMBASE, and The Cochrane Database from inception through March 2018. Pooled incidence and odds ratios (OR) with 95%CI were calculated using a random-effects model. RESULTS: Eighteen studies (2 cohorts from clinical trials and 16 cohort studies) with 5393 heart transplant recipients were enrolled. The pooled estimated incidence of AF in heart transplant was 10.1% (95%CI: 7.6%-13.2%). Meta-analysis based on the type of anastomotic technique demonstrated a pooled estimated incidence of AF following heart transplantation of 18.7% (95%CI: 10.3%-31.5%) and 11.1% (95%CI: 6.5%-18.4%) by biatrial and bicaval techniques, respectively. There was a significant association between AF following a heart transplant and increased mortality risk with a pooled OR of 2.86 (95%CI: 2.08-3.93). Meta-regression analyses showed no significant correlations between the year of study and incidence of AF (P = 0.47) or mortality risk of AF after heart transplantation (P = 0.99). CONCLUSIONS: The overall estimated incidence of AF following heart transplantation is 10.1%. There is a significant association between AF and increased mortality after transplantation. Furthermore, incidence and mortality risk of AF following heart transplant does not seem to decrease over time.

Proton pump inhibitors linked to hypomagnesemia: a systematic review and meta-analysis of observational studies.

BACKGROUND: The reported risk of hypomagnesemia in patients with proton pump inhibitor (PPI) use is conflicting. The objective of this meta-analysis was to assess the association between the use of PPIs and the risk of hypomagnesemia. METHODS: A literature search of observational studies was performed using MEDLINE, EMBASE and Cochrane Database of Systematic Reviews from inception through September 2014. Studies that reported odd ratios or hazard ratios comparing the risk of hypomagnesemia in patients with PPI use were included. Pooled risk ratios (RRs) and 95% confidence interval (CI) were calculated using a random-effect, generic inverse variance method. RESULTS: Nine observational studies (three cohort studies, five cross-sectional studies and a case-control study) with a total of 109,798 patients were identified and included in the data analysis. The pooled RR of hypomagnesemia in patients with PPI use was 1.43 (95% CI, 1.08-1.88). The association between the use of PPIs and hypomagnesemia remained significant after the sensitivity analysis including only studies with high quality score (Newcastle-Ottawa scale score ≥ 8) with a pooled RR of 1.63 (95% CI, 1.14-2.23). CONCLUSIONS: Our study demonstrates a statistically significant increased risk of hypomagnesemia in patients with PPI use. The finding of this meta-analysis of observational studies suggests that PPI use is associated with hypomagnesemia and may impact clinical management of patients who are taking PPIs and at risk for hypomagnesemia related cardiovascular events.

Comparison of renal outcomes in off-pump versus on-pump coronary artery bypass grafting: A systematic review and meta-analysis of randomized controlled trials.

AIM: The objective of this meta-analysis was to compare the effects of off-pump and on-pump coronary artery bypass grafting (CABG) on acute kidney injury (AKI) and the need of dialysis after surgery. METHODS: Comprehensive literature searches for randomized controlled trials (RCTs) of CABG with on-pump and off-pump was performed using MEDLINE, EMBASE, Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials Systematic Reviews and clinicaltrials.gov from inception through September 2014. Primary outcomes were the incidence of AKI and the need of dialysis. Mortality was assessed among the studies that reported renal outcomes. Pooled risk ratios (RRs) and 95% confidence interval (CI) were calculated using a random-effect, generic inverse variance method. RESULTS: Thirty-three RCTs with 17 322 patients were enrolled in our study. Patients in the off-pump CABG group had overall lower incidence of AKI (19.1%) compared with the on-pump CABG group (22.2%). There was a protective effect of off-pump CABG on the incidence of AKI compared with the on-pump CABG group (RR: 0.87; 95% CI: 0.77-0.98). However, there was no significant difference in the need for dialysis in the off-pump group compared with the on-pump group (RR: 0.84; 95% CI 0.63-1.13). Within the selected trials, post hoc analysis assessing the mortality outcome demonstrated a pooled RR of 0.97 (95% CI, 0.77-1.23) in off-pump versus on-pump CABGs. CONCLUSIONS: Our study demonstrates a beneficial effect of off-pump CABG on the incidence of AKI. However, our meta-analysis does not show benefits of the need of dialysis or survival among patients undergoing off-pump CABG.

Preoperative renin-angiotensin system inhibitors use linked to reduced acute kidney injury: a systematic review and meta-analysis.

BACKGROUND: Previous trials of interventions to prevent acute kidney injury (AKI) have been unsuccessful and additional interventions are needed. Existing reviews of preoperative renin-angiotensin system (RAS) inhibitors have suggested harm. We included more recent studies and conducted this meta-analysis to evaluate the risk of postoperative AKI in patients who received preoperative RAS inhibitors. METHODS: A literature search was performed using MEDLINE, EMBASE and Cochrane Database of Systematic Reviews from inception through October, 2014. Studies that reported relative risks, odds ratios or hazard ratios comparing the AKI risk in patients who received preoperative RAS inhibitors versus those who did not were included. We performed the prespecified sensitivity analysis including only propensity score-based studies. Mortality risk was evaluated among the studies that reported AKI outcome. Pooled risk ratios (RR) and 95% confidence interval (CI) were calculated using a random-effect, generic inverse variance method. RESULTS: Twenty-four studies (1 randomized controlled trial and 23 cohort studies) with 102 675 patients were included in the analysis to assess the risk of postoperative AKI and preoperative RAS inhibitors use. The pooled RR of AKI in patients receiving RAS inhibitors was 1.05 (95% CI: 0.92-1.20). The meta-analysis of the RCT and 11 studies with propensity score analysis demonstrated the pooled RR of AKI in patients receiving RAS inhibitors of 0.92 (95% CI: 0.85-0.99). Within the selected studies, preoperative RAS inhibitor therapy was not associated with a significant increase or decrease in mortality (RR: 0.93; 95% CI: 0.80-1.09). CONCLUSIONS: Our meta-analysis demonstrates an association between preoperative RAS inhibitor treatment and lower incidence of AKI.

Periprocedural effects of statins on the incidence of contrast-induced acute kidney injury: a systematic review and meta-analysis of randomized controlled trials.

BACKGROUND: The reports on the efficacy of statins for the prevention of contrast-induced acute kidney injury (CIAKI) remain controversial. The objective of this meta-analysis was to assess the effect of statins for the prevention of CIAKI. METHODS: Comprehensive literature searches for randomized controlled trials (RCTs) of periprocedural statin treatment for prevention of CIAKI were performed using MEDLINE, EMBASE, Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials Systematic Reviews and clinicaltrials.gov from inception until May 2014. The primary outcome was the incidence of CIAKI. RESULTS: Thirteen prospective RCTs were included in our analysis. Of 5803 patients with contrast exposures, 304 patients (5.2%) had CIAKI. Patients in the statin group had an overall lower incidence of CIAKI (3.6%) compared to the control group (6.9%). Intravenous (IV) fluid hydration was used in both groups of all included studies for prevention of CIAKI. There was a significant protective effect of periprocedural statins on the incidence of CIAKI when compared to the control group [risk ratios (RRs): 0.49; 95% CI: 0.37-0.66, I(2) of 25%]. CONCLUSIONS: Our study demonstrates a statistically significant protective effect of statin treatment during procedures with contrast exposures. This finding suggests the use of statins in addition to standard IV crystalloid hydration may be beneficial in the prevention of CIAKI.

The Risk of Coronary Heart Disease in Patients with Kidney Stones: A Systematic Review and Meta-analysis.

BACKGROUND: The reported risk of coronary heart disease (CHD) in patients with a history of kidney stones is conflicting. AIMS: The objective of this meta-analysis was to assess the association between a history of kidney stones and CHD risk. MATERIALS AND METHODS: A literature search was performed using MEDLINE, EMBASE, and Cochrane Database of Systematic Reviews from inception until April 04, 2014. Studies that reported odds ratios or hazard ratios comparing the risk of CHD in patients with a history of kidney stones versus those without a history of kidney stones were included. Pooled risk ratios (RRs) and 95% confidence interval (CI) were calculated using a random-effect, generic inverse variance method. RESULTS: Seven study populations from four cohort studies and one cross-sectional study were identified and included in the data analysis. The pooled risk ratio (RR) of CHD in patients with kidney stones was 1.24 (95% CI, 1.10-1.40). This result remained significant (RR, 1.23 [95% CI, 1.08-1.41]) when the sensitivity analysis was restricted to only cohort studies. A history of kidney stones was associated with increased CHD risk in females (RR, 1.43 [95% CI, 1.12-1.82]), whereas the association was not significant in males (RR, 1.14 [95% CI, 0.94-1.38]). CONCLUSIONS: Our study demonstrates a statistically significant increased risk of CHD in female patients with prior kidney stones. This finding suggests that a history of kidney stones is a risk factor for CHD in females and may impact clinical management.

Associations of sugar-sweetened and artificially sweetened soda with chronic kidney disease: a systematic review and meta-analysis.

BACKGROUND/OBJECTIVES: The risk of chronic kidney disease (CKD) in patients who regularly drink soda is controversial. The objective of this meta-analysis was to evaluate the associations between consumption of sugar-sweetened and artificially sweetened soda and CKD. METHODS: A literature search was performed using MEDLINE, EMBASE and the Cochrane Database of Systematic Reviews from inception until 30 June 2014. Studies that reported odds ratios or hazard ratios comparing the risk of CKD in patients consuming significant amounts of either sugar-sweetened or artificially sweetened soda versus those who did not consume soda were included. Pooled risk ratios (RR) and 95% confidence intervals (CI) were calculated using a random-effects, generic inverse variance method. RESULTS: Five studies were included in our analysis of the association between consumption of sugar-sweetened soda and CKD. The pooled RR of CKD in patients consuming sugar-sweetened soda was 1.58 (95% CI 1.00-2.49). Four studies were selected to assess the association between consumption of artificially sweetened soda and CKD. The pooled RR of CKD in patients consuming artificially sweetened soda was 1.33 (95% CI 0.82-2.15). CONCLUSIONS: Our study demonstrates statistically significant increased risks of CKD in patients consuming sugar-sweetened soda, but not in patients consuming artificially sweetened soda. This finding suggests that sugar-sweetened soda consumption is associated with CKD and may impact clinical management and primary prevention of CKD in high-risk patients.

Oral hydration for prevention of contrast-induced acute kidney injury in elective radiological procedures: a systematic review and meta-analysis of randomized controlled trials.

BACKGROUND: The reports on efficacy of oral hydration treatment for the prevention of contrast-induced acute kidney injury (CIAKI) in elective radiological procedures and cardiac catheterization remain controversial. AIMS: The objective of this meta-analysis was to assess the use of oral hydration regimen for prevention of CIAKI. MATERIALS AND METHODS: Comprehensive literature searches for randomized controlled trials (RCTs) of outpatient oral hydration treatment was performed using MEDLINE, EMBASE, Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials Systematic Reviews, and clinicaltrials.gov from inception until July 4(th), 2014. Primary outcome was the incidence of CIAKI. RESULTS: Six prospective RCTs were included in our analysis. Of 513patients undergoing elective procedures with contrast exposures,45 patients (8.8%) had CIAKI. Of 241 patients with oral hydration regimen, 23 (9.5%) developed CIAKI. Of 272 patients with intravenous (IV) fluid regimen, 22 (8.1%) had CIAKI. Study populations in all included studies had relatively normal kidney function to chronic kidney disease (CKD) stage 3. There was no significant increased risk of CIAKI in oral fluid regimen group compared toIV fluid regimen group (RR = 0.94, 95% confidence interval, CI = 0.38-2.31). CONCLUSIONS: According to our analysis,there is no evidence that oral fluid regimen is associated with more risk of CIAKI in patients undergoing elective procedures with contrast exposures compared to IV fluid regimen. This finding suggests that the oral fluid regimen might be considered as a possible outpatient treatment option for CIAKI prevention in patients with normal to moderately reduced kidney function.