Development and external validation of an acute kidney injury risk score for use in the general population

Development and validation of a medication-based risk prediction model for acute kidney injury in older outpatients

BACKGROUND

Older adults are at an increased risk of acute kidney injury (AKI), particularly in community settings, often due to medications. Effective prevention hinges on identifying high-risk patients, yet existing models for predicting AKI risk in older outpatients are scarce, particularly those incorporating medication variables. We aimed to develop an AKI risk prediction model that included medication-related variables for older outpatients.

METHODS

We constructed a cohort of 2,272,257 outpatients aged ≥65 years using a national claims database. This cohort was split into a development (70%) and validation (30%) groups. Our primary goal was to identify newly diagnosed AKI within one month of cohort entry in an outpatient context. We screened 170 variables and developed a risk prediction model using logistic regression.

RESULTS

The final model integrated 12 variables: 2 demographic, 4 comorbid, and 6 medication-related. It showed good performance with acceptable calibration. In the validation cohort, the area under the receiver operating characteristic curve value was 0.720 (95% confidence interval, 0.692-0.748). Sensitivity and specificity were 69.9% and 61.9%, respectively. Notably, the model identified high-risk patients as having a 27-fold increased AKI risk compared with low-risk individuals.

CONCLUSION

We have developed a new AKI risk prediction model for older outpatients, incorporating critical medication-related variables with good discrimination. This tool may be useful in identifying and targeting patients who may require interventions to prevent AKI in an outpatient setting.

Recovery of kidney function after acute kidney disease-a multi-cohort analysis

BACKGROUND

There are no consensus definitions for evaluating kidney function recovery after acute kidney injury (AKI) and acute kidney disease (AKD), nor is it clear how recovery varies across populations and clinical subsets. We present a federated analysis of four population-based cohorts from Canada, Denmark and Scotland, 2011-18.

METHODS

We identified incident AKD defined by serum creatinine changes within 48 h, 7 days and 90 days based on KDIGO AKI and AKD criteria. Separately, we applied changes up to 365 days to address widely used e-alert implementations that extend beyond the KDIGO AKI and AKD timeframes. Kidney recovery was based on resolution of AKD and a subsequent creatinine measurement below 1.2× baseline. We evaluated transitions between non-recovery, recovery and death up to 1 year; within age, sex and comorbidity subgroups; between subset AKD definitions; and across cohorts.

RESULTS

There were 464 868 incident cases, median age 67-75 years. At 1 year, results were consistent across cohorts, with pooled mortalities for creatinine changes within 48 h, 7 days, 90 days and 365 days (and 95% confidence interval) of 40% (34%-45%), 40% (34%-46%), 37% (31%-42%) and 22% (16%-29%) respectively, and non-recovery of kidney function of 19% (15%-23%), 30% (24%-35%), 25% (21%-29%) and 37% (30%-43%), respectively. Recovery by 14 and 90 days was frequently not sustained at 1 year. Older males and those with heart failure or cancer were more likely to die than to experience sustained non-recovery, whereas the converse was true for younger females and those with diabetes.

CONCLUSION

Consistently across multiple cohorts, based on 1-year mortality and non-recovery, KDIGO AKD (up to 90 days) is at least prognostically similar to KDIGO AKI (7 days), and covers more people. Outcomes associated with AKD vary by age, sex and comorbidities such that older males are more likely to die, and younger females are less likely to recover.

Acute Kidney Injury after Endoscopic Retrograde Cholangiopancreatography-A Hospital-Based Prospective Observational Study

Background: Endoscopic retrograde cholangiopancreatography (ERCP) represents a major pivotal point in gastrointestinal endoscopy. Little is known about acute kidney injury (AKI) post-ERCP. This study analyses the incidence, risk factors, and prognosis of post-ERCP AKI. Methods: A total of 396 patients were prospectively studied. AKI was defined by an increase in serum creatinine (SCr) ≥ 0.3 mg/dL or by an increase in SCr ≥ 50% in the first 48 h post-ERCP. Logistic regression analysis was used to identify the predictors of AKI and in-hospital mortality. A two-tailed p value < 0.05 was considered significant. Results: One hundred and three patients (26%) developed post-ERCP AKI. Estimated glomerular filtration rate (adjusted odds ratio (aOR) = 0.95, 95% confidence interval (CI): 0.94−0.96, p < 0.001), nonrenal Charlson Comorbidity Index (Aor = 1.19, 95% CI: 1.05−1.35, p = 0.006), choledocholithiasis (aOR = 4.05, 95% CI: 1.98−8.29, p < 0.001), and bilirubin (aOR = 1.1, 95% CI: 1.05−1.15, p < 0.001) were associated with post-ERCP AKI. Post-ERCP AKI was associated with longer hospital stay (p < 0.001) and with increased in-hospital mortality (7.76% versus 0.36%, p < 0.001). Moderate-to-severe (stage 2 and 3) AKI was independently associated with in-hospital mortality (aOR = 6.43, 95% CI: 1.48−27.88, p < 0.013). Conclusions: Post-ERCP AKI represented an important complication associated with longer hospital stay. Moderate-to-severe post-ERCP AKI was an independent risk factor for in-hospital mortality.

Artificial Intelligence in Acute Kidney Injury Prediction

The use of artificial intelligence (AI) in nephrology and its associated clinical research is growing. Recent years have seen increased interest in utilizing AI to predict the development of hospital-based acute kidney injury (AKI). Several AI techniques have been employed to improve the ability to detect AKI across a variety of hospitalized settings. This review discusses the evolutions of AKI risk prediction discussing the static risk assessment models of yesteryear as well as the more recent trend toward AI and advanced learning techniques. We discuss the relative improvement in AKI detection as well as the relative dearth of data around the clinical implementation and patient outcomes using these models. The use of AI for AKI detection and clinical care is in its infancy, and this review describes how we arrived at our current position and hints at the promise of the future.

SGLT2 Inhibitors and the Risk of Acute Kidney Injury in Older Adults With Type 2 Diabetes

RATIONALE & OBJECTIVE

Sodium-glucose cotransporter 2 (SGLT2) inhibitors have been found to have many benefits for patients with type 2 diabetes. However, whether SGLT2 inhibitors increase the risk of acute kidney injury (AKI) remains unknown. We examined the association of AKI hospitalization with prior initiation of an SGLT2 inhibitor compared with initiation of a dipeptidyl peptidase 4 (DPP-4) inhibitor or a glucagon-like peptide 1 receptor agonist (GLP-1RA) among older adults with type 2 diabetes in routine practice.

STUDY DESIGN

Population-based cohort study.

SETTING & PARTICIPANTS

Older adults aged at least 66 years with type 2 diabetes enrolled in Medicare fee-for-service and who were new users of SGLT2 inhibitor, DPP-4 inhibitor, or GLP-1RA agents in the interval from March 2013 to December 2017.

EXPOSURES

New use of an SGLT2 inhibitor versus new use of a DPP-4 inhibitor or GLP-1RA.

OUTCOME

The primary outcome was hospitalization for AKI, defined as a discharge diagnosis of AKI in the primary or secondary position.

ANALYTICAL APPROACH

New users of SGLT2 inhibitors were matched at a 1:1 ratio to new users of DPP-4 inhibitors or GLP-1RAs using propensity scores in 2 pairwise comparisons. Cox proportional hazards regression models generated hazard ratios (HRs) with 95% CIs in propensity score-matched groups.

RESULTS

Totals of 68,130 and 71,477 new users of SGLT2 inhibitors were matched to new users of DPP-4 inhibitors or GLP-1RAs, respectively. Overall, the mean age of study participants was 72 years. The risk of AKI was lower in the SGLT2 inhibitor group than in the DPP-4 inhibitor group (HR, 0.71 [95% CI, 0.65-0.76]) or the GLP-1RA group (HR, 0.81 [95% CI, 0.75-0.87]).

LIMITATIONS

Residual confounding and lack of laboratory data.

CONCLUSIONS

Among older adults with type 2 diabetes, initiation of an SGLT2 inhibitor was associated with a reduced risk of AKI compared with initiation of a DPP-4 inhibitor or a GLP-1RA.

Development and Validation of a Model to Predict Severe Hospital-Acquired Acute Kidney Injury in Non-Critically Ill Patients

BACKGROUND

The current models developed to predict hospital-acquired AKI (HA-AKI) in non-critically ill fail to identify the patients at risk of severe HA-AKI stage 3.

OBJECTIVE

To develop and externally validate a model to predict the individual probability of developing HA-AKI stage 3 through the integration of electronic health databases.

METHODS

Study set: 165,893 non-critically ill hospitalized patients. Using stepwise logistic regression analyses, including demography, chronic comorbidities, and exposure to risk factors prior to AKI detection, we developed a multivariate model to predict HA-AKI stage 3. This model was then externally validated in 43,569 non-critical patients admitted to the validation center.

RESULTS

The incidence of HA-AKI stage 3 in the study set was 0.6%. Among chronic comorbidities, the highest odds ratios were conferred by ischemic heart disease, ischemic cerebrovascular disease, chronic congestive heart failure, chronic obstructive pulmonary disease, chronic kidney disease and liver disease. Among acute complications, the highest odd ratios were associated with acute respiratory failure, major surgery and exposure to nephrotoxic drugs. The model showed an AUC of 0.906 (95% CI 0.904 to 0.908), a sensitivity of 89.1 (95% CI 87.0-91.0) and a specificity of 80.5 (95% CI 80.2-80.7) to predict HA-AKI stage 3, but tended to overestimate the risk at low-risk categories with an adequate goodness-of-fit for all risk categories (Chi2: 16.4, p: 0.034). In the validation set, incidence of HA-AKI stage 3 was 0.62%. The model showed an AUC of 0.861 (95% CI 0.859-0.863), a sensitivity of 83.0 (95% CI 80.5-85.3) and a specificity of 76.5 (95% CI 76.2-76.8) to predict HA-AKI stage 3 with an adequate goodness of fit for all risk categories (Chi2: 15.42, p: 0.052).

CONCLUSIONS

Our study provides a model that can be used in clinical practice to obtain an accurate dynamic assessment of the individual risk of HA-AKI stage 3 along the hospital stay period in non-critically ill patients.

Development and Validation of a Nomogram for Predicting Drug-Induced Acute Kidney Injury in Hospitalized Patients: A Case-Control Study Based on Propensity-Score Matching

Background: Drug-induced acute kidney injury (D-AKI) is associated with increased mortality and longer hospital stays. This study aims to establish a nomogram to predict the occurrence of D-AKI in hospitalized patients in a multi-drug environment.

Methods: A single center retrospective study among adult hospitalized patients was conducted from July 2019 to September 2019 based on the Adverse Drug Events Active Surveillance and Assessment System-2 developed by our hospital. According to the propensity score matching algorithm, four controls per case were matched to eliminate the confounding bias caused by individual baseline variables. The predictors for D-AKI were obtained by logistic regression equation and used to establish the nomogram.

Results: Among 51,772 hospitalized patients, 332 were diagnosed with D-AKI. After matching, 288 pairs and 1,440 patients were included in the study, including 1,005 cases in the development group and 435 cases in the validation group. Six variables were independent predictors for D-AKI: alcohol abuse, the concurrent use of nonsteroidal anti-inflammatory drugs or diuretics, chronic kidney disease, lower baseline red blood cell count and neutrophil count ≥7 × 10⁹/L. The area under the curve (AUC) of the prediction model in the development group and validation group were 0.787 (95%CI, 0.752–0.823) and 0.788 (95%CI, 0.736–0.840), respectively. The GiViTI calibration belts showed that the model had a good prediction accuracy for the occurrence of D-AKI (p > 0.05).

Conclusion: This nomogram can help identify patients at high risk of D-AKI, which was useful in preventing the progression of D-AKI and treating it in the early stages.

Artificial Intelligence in Acute Kidney Injury: From Static to Dynamic Models

Artificial intelligence (AI) is the development of computer systems that normally require human intelligence. In the field of acute kidney injury (AKI) AI has led to an evolution of risk prediction models. In the past, static prediction models were developed using baseline (eg, preoperative) data to evaluate AKI risk. Newer models which incorporated baseline as well as evolving data collected during a hospital admission have shown improved predicative abilities. In this review, we will summarize the advances made in AKI risk prediction over the last several years, including a shift toward more dynamic, real-time, electronic medical record-based models. In addition, we will be discussing the role of electronic AKI alerts and decision support tools. Recent studies have demonstrated improved patient outcomes through the use of these tools which monitor for nephrotoxin medication exposures as well as provide kidney focused care bundles for patients at high risk for severe AKI. Finally, we will briefly discuss the pitfalls and implications of implementing these scores, alerts, and support tools.