Integrating electronic health data records to develop and validate a predictive model of hospital-acquired acute kidney injury in non-critically ill patients

Abstract

Background Models developed to predict hospital-acquired AKI (HA-AKI) in non-critically ill patients have a low sensitivity, do not include dynamic changes of risk factors and do not allow to establish a time relationship between exposure to risk factors and AKI. We developed and externally validated a predictive model of HA-AKI, integrating electronic health databases and recording the exposure to risk factors prior to the detection of AKI. Methods Study set: 36,852 non-critically ill hospitalized patients admitted from January to December 2017. Using stepwise logistic analyses, including demography, chronic comorbidities, and exposure to risk factors prior to AKI detection, we developed a multivariate model to predict HA-AKI. This model was then externally validated in 21,545 non-critical patients admitted to the validation center in the period from June 2017 to December 2018. Results The incidence of AKI in the study set was 3.9%. Among chronic comorbidities, the highest odds ratios, were conferred by chronic kidney disease, urologic and liver disease. Among acute complications, the highest odd ratios were associated with acute respiratory failure, anemia, systemic inflammatory response syndrome (SIRS), circulatory shock and major surgery. The model showed an AUC of 0.907(95% CI 0.902 to 0.908), a sensitivity of 82.7 (95% CI 80.7–84.6) and a specificity of 84.2 (95% CI 83.9-84.6) to predict HA-AKI, with an adequate goodness-of-fit for all risk categories (Chi2:6.02, p:0.64). In the validation set, prevalence of AKI was 3.2%. The model showed an AUC of 0.905 (95% CI 0.904-0.910) a sensitivity of 81.2 (95% CI 79.2–83.1) and a specificity of 82.5 (95% CI 82.2- 83) to predict HA-AKI and had an adequate goodness-of-fit for all risk categories (Chi2:4.2, p:0.83). An online tool predaki.amalfianalytics.com is available to calculate the risk of AKI in other hospital environments. Conclusions By using electronic health data records, our study provides a model that can be used in clinical practice to obtain an accurate dynamic and updated assessment of the individual risk of HA-AKI along the hospital admission period in non-critically ill patients.

Definitions, phenotypes, and subphenotypes in acute kidney injury—Moving towards precision medicine

“…These risk factors modify prognosis in patients who develop AKI. 46 CKD is the most important AKI risk factor in most studies, 4,[47][48][49] and probably often occurs on a background of subclinical CKD. 50 Available studies suggest that risk factors should be evaluated in any setting where there is an 'exposure' and used to triage patients into high and low AKI risk categories.…”

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

“…In patients with a code of AKI = 1, the exposure to these risk factors only was classified as =1 when it occurred within a maximum period of time prior to HA-AKI stage 3 detection (48 h for anemia, SIRS and shock, 72 h for contrast dyes and surgery and 7 days for nephrotoxic drugs). The procedures for the interrelation among the different electronic databases carried out to obtain the information on the clinical variables along hospital stay have been detailed in a previous report [28]. Unlike the hemoglobin level, arterial oxygen saturation, heart rate, respiratory rate or blood pressure level, that being numerical variables could be directly transferred to the general database, both circulatory shock and SIRS are complex variables that, to be automatically detected using a software-guided detection code, required the integration of data from various electronic records and the definition of classification algorithms.…”

“…Unlike the hemoglobin level, arterial oxygen saturation, heart rate, respiratory rate or blood pressure level, that being numerical variables could be directly transferred to the general database, both circulatory shock and SIRS are complex variables that, to be automatically detected using a software-guided detection code, required the integration of data from various electronic records and the definition of classification algorithms. In both cases, before using them in statistical analyses, we analyzed the accuracy of the automatic detection systems in a sample of 3426 patients, as previously detailed [28].…”

“…Table 3 (Chi 2 : 16.4, p: 0.034). The results of the stepwise forward procedures done to develop the final logistic model, including changes in the likelihood ratios, Cox and Snell R 2 , Nagelkerke R 2 and AIC are summarized in the previous report [28].…”

“…Wu L. et al recently published an article where the risk factors that predict the presentation of severe AKI were defined, but they included both ICU and non-ICU patients and no external validation was performed [27]. Our group recently developed a model that overcame some of those limitations and provides an accurate dynamic assessment of the individual risk of suffering AKI along the whole hospital stay period in patients admitted into non-critical hospitalization wards [28]. However, although this model allows AKI to be accurately predicted, because of a lack of statistical power, it does not allow to detect the risk of developing AKI-3 severity stage, which is the one associated with greater morbidity, related to the severity of complications and, in many cases, to the need for replacement of kidney function.…”