17
views
0
recommends
+1 Recommend
0 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      Prediction of Acute Kidney Injury With a Machine Learning Algorithm Using Electronic Health Record Data

      research-article

      Read this article at

      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          Background:

          A major problem in treating acute kidney injury (AKI) is that clinical criteria for recognition are markers of established kidney damage or impaired function; treatment before such damage manifests is desirable. Clinicians could intervene during what may be a crucial stage for preventing permanent kidney injury if patients with incipient AKI and those at high risk of developing AKI could be identified.

          Objective:

          In this study, we evaluate a machine learning algorithm for early detection and prediction of AKI.

          Design:

          We used a machine learning technique, boosted ensembles of decision trees, to train an AKI prediction tool on retrospective data taken from more than 300 000 inpatient encounters.

          Setting:

          Data were collected from inpatient wards at Stanford Medical Center and intensive care unit patients at Beth Israel Deaconess Medical Center.

          Patients:

          Patients older than the age of 18 whose hospital stays lasted between 5 and 1000 hours and who had at least one documented measurement of heart rate, respiratory rate, temperature, serum creatinine (SCr), and Glasgow Coma Scale (GCS).

          Measurements:

          We tested the algorithm’s ability to detect AKI at onset and to predict AKI 12, 24, 48, and 72 hours before onset.

          Methods:

          We tested AKI detection and prediction using the National Health Service (NHS) England AKI Algorithm as a gold standard. We additionally tested the algorithm’s ability to detect AKI as defined by the Kidney Disease: Improving Global Outcomes (KDIGO) guidelines. We compared the algorithm’s 3-fold cross-validation performance to the Sequential Organ Failure Assessment (SOFA) score for AKI identification in terms of area under the receiver operating characteristic (AUROC).

          Results:

          The algorithm demonstrated high AUROC for detecting and predicting NHS-defined AKI at all tested time points. The algorithm achieves AUROC of 0.872 (95% confidence interval [CI], 0.867-0.878) for AKI detection at time of onset. For prediction 12 hours before onset, the algorithm achieves an AUROC of 0.800 (95% CI, 0.792-0.809). For 24-hour predictions, the algorithm achieves AUROC of 0.795 (95% CI, 0.785-0.804). For 48-hour and 72-hour predictions, the algorithm achieves AUROC values of 0.761 (95% CI, 0.753-0.768) and 0.728 (95% CI, 0.719-0.737), respectively.

          Limitations:

          Because of the retrospective nature of this study, we cannot draw any conclusions about the impact the algorithm’s predictions will have on patient outcomes in a clinical setting.

          Conclusions:

          The results of these experiments suggest that a machine learning–based AKI prediction tool may offer important prognostic capabilities for determining which patients are likely to suffer AKI, potentially allowing clinicians to intervene before kidney damage manifests.

          Abrégé

          Contexte:

          Une des principales difficultés liées au traitement de l’insuffisance rénale aiguë (IRA) est le fait que les critères cliniques diagnostiques sont des marqueurs d’une lésion ou d’une dysfonction rénale déjà établie. Il est souhaitable d’intervenir avant une telle issue. En dépistant les patients à risque d’IRA ou atteints d’IRA débutante, les cliniciens seraient en mesure d’intervenir précocement et ainsi prévenir les lésions rénales permanentes.

          Objectif de l’étude:

          L’étude visait à évaluer un algorithme d’apprentissage automatique destiné à la prédiction des cas d’IRA et à sa détection précoce.

          Type d’étude:

          Nous avons employé une technique d’apprentissage automatique, soit des ensembles d’arbres décisionnels amplifiés, pour entrainer un outil de prédiction de l’IRA à partir de données rétrospectives provenant de plus de 300 000 consultations auprès de patients hospitalisés.

          Cadre de l’étude:

          Les données ont été colligées à partir des dossiers des unités d’hospitalisation du centre médical de l’université Stanford et de l’unité des soins intensifs du centre médical Beth Israel Deaconess.

          Participants:

          Ont été inclus dans l’étude tous les patients adultes dont l’hospitalisation avait duré de 5 à 1 000 heures et pour lesquels on disposait d’au moins une mesure parmi les suivantes : pouls, rythme respiratoire, température corporelle, taux de créatinine sérique (SCr) et score de Glasgow.

          Mesures:

          Nous avons testé l’efficacité de l’algorithme à détecter l’IRA dès son apparition, et à la prédire 12, 24, 48 et 72 heures avant qu’elle ne se manifeste.

          Méthodologie:

          L’algorithme du NHS England a servi de référence pour tester l’efficacité de notre algorithme de prédiction et de détection de l’IRA. Nous avons également testé l’efficacité de notre algorithme à détecter l’IRA telle que définie par les Recommandations de Bonnes Pratiques Cliniques du KDIGO ( Kidney Disease: Improving Global Outcomes ). Nous avons utilisé la surface sous la courbe ROC ( Receiver Operating Characteristic) pour comparer le score SOFA à l’efficacité de validation croisée tripartite de notre algorithme.

          Résultats:

          L’algorithme a démontré une SSROC (surface sous la courbe ROC) élevée pour la détection et la prédiction de l’IRA (telle que définie par le NHS) pour tous les moments testés. En détection de la maladie à son apparition, l’algorithme a obtenu une SSROC de 0,872 (IC 95 % : 0,867-0,878). En prédiction, l’algorithme a obtenu une SSROC de 0,800 (IC 95 % : 0,792-0,809) à 12 heures, de 0,795 à 24 heures (IC 95 % : 0,785-0,804), de 0,761 (IC 95 % : 0,753-0,768) à 48 heures et de 0,728 (IC 95 % : 0,719-0,737) à 72 heures avant l’apparition des premiers symptômes.

          Limites de l’étude:

          La nature rétrospective de l’étude ne nous permet pas de tirer de conclusions sur les conséquences qu’auront les prédictions de l’algorithme sur les résultats cliniques des patients.

          Conclusion:

          Les résultats de nos essais laissent supposer qu’un outil de prédiction de l’IRA fondé sur l’apprentissage automatique pourrait offrir d’importantes fonctions pronostiques pour détecter les patients susceptibles de développer une IRA en vue d’une intervention précoce.

          Related collections

          Most cited references17

          • Record: found
          • Abstract: found
          • Article: not found

          Acute kidney injury, mortality, length of stay, and costs in hospitalized patients.

          The marginal effects of acute kidney injury on in-hospital mortality, length of stay (LOS), and costs have not been well described. A consecutive sample of 19,982 adults who were admitted to an urban academic medical center, including 9210 who had two or more serum creatinine (SCr) determinations, was evaluated. The presence and degree of acute kidney injury were assessed using absolute and relative increases from baseline to peak SCr concentration during hospitalization. Large increases in SCr concentration were relatively rare (e.g., >or=2.0 mg/dl in 105 [1%] patients), whereas more modest increases in SCr were common (e.g., >or=0.5 mg/dl in 1237 [13%] patients). Modest changes in SCr were significantly associated with mortality, LOS, and costs, even after adjustment for age, gender, admission International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis, severity of illness (diagnosis-related group weight), and chronic kidney disease. For example, an increase in SCr >or=0.5 mg/dl was associated with a 6.5-fold (95% confidence interval 5.0 to 8.5) increase in the odds of death, a 3.5-d increase in LOS, and nearly 7500 dollars in excess hospital costs. Acute kidney injury is associated with significantly increased mortality, LOS, and costs across a broad spectrum of conditions. Moreover, outcomes are related directly to the severity of acute kidney injury, whether characterized by nominal or percentage changes in serum creatinine.
            Bookmark
            • Record: found
            • Abstract: found
            • Article: not found

            The RIFLE criteria and mortality in acute kidney injury: A systematic review.

            In 2004, the Acute Dialysis Quality Initiative workgroup proposed a multilevel classification system for acute kidney injury (AKI) identified by the acronym RIFLE (Risk, Injury, Failure, Loss of kidney function, and End-stage kidney disease). Several studies have been published aiming to validate and apply it in clinical practice, verifying whether outcome progressively worsened with the severity of AKI. A literature search from August 2004 to June 2007 was conducted: 24 studies in which the RIFLE classification was used to define AKI were identified. In 13 studies, patient-level data on mortality were available for Risk, Injury, and Failure patients, as well as those without AKI (non-AKI). Death was reported at ICU discharge, hospital discharge, 28, 30, 60, and 90 days. The pooled estimate of relative risk (RR) for mortality for patients with R, I, or F levels compared with non-AKI patients were analyzed. Over 71 000 patients were included in the analysis of published reports. With respect to non-AKI, there appeared to be a stepwise increase in RR for death going from Risk (RR=2.40) to Injury (RR=4.15) to Failure (6.37, P<0.0001 for all). There was significant intertrial heterogeneity as expected with the varying patient populations studied. The RIFLE classification is a simple, readily available clinical tool to classify AKI in different populations. It seems to be a good outcome predictor, with a progressive increase in mortality with worsening RIFLE class. It also suggests that even mild degrees of kidney dysfunction may have a negative impact on outcome. Further refinement of RIFLE nomenclature and classification is ongoing.
              Bookmark
              • Record: found
              • Abstract: not found
              • Article: not found

              Kidney Disease:Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group KDIGO clinical practice guideline for acute kidney injury

                Bookmark

                Author and article information

                Journal
                Can J Kidney Health Dis
                Can J Kidney Health Dis
                CJK
                spcjk
                Canadian Journal of Kidney Health and Disease
                SAGE Publications (Sage CA: Los Angeles, CA )
                2054-3581
                08 June 2018
                2018
                : 5
                : 2054358118776326
                Affiliations
                [1 ]Dascena, Inc, Hayward, CA, USA
                [2 ]Department of Emergency Medicine, University of California, San Francisco, USA
                [3 ]Kaiser Permanente South San Francisco Medical Center, CA, USA
                [4 ]Department of Medicine, Stanford University School of Medicine, CA, USA
                Author notes
                [*]Anna Lynn-Palevsky, Dascena, Inc, 22710 Foothill Boulevard, Suite #2, Hayward, CA 94541, USA. Email: anna@ 123456dascena.com
                Article
                10.1177_2054358118776326
                10.1177/2054358118776326
                6080076
                30094049
                7f62268c-3e79-409f-9cc3-e17b1bbba13c
                © The Author(s) 2018

                This article is distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 License ( http://www.creativecommons.org/licenses/by-nc/4.0/) which permits non-commercial use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages ( https://us.sagepub.com/en-us/nam/open-access-at-sage).

                History
                : 11 October 2017
                : 28 March 2018
                Categories
                Original Research Article
                Custom metadata
                January-December 2018

                acute kidney injury,machine learning
                acute kidney injury, machine learning

                Comments

                Comment on this article