This README.txt file describes the dataset that accompanies: He J., Hu Y., Zhang X., Wu L., Waitman LR., Liu M. Multi-perspective predictive modeling for acute kidney injury in general hospital populations using electronic medical records ------------------- GENERAL INFORMATION ------------------- 1. Title of Dataset: AKI_Prediction_Scrubbed_EMR 2. Author Information Corresponding Author Contact Information Name: Mei Liu Institute: University of Kansas Medical Center Address: 3901 Rainbow Boulevard, Kansas City, KS 66160, USA Email: meiliu@kumc.edu First Author Contact Information Name: Yong Hu Institute: Big Data Decision Institute, Jinan University Address: Tianhe, Guangzhou, China Email: henryhu200211@163.com 3. Date of data collection: 2007/11 to 2016/12 4. Geographic location of data collection: Kansas City, KS 66160 5. Information about funding sources that supported the collection of the data: The dataset used for analysis described in the study was obtained from the University of Kansas Medical Center's HERON clinical data repository which is supported by institutional funding and by the KUMC CTSA grant UL1TR002366 from NCRR/NIH. -------------------- DATA & FILE OVERVIEW -------------------- 1. File List A. Filename: Perspective_scrb_1.csv Description: All samples for perspective #1 AKI prediction Number of variables: 233 Number of cases/rows: 76957 B. Filename: Perspective_scrb_2.csv Description: All samples for perspective #2 AKI prediction Number of variables: 303 Number of cases/rows: 76957 C. Filename: Perspective_scrb_3.csv Description: All samples for perspective #3 AKI prediction with prediction window between admission to 1-day after admission day. Number of variables: 303 Number of cases/rows: 76957 D. Filename: Perspective_scrb_4.csv Description: All samples for perspective #4 AKI prediction with prediciton point on 1-day after admission day. Number of variables: 283 Number of cases/rows: 72846 2. Variable List Data dictionary is in AKI_dataset_attributes_map.xlsx. For example, 'X0' in data files above corresponds to 'Attribute 1' in the dictionary map and 'X1917' is the class label. -------------------- DATA USE INSTRUCTION -------------------- Our de-identified dataset contain total 1287 variables. However, to anonymize the dataset for public sharing, we further scrubbed the de-identified dataset by removing variables. For access to the full de-identified dataset used in our study, please contact the corresponding author Mei Liu at meiliu@kumc.edu. Our institutional IRB requires a signed data use agreement before data release.