readmission.RdClinical care data from 130 U.S. hospitals in years 1999-2008. Each row describes an "encounter" with a patient with diabetes, including variables on demographics, medications, patient history, diagnostics, payment, and readmission.
readmissionA data frame with 71,515 rows and 12 columns:
Whether the patient was readmitted within the 30 days
following discharge. A factor with levels "Yes" and "No".
Reported race of the patient. Source data does not document
data collection strategy. A factor with levels "African American",
"Asian", "Caucasian", "Hispanic", "Other", and "Unknown".
Reported sex of the patient. Source data does not document
data collection strategy. A factor with levels "Female" and "Male".
Age range for the patient, binned in 10-year intervals. A factor
with levels "[0-10)", "[10-20)", "[20-30)", "[30-40)", "[40-50)",
"[50-60)", "[60-70)", "[70-80)", "[80-90)", and "[90-100)".
Whether the patient was referred from a physician,
admitted via the ER, or arrived via some other source. A factor with
levels "Emergency", "Other", and "Referral".
Results from an A1C test, estimating the
patient's average blood sugar over the past 2-3 months. Higher estimated
average blood glucose levels are linked to diabetes complications. A factor
with levels "Normal", "High", and "Very High", and many missing values.
The health insurance provider (or lack thereof,
via "Self-Pay") for the patient. A factor with levels
"Medicaid", "Medicare", "Private", and "Self-Pay", and many missing
values.
Number of days in the hospital between admission and discharge.
Number of emergency, inpatient, and outpatient visits in the year preceding the encounter.
"Number of diagnoses entered to the system" during the encounter.
"Number of procedures (other than lab tests) performed" during the encounter.
"Number of distinct generic names administered" during the encounter.
Original source data from the following paper (CC BY 3.0):
Strack, B., DeShazo, J. P., Gennings, C., Olmo, J. L., Ventura, S., Cios, K. J., & Clore, J. N. 2014. Impact of HbA1c measurement on hospital readmission rates: analysis of 70,000 clinical database patient records. BioMed research international, 781670. doi:10.1155/2014/781670.
Shared freely through the UCI Machine Learning Repository (CC BY 4.0):
Clore, J., Cios, K., DeShazo, J. P., and Strack, B. 2014. Diabetes 130-US hospitals for years 1999-2008. UCI Machine Learning Repository. doi:10.24432/C5230J.
Downloaded from resources shared by the Fairlearn team (MIT):
Weerts, H., Dudík M., Edgar, R., Jalali, A., Lutz, R., & Madaio, M. 2023. Fairlearn: Assessing and Improving Fairness of AI Systems. Journal of Machine Learning Research, 24(257):1-8.
str(readmission)
#> tibble [71,515 × 12] (S3: tbl_df/tbl/data.frame)
#> $ readmitted : Factor w/ 2 levels "Yes","No": 1 2 1 2 2 2 1 2 2 2 ...
#> $ race : Factor w/ 6 levels "African American",..: 1 3 3 3 3 3 1 3 3 3 ...
#> $ sex : Factor w/ 2 levels "Female","Male": 2 1 1 1 1 2 1 1 2 1 ...
#> $ age : Factor w/ 10 levels "[0-10)","[10-20)",..: 7 6 8 9 8 6 8 3 7 9 ...
#> $ admission_source : Factor w/ 3 levels "Emergency","Other",..: 3 1 3 3 3 1 3 1 2 3 ...
#> $ blood_glucose : Factor w/ 3 levels "Normal","High",..: NA 1 NA NA NA 3 NA NA NA NA ...
#> $ insurer : Factor w/ 4 levels "Medicaid","Medicare",..: NA 3 2 3 NA NA 3 NA NA 2 ...
#> $ duration : num [1:71515] 7 4 5 5 4 2 3 1 12 1 ...
#> $ n_previous_visits: num [1:71515] 2 0 2 0 0 0 0 7 0 0 ...
#> $ n_diagnoses : num [1:71515] 4 9 9 9 5 2 9 9 9 4 ...
#> $ n_procedures : num [1:71515] 0 0 0 3 1 3 3 0 2 2 ...
#> $ n_medications : num [1:71515] 16 15 14 26 15 25 22 10 17 6 ...
head(readmission)
#> # A tibble: 6 × 12
#> readmitted race sex age admission_source blood_glucose insurer duration
#> <fct> <fct> <fct> <fct> <fct> <fct> <fct> <dbl>
#> 1 Yes Africa… Male [60-… Referral NA NA 7
#> 2 No Caucas… Fema… [50-… Emergency Normal Private 4
#> 3 Yes Caucas… Fema… [70-… Referral NA Medica… 5
#> 4 No Caucas… Fema… [80-… Referral NA Private 5
#> 5 No Caucas… Fema… [70-… Referral NA NA 4
#> 6 No Caucas… Male [50-… Emergency Very High NA 2
#> # ℹ 4 more variables: n_previous_visits <dbl>, n_diagnoses <dbl>,
#> # n_procedures <dbl>, n_medications <dbl>