These data come from a large, multi-hospital study of adult patients with critical, acute conditions such as acute respiratory failure and multiple organ system failure. The primary outcome is time until death. A variety of potentially useful predictors, both demographic and lab values, are available. The primary goal of the study is to develop an algorithm for predicting survival in these patients.
Death: Death indicatorTime: Time on study (days)Age: Age in years at beginning of studyNCom: Number of comorbiditiesEdu: Education level (years of schooling completed)Income: (under $11k, $11-$25k, $25-$50k, >$50k)Race: (white, black, hispanic, asian, other)HDay: Number of days spent in the hospital prior to being admitted into the study. For many patients, this is 1 - they were entered into the study at the beginning of their stay in the hospital. Other patients, however, stayed in the hospital for days/week/months prior to their entry in the study.MAP: Mean arterial pressure (mm Hg)WBC: White blood cell count (thousands/mm3)Pulse: Heart rate (beats/minute)Resp: Breathing rate (breaths/minute)Temp: Temperature (degrees Celsius)PAFI: Ratio between the partial pressure of oxygen in arterial blood (PaO2) and the fraction of inspired oxygen (FiO2), PaO2/FiO2.Alb: Serum albumin (g/dL)Bili: Serum bilirubin (mg/dL)Crea: Serum creatine (mg/dL)Sod: Serum sodiumPH: Arterial PHGlucose: Serum glucose (mg/dL)BUN: Blood urea nitrogen (mg/dL)Urine: Urine output (cc/day)I obtained the data from http://biostat.mc.vanderbilt.edu; the original publication is:
Knaus WA et al. (1995) The SUPPORT prognostic model: Objective estimates of survival for seriously ill hospitalized adults. Annals of Internal Medicine, 122:191-203.