Skip to content

Hip fractures dataset to learn data wrangling and modeling

License

Notifications You must be signed in to change notification settings

rstats-tartu/hip-fractures

Repository files navigation

Hip fracture data

Estonian hip fracture data from Jan 2009-Oct 2017 (modified) Structure of the dataset and descriptions of the variables

ABBREVIATIONS:

PAC – post-acute care LOS – length of hospital stay (measured in days) n/a – not available or missing value

SUMMARY FILE (COLUMNS)

Patients' characteristics:

  • id -- anonymised patient's identification number
  • sex -- f - female; m - male
  • age -- in years
  • county -- county of residence
  • Comorbidities. 19 disease categories, all binary variables (1 – disease, 0 – no disease):
    • comorbidity -- Charlson comorbidity score
    • myocardial_infarction,
    • congestive_heart_failure,
    • peripheral_vascular_disease,
    • cerebrovascular_disease,
    • dementia,
    • chronic_pulmonary_disease,
    • rheumatic_disease,
    • peptic_ulcer_disease,
    • diabetes,
    • hemi-paraplegia,
    • renal_disease,
    • any_malignancy,
    • liver_disease,
    • metastatic_solid_tumor,
    • aids-hiv,
    • alcohol_abuse,
    • obesity,
    • psychoses,
    • depression
  • Fracture
  • Time
  • date_of_fracture -- actually it is date of hospitalisation, but mostly they are the same
  • year -- year of fracture
  • Type
    • fracture_type -- 3 values: femoral_neck, pertrochanteric, subtrochanteric
  • Acute care (1st step care). Fracture treatment some variable include missing values (N/A) for operated patients
    • operation_date
    • days_to_operation – time from acute hospitalisation to operation in days
    • operation_weekday – 1 – Monday, 2 Tuesday …
    • management_method –values: conservative, hemiarthroplasty, total hip arthroplasty, osteosynthesis
  • Length of stay and physical therapy received
    • acute_LOS – integer variable, acute length of stay in days
    • acute_therapy – scale variable, physical therapy received during acute care
  • Post-acute care (2nd step care). Length of hospital stay
    • post-acute_LOS – integer variable, total post-acute length of stay in days (includes all hospitalisations)
    • post-acute_therapy – scale variable, total post-acute physical therapy received in hours (includes all in- and outpatient settings)
    • PAC_type – post-acute care type: patients were subgrouped by received post-acute in- and outpatient care
  • Survival. Variables presented as time (time period, m – month [in days or in months]) and status (1 – event of death occurred; 0 – alive at that time point). - time_3m - status_3m - time_12m - status_12m

POST-ACUTE DATA ONLY (ROWS)

  • Patients' characteristics
    • id -- anonymised patient's identification number
    • invoice_number – post-acute invoices were numbered consecutively by date
    • invoice_start – invoice start date
    • invoice_end – invoice end date
    • post-acute_type – type of setting (i simplified this part for the course, since defining these types would take much time)
    • LOS – length of stay in days in this setting, zeros are for outpatient settings
    • therapy – total physical therapy received in that setting

About

Hip fractures dataset to learn data wrangling and modeling

Resources

License

Stars

Watchers

Forks

Packages

No packages published