CEC-UW Data Management

Overview

Here’s the summary of the data from the CEC-UW project. It includes more than 1.6 million COVID-19 patients. To view data results, click here.

Specialized computer programs were developed by CEC-UW project programmers to extract EHR-based patient data at each of the 21 participating health systems. Extracted datasets were securely transferred to the CEC-UW Coordinating Center for cleaning, processing, and harmonizing.

Each data transfer included EHR data looking back to February 1, 2020. Each patient was assigned a 64-character unique and private patient ID. The final transfer of data from the health systems covered the period February 1, 2020, to January 31, 2022.

• Across the 21 health systems, there were a total of 105 extracted datasets (21 health systems X 5 datasets = 105) transferred to the CEC-UW Coordinating Center. The five datasets transferred from each health system included:

• • Patient Information Variables
  • Encounter Variables
  • Diagnosis Variables
  • Medication Variables
  • Lab Test Result Variables

The five datasets were used to create two merged datasets for analysis:

• Overall Sample Dataset consists of all patients meeting the following criteria:
  • COVID Index Date (the date that a patient was first identified by a healthcare system as having COVID-19 via ICD-10 diagnosis or lab test (PCR or Antigen test)
  • Identifiable Index Encounter that corresponds to the COVID Index Date
  • Must have either of the following:
    • ICD-10 COVID-19 diagnosis (U07.1 or J12.82)
    • Positive PCR test close in time to the COVID Index Date
    • Positive Antigen test result close in time to the COVID Index Date

• Hospitalized Sample Dataset is a subset of the Overall Sample that consists of patients who were admitted to a hospital (“inpatient encounter”) and who met the following criteria:
  • The inpatient encounter must have been the first inpatient hospitalization and the duration of the inpatient encounter must have been at least 24 hours (or, if < 24 hours, admission to ICU or death during the hospitalization)
  • Must have either of the following:
    • ICD-10 diagnosis of COVID-19 (U07.1 or J12.82) assigned during the initial inpatient encounter
    • Positive PCR test result within a 14-day time window starting 7 days prior to admission through the end of the first week of the initial inpatient encounter
  • The patient must have had prior contact with the health system for purposes of tabulating ICD-10 diagnoses and calculating the Elixhauser Comorbidity Score

Variables for Analysis

Patient Information Variables

• • Health System Information (e.g., health system name, regional location, EHR platform, etc.)
  • Date when patient met eligibility criteria for entry into the cohort
  • For patients who died, date of death
  • Patient sociodemographic characteristics and related variables (e.g., sex, race, ethnicity, insurance status, etc.)
  • Blood type and RH factor
  • COVID-19 vaccination variables
  • Patient BMI scores (or percentiles depending on age) and BMI groups (e.g., underweight, healthy weight, etc.)
  • Location variables based on patient 5-digit ZIP code (e.g., city, state, FIPS code)
  • Rural-Urban Commuting Area (RUCA) variables (RUCA codes based on classification of U.S. census tracts using measures of population density, urbanization, and daily commuting).
  • Social Deprivation Index (SDI) and related variables (the SDI is a composite measure of area level deprivation based on seven demographic characteristics collected in the American Community Survey)
  • Variables related to patient-based Core-based Statistical Areas (CBSA); a CBSA is a geographic area consisting of one or more counties anchored by an urban center of at least 10,000 people plus adjacent counties that are socioeconomically tied to the urban center by commuting

Encounter Variables

• • Timing variables (e.g., encounter start and end dates, duration of the encounter)
  • Zip Code-related variables for the treatment location (e.g., city, state, FIPS code)
  • Encounter type variables (inpatient, outpatient, emergency department, urgent, or other)
  • Admission to ICU during the encounter?
  • Variables related to ventilatory support (e.g., intubation)
  • Serious conditions during the encounter (e.g., septic shock, pneumonia)
  • Patient death during encounter (limited to mortality during hospitalization)
  • Tobacco- and smoking-related variables
  • Vital signs at time of admission (e.g., temperature, pulse, blood pressure, etc.)
  • Other signs and symptoms at time of admission (e.g., chills, muscle aches, shortness of breath, smell impairment, ataxia, seizure, etc.)
  • ICD-10 diagnoses assigned during the encounter
  • CPT and HCPCS Procedure Codes assigned during the encounter
  • Discharge disposition (primarily for hospital admissions)

Diagnosis Variables

Selected ICD-10 diagnoses were determined for each encounter (inpatient, outpatient, etc.) as well as during the following time periods:

• • Past year prior to first diagnosis of COVID-19
  • Past 5 years prior to first diagnosis of COVID-19
  • Ever prior to first diagnosis of COVID-19
  • After first diagnosis of COVID-19

• • • Elixhauser Comorbidity Score
    • Charlson Comorbidity Score
    • COVID-19 Diagnoses
    • Multisystem inflammatory syndrome
    • Non-COVID-19 Coronavirus Diagnoses
    • Diabetes Diagnoses
    • Obesity and Overweight Diagnoses
    • Pulmonary Diagnoses
    • Hypertension Diagnoses
    • Heart Disease Diagnoses
    • Cerebrovascular Disease Diagnoses
    • Kidney Disease Diagnoses
    • Cancers of the Lip, Mouth, Tongue, Parotid Gland, and Salivary Gland
    • Cancers of the Upper Airway Structures
    • Cancers of the Esophagus and Gastrointestinal Tract
    • Head and Neck Cancer Diagnoses
    • Respiratory Cancer Diagnoses
    • Non-Respiratory Thoracic Cavity Cancer Diagnoses
    • Melanoma
    • Mesothelial and Soft Tissue Cancer Diagnoses
    • Breast Cancer
    • Female Genital Cancer Diagnoses
    • Male Genital Cancer Diagnoses
    • Urinary Cancer Diagnoses
    • Eye and CNS Cancer Diagnoses
    • Endocrine Cancer Diagnoses
    • Cancer: Ill-defined, Other Secondary and Unspecified Sites
    • Neuroendocrine Cancer Diagnoses
    • Lymphoma Diagnoses
    • Multiple Myeloma
    • Leukemia Diagnoses
    • Any Kind of Cancer Diagnosis
    • Dementia Diagnoses
    • Arrythmias
    • Trisomy 21
    • Immunosuppression
    • Sarcoidosis
    • Lupus
    • Rheumatoid Arthritis
    • Organ Transplant
    • HIV
    • Crohn’s Disease
    • Ulcerative Colitis
    • Cystic Fibrosis
    • Any Immunocompromising Disease
    • Pregnancy
    • Sickle Cell Disease
    • Septic Shock
    • Depressive Disorder Diagnoses
    • Anxiety Disorder Diagnoses
    • Alcohol Use Disorder Diagnoses
    • Cannabis Use Disorder Diagnoses
    • Any Psychotic Disorder
    • Nicotine Dependence and Tobacco-Related Disorders

Medication Variables

Medication variables are available for medications ordered during the year prior to the date of first COVID-19 diagnosis as well as medications ordered during any encounter (inpatient, outpatient, etc.).

• • COVID-19 Medications: Remdesivir, Convalescent Plasma, Hydroxychloroquine, Chloroquine, Tocilizumab, Vitamin C, Baricitinib, Casirivimab, Imdevimab, Casirivimab + Imdevimab, Bamlanivimab, Etesevimab, Bamlanivimab + Etesevimab, Sarilumab, Molnupiravir, Nirmatrelvir, Ritonavir, Lopinavir, Nirmatrelvir + Ritonavir, Sotrovimab, Cilgavimab, Tixagevimab, Cilgavimab + Tixagevimab, Fluoxetine, Fluvoxamine, Dexamethasone
  • Blood Thinners: Enoxaparin, Heparin, Warfarin
  • Steroid Medications: Hydrocortisone, Methylprednisolone, Prednisolone, Prednisone
  • Smoking Cessation Medications: Nicotine Replacement Therapy (NRT), Varenicline
  • Angiotensin-converting enzyme (ACE) inhibitors: Benazepril, Captopril, Enalapril, Fosinopril, Lisinopril, Moexipril, Perindopril, Quinapril, Ramipril, Trandolapril
  • Angiotensin receptor blockers (ARBs): Azilsartan, Candesartan, Eprosartan, Irbesartan, Losartan, Olmesartan, Telmisartan, Valsartan
  • Bronchodilator Combos: Aclidinium + Formoterol, Formoterol + Glycopyrrolate, Albuterol + Ipratropium, Olodaterol + Tiotropium, Glycopyrrolate + Indacaterol
  • Adrenergic Short Bronchodilators: Albuterol, Epinephrine, Levalbuterol, Isoproterenol, Metaproterenol, Pirbuterol
  • AdrenergicLong Bronchodilators: Olodaterol, Formoterol, Salmeterol, Indacaterol, Terbutaline, Arformoterol, Vilanterol
  • Anticholinergic Bronchodilators: Ipratropium, Tiotropium, Aclidinium, Umeclidinium, Glycopyrrolate, Revefenacin
  • Inhaled Corticosteroids: Fluticasone _Salmeterol, Budesonide +Formoterol, Fluticasone + Umeclidinium + Vilanterol, Formoterol + Mometasone, Fluticasone + Vilanterol, Beclomethasone, Ciclesonide, Flunisolide, Mometasone

Lab Test Result Variables

Lab test variables are available for lab tests ordered during any encounter (inpatient, outpatient, etc.).

• • COVID-19 Antibody Test
  • COVID-19 Antigen Test
  • Albumin
  • Alanine aminotransferase (ALT)
  • Bicarbonate
  • Blood Urea Nitrogen (BUN)
  • Calcium
  • C-Reactive Protein
  • High Sensitivity C-Reactive Protein
  • Creatinine
  • D-Dimer
  • Erythrocyte Sedimentation Rate (ESR)
  • Ferritin
  • Hematocrit
  • Hemoglobin A1c
  • International Normalized Ratio (INR)
  • Lactate Dehydrogenase (LDH)
  • Leukocytes
  • Platelet Count
  • Potassium
  • Procalcitonin
  • Sodium
  • Troponin I
  • Troponin T

Data Availability

The existing Data Transfer and Use Agreements (DTUAs) negotiated with each of the 21 participating health systems precludes the University of Wisconsin (UW) from sharing CEC-UW data with any entity at this time. This was a requirement of the UW IRB prior to health systems transferring their data sets to UW. Starting in 2023, investigators desiring access to CEC-UW data will need to apply for access to the data via a National Cancer Institute (NCI) contractor organization, Information Management Services (IMS; https://www.imsweb.com/).