NTDB National Trauma Data Bank (TQP Trauma Quality Program) - onetomapanalytics/Meta_Data GitHub Wiki

NTDB - National Trauma Data Bank (TQP - Trauma Quality Program)

General description

  1. Database primary purpose - To be a repository of trauma-related data voluntarily reported by participating trauma centers.
  2. Overall data type - Health outcomes
  3. Dataset type - Cross-sectional
  4. Data source - Registry
  5. Data level - Patient level
  6. Geographic location of the data collection sites - United States
  7. Sponsor, manager, or home institution - American College of Surgeons
  8. Date range - 2011 - 2019
  9. Geolocation data - Geographic region for the hospital (midwest, northeast, south, west)
  10. Dates - Year
  11. Financial variables - Patient’s primary method of payment
  12. Clinical areas of interest - Trauma. Specifically, ICD-9-CM discharge diagnosis 800.00–959.9.
  13. Variables that are uniquely present in this dataset - NTDB is one of the primary datasets in the United States pertaining exclusively to Trauma.
  14. Database caveats and limitations - (1) The data quality of NTDB is dependent on how well the National Trauma Data Standard (NTDS) is implemented by the individual hospital. (2) NTDB is not a population-based dataset, so it is subject to the limitations of all “convenience samples.” (3) Since hospitals may have differing criteria for inclusions, NTDB is subject of selection bias (i.e., an apparent difference between two groups that is actually caused by different inclusion criteria) and information bias (i.e., an apparent difference between two groups that is actually caused by a difference in the data available to compare them). (4) The proportion of missing data varies across data elements. (5) NTDB is not nationally representative since its data are voluntarily submitted. However, all level I/II trauma centers now provide data to the NTDB, so the data can be considered nationally representative of those facilities. (6) The NTDB is an incident-based record; each incident is recorded independently of repeated injuries. (7) NTDB does not include data on costs, laboratory values, and long-term outcomes, including readmissions and functional outcomes.

Applicable methods

  1. Association analysis such as multivariable logistic regression (1, 2, 3), linear regression (4)
  2. Cochran-Armitage trend analysis (5)
  3. Machine learning (6, 7)
  4. Propensity score (8, 9)
  5. Univariate analysis (2)

High-impact designs

  • Evaluate blunt mechanism of injury (10)

  • Investigate predictors of mortality (1)

  • Assess undertriage (UT) rate and characteristics of UT patients (2)

  • Determine incidence, risk factors, and outcomes (11)

  • Association of trauma center volume and outcomes in emergency surgery (3)

  • Recent trends in the management of injuries (5)

  • Assess facility disparities (12)

  • Practical guide to surgical datasets (13)

Data dictionary

To access the data dictionary, click here

Variable categories

  1. Patient demographics (e.g., age, sex, race, ethnicity)
  2. Pre-hospital information (e.g., EMS response, EMS vital sign)
  3. Emergency department (e.g., ED vital sign, elapsed time in ED, death)
  4. Hospital procedure information (e.g., procedure codes, time to procedure)
  5. Diagnosis (see "injury" below, includes complications during patient treatment)
  6. Injury severity information (e.g., Injury Severity Score, diagnosis code, nature of injury, external cause, body region, protective devices)
  7. Hospital discharge records (e.g., year, total length of stay, ICU days, ventilator days)
  8. Financial information (primary payment method)
  9. Pre-existing conditions (e.g., comorbidities)
  10. Facility data (e.g., tax status, teaching hospital, ACS verification, number/type of beds, number of surgeons)

Linkage to other datasets

  • Linkages of NTBD data to any other data are not allowed.