National Trauma Data Bank - www.ntdsdictionary.org
As we receive questions from both hospitals and software developers, we will try to update this section. If you have any questions, please feel free to contact the NTDB. We will try to update this section as often as possible.
- Will the new dataset become the NTDB and what is the difference between the two?
- When will facilities have to be completely NTDS compliant?
- Is there discussion about adding the AIS 6 digit code to the national dataset?
- If a patient dies in the ED, the registrar would enter "Expired" under ED Disposition. For Hospital Disposition, would the registrar enter "Not Applicable" because they were never admitted to the hospital or does every patient get a hospital discharge disposition?
- Registrars are questioning whether IV's, wound dressings, foleys, and
other non-major procedures need to be entered in the Hospital Procedures section?
- Regarding the ED DEATH variable: Does the 15 minute resuscitation criteria start from the time the patient arrived at the reporting hospital? Or should registrars also take into account any prehospital or referring facility resuscitation time?
- I have heard that the NTDS data dictionary and the new national pre-hospital data dictionary (NEMSIS) are very similar. Please explain how this similarity helps me as an abstractor?
- Based on the NTDS, how is total ICU LOS calculated when a patient stays a partial day?
- The NTDS Data Dictionary definition for Operative and/or Essential Procedures is vague and seems inadeaquate?
Think of the NTDB as a bank….that conducts deposits and withdraws. Data are deposited in the NTDB and reports are generated from the data and provided to researchers. The new data dictionary is simply that, a new data standard defining the structure and format of the data deposited within the NTDB. It represents variables characterizing trauma events that are important to aggregate at the national level.
Starting with NTDB's 2008 Call for Data, participating facilities will need to have a translation program in place to map their current data to the new data standard. Facilities should then work with their vendors to have their registry software updated with the new XML by 2009.
You are correct that the new national database relies upon ICD-9-CM codes to calculate AIS and ISS. You are also right that AAAM AIS and ICD-9-CM AIS are not likely to always agree. However, the new national database relies upon ICD-9-CM for AIS because many hospitals do not use the AAAM system. Although ICD-9-CM AIS systems have a number of shortcomings, ICD-9-CM codes are available at all hospitals and ensures that AIS and ISS codes are consistent and can be compared across hospitals if calculated in the same way. Thus, the National database utilizes a ICD-9-CM based AIS. We have, however, added optional varibales to the NTDS that capture the 6-digit AIS code. Thus, hospitals that do calculate AIS will be able to report that data to the NTDB.
If the patient dies in the ED, two variables are completed: "ED Discharge Disposition" and "ED Death". Enter "Expired" under ED Disposition. Under Hospital Disposition, you would list "Not Applicable".
We knew that there would be some variation in how folks interpreted "operative and/or essential" (O/E) procedures. The other option was to list every possible O/E procedure and try to keep that list current...which would be a daunting task. The term "operative" should be somewhat very time consuming. When attempting to define O/E, I think you can safely exclude IV's, wound dressings, foleys, and other non-major procedures unless the abstractor, based upon the constellation of injuries, considered that procedure to be vital to the stabilization, treatment, survival of the patient. These specific circumstances would be rare. Thus, I suggest excluding these items.
This field was meant to refer to activities occurring in the reporting hospital.
The NTDS data dictionary contains a number of variables that are to be obtained from the pre-hospital record. The programming language that defines variables in NTDS and NEMSIS were designed to exactly match. This was done so that "next generation" trauma registry software and pre-hospital software could exchange data. Thus, in the future, trauma registrars may open a new record on a patient, only to find that all of the pre-hospital variables have "auto-populated" the record. Similarly, when an abstractor finishes a trauma registry record, ED and hospital outcome data could "back-populate" the EMS record, allowing EMS to evaluate QA topics for transported patients. For a detailed description of the overlap between the two data bases, click on this link. In the linked document, green variables demonstrate direct XML and XSD compatability when moving from NEMSIS to NTDS. Yellow indicates variables in NEMSIS that could “inform” trauma registry abstractors, but no direct data transfer is possible (i.e., there is non-matching XML and XSDs).
The spirit of this question is to ensure that we not only document time spent in the ICU, but that we also document all who were transferred to the ICU even if for a very short period of time. Thus, if a patient is transferred into an ICU and then is transferred out in less than one day…the total ICU LOS should be calculated as one day. If the patient stays more than one day, or experiences multiple admissions to an ICU, the total time spent in the ICU should be calculated (in hours) and rounded to the next full day increment. Thus, if a patient is admitted to the ICU on two different occasions for a total of 31 hours, the total ICU LOS would be recorded as 2 days.
This question is one that content experts developing the NTDS data dictionary really wrestled with. In summary, the content experts determined that it would be very difficult to maintain a lengthy list of procedures performed in an Operating Suite, Emergency Department, or Intensive Care Unit that would be considered important to the diagnoses, stabilization, or treatment of the patients with specific injuries. New procedures are developed continuously and procedures important to the care of trauma differ somewhat based upon the type and severity of trauma....the list would be very, very long!
The value of a standardized list of procedures (in the past) was that researchers could use the NTDB to determine if patients "expected" to receive a particular procedure...actually received it. Nevertheless, a major bias with the NTDB was always that researchers did not know if a patient didn’t receive a needed procedure...or if it was just not recorded. Conducting this type of research using a retrospective dataset is almost impossible.
The approach that was taken allows hospital-based registries some flexibility. Most trauma registries in designated trauma centers are collecting data on a set of procedures "they" consider important or rely on administrative procedure codes to document procedures the patient receives. The approach adopted by the NTDB requires that abstractors have some idea of what procedures were considered "essential to the diagnoses, stabilization, or treatment of the patient's specific injuries" or work with clinicians in there institutions to develop a list to capture the appropriate procedures.
It is understandable that the current definition for this variable appears somewhat inadequate. It was meant to be flexible. We would admonish everyone to work with thier facility to determine what list they would like use (or method) for abstracting this information.