First and foremost, data validation is an educational tool that sharpens the skills of the Trauma Registry Professional and in turn, develops better data for the trauma centers to use to improve patient care. Data validation highlights areas of weakness and motivates the registrar to pursue additional training, ensuring the accuracy of data points
There are many ways to do data validation on the trauma registry, so there are multiple tools that one can find to fit their trauma center needs. The main goal is to train the Trauma Registry Professionals to get the most accurate data from the EHR to the registry.
The Guidelines:
The National Trauma Data Standard (NTDS) serves as the essential data dictionary for all trauma registrars. It details all required data fields, providing definitions, element values, and additional information to ensure correct data entry. The NTDS includes a data source hierarchy guide, directing registrars to the appropriate documents for data retrieval. It also outlines associated edit checks, specifying that Level 1 and 2 edit checks must be corrected before data can be uploaded to the National Trauma Data Bank (NTDB). Statewide trauma data dictionaries, for states that track additional data, adhere to the same format as the NTDS.
Hospital-based data dictionaries are designed to indicate where to obtain data from the EHR. They should include an additional column specifying the exact location in the EHR to pull the data. This will assist new registrars in accurately entering data into the trauma registry.
According to the American College of Surgeons (ACS), up to 10% of the total charts per month must be validated. There are two schools in data validation of the trauma registry:
100% chart audit:
This is used for the new registrar. This will assist in finding out what areas they are weak in and in turn get them educated on those areas.
Some trauma centers still do 100% chart audits on all their records
20-25 data points:
Once the new registrar has shown proficiency in the 100% validations, you can then choose 20-25 data fields to perform the validation. You can choose what data fields to track, but it is important to cover the ones that help show the Probability of Survival.
Age
Mechanism
Blunt vs Penetrating
Revised Trauma Score
Injury Severity Score (ISS)
Based on 100%, you can add a percentage to take off for any missing or incorrect data that is found.
Inter-rater Reliability (IRR) is the percentage of accuracy the Trauma Registry Professional has obtained in the validation.
If you select 20-25 data points and your registrars consistently achieves 100% accuracy, you can replace one of those data points with a new one to maintain engagement. Trauma Registry Professionals are aware of the data points being monitored, which helps ensure their integrity.
One method to do data validation is that you have the data analyst and Trauma Registry Professional and PI RN, or other trauma registrar in the same room. The data analyst gives a registry number to the Trauma Registry Professional who in turn tells the PI RN the medical record number. After the PI RN opens the EHR, the analyst then goes through each data point that is being tracked. The PI RN finds it in the medical record, then answers the question and the Trauma registry Professional has to either acknowledge it as correct, or say it was missed in the registry. Then they can discuss where the registrar found the data.
In data validation, you aim to identify patterns of missed or incorrect data. If a pattern is detected, you can guide the registrar to relevant courses to improve their understanding. Options include trauma registrar courses, ICD-10 courses, and AIS courses. Sharing validation scores on a shared drive allows the Trauma Program Manager and Trauma Medical Director to quickly review the inter-rater reliability (IRR) of the trauma registrars, ensuring high-quality reports from the registry.
In conclusion there are many ways to validate a record for the Trauma Registry Professional. The main point is to be educational because no one is perfect. It should be a two-way street where the reviewer identifies the missing/incorrect data, presents it to the registrar and the registrar can show the reviewer where they got the data. All of this is done to help the registrar abstract and enter high quality data to help the hospitals improve the treatment of the injured patient.
