How to use registry reports to increase trauma data accuracy


The whole purpose of a trauma registry is to use your data to improve patient care and make better program management decisions. But as the Orange Book points out, the only way to transform your registry into useful information is to use registry reports and validation to maintain quality data. Unfortunately, report writing and data validation often take a back seat to other priorities. When you are struggling to maintain a concurrent trauma registry, does it make sense to set aside time for running reports and doing quality checks?

This question was answered for me a few years ago when I was the registry manager at a large Level I trauma center. As part of my job I regularly ran several registry reports, including monthly trending reports for MOI, admit unit, complication counts, physician admit service, etc.

Like most registry teams, we often dealt with chart backlogs, so I decided to cut back on reporting and help with abstraction. The plan was that devoting more resources on the front end would help us get caught up with our data.

However, within just a few weeks I realized things were not working out as expected. While I was saving time on running reports, I was spending more time than ever cleaning up data errors on the back end. I saw in hindsight that the effort I had been putting into reporting was actually paying off several-fold in better chart abstracting, higher quality data and a more useful trauma registry.

The bottom line is that data reporting is an absolutely critical part of data validation. Report writing—used in combination with trending analysis, consistency checks and logic—lets you spot the front-end problems that lead to bad data. The more frequently you look at your data through registry reports, the more accurate your trauma registry data will be.

Most trauma centers do not devote enough resources to data reporting. The good news is that even a basic reporting strategy can lead to significant improvements in data accuracy. Here are three suggestions for creating an ongoing reporting program that improves data quality while providing valuable information to trauma leaders.

1. Create monthly trending reports on key PI data points
The Orange Book requires ACS-verified trauma centers to use registry data to support performance improvement, so PI is an ideal place to begin with data reporting. Run monthly registry reports on the data points that have the biggest impact on quality, outcomes and safety in your facility. For example:

Mechanism of injury. Clinical leaders need mechanism of injury data to analyze patient cohorts for performance improvement. When reviewing your monthly injury mechanism report, validate the underlying data by making sure mechanism types match E-codes.

Admitting service. Monthly admission reports help program leaders make sure patients are receiving the right specialty care. Double-check the charts for all multisystem injury patients who were admitted from a non-surgical service.

Ventilator days. Like most registry data points, vent days must align logically with other data fields. If your monthly report says a patient was intubated in the ED and spent time in the ICU but ventilator days were “0”, the registry data may be flawed. Check the chart to confirm that the intubation did not continue through the ICU visit.

2. Use reports to check data submitted to national registries
If your hospital participates in the Trauma Quality Improvement Program (TQIP) or submits data to the National Trauma Data Bank (NTDB), regular internal reporting can help you ensure the quality of your data submissions. Develop ad hoc reports for:

Key fields and exclusion triggers. Begin by writing a custom report of required data elements in the NTDB dictionary. Many of these elements will overlap with your key PI data, like mechanism of injury and E-code. Other data fields, such as ED discharge disposition, are important because they can trigger exclusion from TQIP.

Comorbid conditions and complications. Your ad hoc reports should also include comorbidity and complication data fields, since these affect risk adjustment in NTDB and TQIP benchmarking. Watch for patterns to identify potential data problems. For example, your monthly comorbidity reports might show a fairly consistent percentage of older patients with hypertension. If you suddenly begin to see wide fluctuations in this field, check the patient charts to make sure they were abstracted correctly.

Data submission frequency reports. NTDB/TQIP submission frequency reports provide an excellent opportunity to use reporting to drive accuracy. Review the reports to identify any inappropriate use of null values. For example, the “report of physical abuse” field should always contain either yes or no, never “not known” or “not recorded.”

Benchmark reports. You should also examine NTDB/TQIP benchmark reports to identify the data fields that have the highest frequency of use and are most likely to trigger an exclusion. The following data fields typically have the greatest overall impact on benchmark and research reports for these programs:

Data fieldQuality check
NameVerify there are no duplications (as an internal validation to check for duplicate accounts prior to submission)
AgeCheck for blank fields
RaceCheck for blank fields
Arrival dateEnsure correlation with appropriate month of reconciliation
Date of dischargeEnsure correlation with appropriate month of reconciliation
CauseEnsure consistency through all three cause fields
MOIEnsure consistency through all three cause fields
Primary E-codeEnsure consistency through all three cause fields
Protective deviceConfirm correlation with appropriate cause/E-code
Other protective deviceConfirm correlation with appropriate cause/E-code
ISSVerify blank fields
Post-ED destinationConfirm location is logical in comparison to admit unit
Admit unitConfirm location is logical in comparison to post-ED destination

3. Develop reports that support hospital and physician goals
Data reporting is also an opportunity to prove the value of the trauma registry to hospital and physician leaders. Consider beginning with:

Utilization data. Monthly registry reports on hospital length of stay (LOS) and readmissions will help administrators manage resource utilization. Validate the data through selective checks. For instance, if a report shows that a trauma patient with a low ISS is readmitted to the hospital, check the records to confirm the discharge process. Was the patient sent home too early? Did he or she have enough pain medication and receive the right follow-up care? Were the appropriate physicians involved in the discharge process?

Physician benchmarking. Using registry data to develop a Physician Report Card will help the trauma medical director evaluate the core team more objectively. This report could include each physician’s monthly volume, patient count by ISS score (≤15 and >15), complications, fall-outs from patient audit filters, ICU LOS and hospital LOS. Cross-check ISS against procedures and complications and double-check all outlier LOS values. This exercise will help ensure physicians are evaluated fairly. It will also serve as another data validation checkpoint for the trauma registry.

Close the loop
Not every registry team will be able to tackle all these reports at once. That is why it often makes sense to work with a trauma registry outsource vendor with the resources to keep your registry current—and the expertise to provide you with a broad range of accurate and actionable reports.

But no matter how you proceed, getting started with a handful of key reports will help you transform your registry into a useful tool. The important point is to use data reporting to drive registrar learning. Reports help catch errors, but ultimately you need to use error findings to “close the loop” with registrars.

To find out how to create a practical and effective registrar feedback loop, download The Step-by-Step Guide to Improving Trauma Data Accuracy with Ongoing Registrar Learning. This free ebook from himagine solutions will show you how to develop an ongoing registrar feedback process that addresses the root causes of abstraction errors, reduces variability and improves trauma data quality.

Irene Lopez, BSN, RN, CSTR is trauma registry manager at himagine solutions.