A Wider Perspective on Data Validation: A Multiregion Collaboration

Abstract

INNOVATIONS BACKGROUNDThe trauma registry data are key to informing quality improvement, prevention initiatives, research, clinical guidelines, and other topics related to the in-jured patient at all levels: from the individual facility to nationally (Porgo et al., 2016;Protetch & Chappel, 2008). The usefulness of the registry relies on data quality, including completeness and accuracy. Thus, there should be ongoing review of registry data. Resources for Optimal Care of the Injured Patient (American College of Surgeons & Committee on Trauma, 2014) recommends that hospitals validate registry data by reabstracting 5%-10% of patient records.There are additional practices beyond full reabstraction to support validation. Centers can build validation rules within their registries to facilitate concurrent review and support entry of valid data from the outset (Rees, 2020). Targeting key data fields most relevant to an individual trauma program can be helpful for centers with high patient volumes for whom full reabstraction may be a challenge (Fojut, 2015).Impact of data quality issues starts at the trauma center level. Errors can contribute to inaccurate trends and misclassifications, affecting accurate tracking of performance improvement efforts (Hlaing et al., 2006). A study looking at heterogeneity of registry data across trauma centers in one state found that erroneous data varied widely by center, which affects the homogeneity needed for good comparison and benchmarking among centers (Dente et al., 2016).

Pilot Study of a Software Application to Identify Trauma Registry Inconsistencies

“…Consistent with this study, a recent preliminary report suggests that logic-based data validations can supplement existing validation processes (Houck et al, 2022). Houck et al formed a state-level workgroup to perform 14 logic-based validations across two project phases.…”