Ensuring Data Quality and Integrity in CathPCI Abstraction
Boosting Accuracy through Expert Review and Process Improvement
Understanding the importance of data quality and integrity, an established client wished to review the data being collected by their existing CathPCI abstractors. The decision was made to use the subject matter experts at Registry Partners as a non-bias, third party to maintain impartiality and focus solely on the data collection and areas for improvement.
The review of the two on-site abstractors identified opportunities for process improvement initiatives and focused education for their team, leading to improved accuracy scores and data integrity.
Registry Partners performed an initial IRR review of 5 cases from each abstractor. Once the review was completed, findings were reviewed and education was provided by a member of the Registry Partners team. Following this review a focused re-abstraction of Q2 data was performed.
Q3 was re-abstracted by the on-site abstractors. Once that was complete, IRR was focused on Q4 abstractions.
The overall accuracy scores for DA 1 improved from 80.92% to 90.54%. DA 2’s overall accuracy went from 81.47% to 97.16%.
Accuracy Scores By Abstractor
During our review of the CathPCI data, we identified several key abstraction issues that were impacting both data accuracy and completeness. These opportunities for improvement, detailed below, were addressed through targeted education and process adjustments. By focusing on high-frequency, high-impact sequences, we were able to drive significant improvements in data quality.
Sequence 7400, Indications for Cath Lab Visit:
High frequency/high impact. AUC & risk adjusted mortality. *Registry Data Template did not follow NCDR definitions
Sequence 7825, Percutaneous Coronary Intervention Indication:
High frequency/high impact. AUC & risk-adjusted mortality. Capturing of Immediate PCI for STEMI. *Registry Data Template did not follow NCDR definitions
Sequence 5203, Stress Test Risk of Ischemia:
High frequency/high impact. AUC. Patients labeled as unclassifiable, may be appropriate, rarely appropriate. *Coding unavailable instead of referencing definition/crosswalk to make determination per NCDR data dictionary.
Sequence 6986/6991, Pre-procedure Medication:
High frequency/high impact. AUC. *Not capturing pre-hospital medications and/or utilizing MAR to capture antianginals.
Sequence 8001, Native Lesion Segment Number:
High frequency/low impact. The registry advises that sequentially treated segments be captured as one lesion (Seq# 8000) when the segments are treated at the same time – representative of a disease vessel region & the highest / worst percent stenosis of the lesion would be coded for Seq# 8004 (Stenosis Immediately prior to Treatment).
Sequence 3005/3010, Health Insurance:
High frequency/low impact: Capturing all that apply.
Sequence 8007, Pre TIMI flow:
High frequency/high impact: Missing documentation. Coding based on abstractor viewing cath images. Must be documented in the chart.
Sequence 7220, DAP:
High frequency/high impact: Incorrect units being abstracted into data tool. Using Ugy when GYcm2 was unit in the cath log.
Sequence 7415, Cardiovascular Instability:
High frequency/high impact: Coding for stable patients and not coding for STEMI.
Sequence 6095, Troponin Pre Procedure:
High frequency/high impact: Coding transfer facility troponins and not converting high sensitivity troponins because unit is ng/mL and in Epic ng/L.
Sequence 6100/6105, Total Cholesterol/HDL:
High frequency/low impact: Target value not being met. Coding these labs that were greater than 30 days pre cath.
Sequence 7841, Transferred in for Primary PCI for STEMI:
High frequency/high impact: Not capturing a STEMI transfer which impacted DTB time.
Sequence 5037, Pre EKG:
High frequency/high impact: Only capturing the EKG closest to procedure. The target value for 5034 states: “all values in 30 days OR between the last procedure and current one.”
Sequence 10060/10061, Discharge labs:
High Frequency/low impact: Capturing the post PCI lab
in both post & DC elements when only one lab was drawn post-procedure. It must be a unique specimen.
Sequence 7820, PCI multi-vessel disease:
High frequency/high impact: Not capturing as multi vessel per data dictionary.
Based on our findings, we provided the client with a comprehensive set of recommendations to ensure sustained data integrity moving forward. These process improvement initiatives were designed to enhance documentation practices, refine abstraction techniques, and create a more consistent and accurate data collection process. Below are the key recommendations that were implemented as part of this initiative.
- Continue use of case completion log capturing missing documentation & complications for ease of review and Physician follow up
- Ongoing STEMI case validation
- Submission weekly or biweekly for review instead of reviewing quarterly
- Drill down of AUC metrics to review the cath lab indications chosen were appropriate
- Attend bimonthly CathPCI NCDR RSM calls
- Review announcements on the NCDR home page weekly
- Send FAQ’s via “Contact Us” in the NCDR
After identifying the areas for improvement and implementing our recommendations, the client took actionable steps to refine their data abstraction process. These key improvements were successfully integrated into their workflows, leading to enhanced accuracy and more reliable data collection. Below are the specific process improvements that were put into practice.
- MD Documentation of Pre TIMI flow
- Cath Lab staff documentation of correct DAP and unit in Cath log
To Learn More About This Project and Explore Our Services