Abstract Details

Title Validation of a Claims-based Algorithm to Identify Patients With Diabetic Peripheral Neuropathy (DPN) and Painful DPN Using Medical Charts

Topic Pain

Presentation(s) P8 - Poster Session 8 (11:45 AM-12:45 PM)

Poster/Presentation
Number 14-009

Objective To evaluate the accuracy of a claims-based algorithm for identifying diabetic peripheral neuropathy (DPN) and painful DPN patients using medical chart data.

Background Claims databases are frequently utilized for real-world studies of DPN and painful DPN patients, but no validated algorithm currently exists to identify DPN patients, and ICD-10-CM codes alone do not capture pain status. Validating a claims-based algorithm can improve the accuracy in identifying and characterizing DPN and painful DPN populations.

Design/Methods A retrospective study using medical and pharmacy claims data from the Optum Research Database (ORD) was conducted to validate potential DPN and painful DPN patients identified in claims in the year 2022 against linked abstracted medical chart data. The claims-based algorithm utilized ICD-10-CM codes for DPN to identify potential DPN, and pharmacy claims for pain medication (within 60 days on or after DPN claim) to identify potential painful DPN. DPN and painful DPN status using the claims-based algorithm was validated against diagnosis, treatments, and symptoms noted in medical charts. Positive predictive value (PPV), negative predictive value (NPV), sensitivity, and specificity were calculated.

Results Among 300 patients identified by the claims-based algorithm as having potential DPN, 95% had DPN confirmed by chart review. 118 (39%) of the 300 potential DPN patients were classified as having potential painful DPN by the claims-based algorithm. Chart review confirmed painful DPN in 82% of these patients. However, only 60% of 161 chart-identified painful DPN cases were captured by the claims-based algorithm.

Conclusions DPN patients identified via a claims-based algorithm are highly likely to have confirmed DPN, suggesting this algorithm has high utility for use in future real-world claims-based studies of DPN. Although the painful DPN claims-based algorithm accurately identifies painful DPN patients, some DPN patients experiencing pain may be missed when using diagnosis and prescription claims alone. Further refinements to the algorithm should be explored.

Authors/Disclosures
Deven M. Patel, PhD PRESENTER	Dr. Patel has received personal compensation for serving as an employee of Vertex Pharmaceuticals.
Hannah Bouldin Olivet, MD	Dr. Bouldin Olivet has received personal compensation for serving as an employee of Vertex. Dr. Bouldin Olivet has stock in Vertex.
Maureen Carlyle, MPH	Ms. Carlyle has received personal compensation for serving as an employee of Optum. Ms. Carlyle has stock in United Health Group.
Milada Mahic, PhD	Dr. Mahic has nothing to disclose.

��ɫ��

��ɫ��