Abstract: FR-PO1091
Use of Tokenization to Overcome Coding Challenges in Real-World Data: Identification and Assessment of Remote Patient Monitoring (RPM) in Peritoneal Dialysis (PD)
Session Information
- Home Dialysis: Modalities, Barriers, Utilization, and Clinical Outcomes
October 23, 2026 | Location: Exhibit Hall A, Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Dialysis
- 802 Dialysis: Home Dialysis and Peritoneal Dialysis
Authors
- Evans, Kathryn, PPD™ Evidera™ Real-World Data & Scientific Solutions, Thermo Fisher Scientific, Waltham, Massachusetts, United States
- Shaw, Rhonda, Vantive Holding B.V., Utrecht, Netherlands
- Keller, Brad, Vantive US Healthcare LLC, Deerfield, Illinois, United States
- Brown, Julie, Vantive US Healthcare LLC, Deerfield, Illinois, United States
- Udayachalerm, Sariya, PPD™ Evidera™ Real-World Data & Scientific Solutions, Thermo Fisher Scientific, Waltham, Massachusetts, United States
- Danielson, Vanessa, Vantive Limited, Liverpool, United Kingdom
- Huang, Qiushi, PPD™ Evidera™ Real-World Data & Scientific Solutions, Thermo Fisher Scientific, Waltham, Massachusetts, United States
- Berger, Ariel, PPD™ Evidera™ Real-World Data & Scientific Solutions, Thermo Fisher Scientific, Waltham, Massachusetts, United States
Background
Healthcare claims are a well-established and timely way to examine disease epidemiology and outcomes but rely on availability and specificity of relevant codes. We examined how tokenization may "bridge" instances where such codes are unavailable, using as an example the Sharesource® Platform (SSP)—a two-way RPM platform that allows healthcare professionals to remotely monitor patients who use PD at home —that does not currently have specific procedure codes associated with its use.
Methods
We tokenized SSP data to Norstella healthcare claims. Adult patients who used PD between 1 January 2020 and 31 December 2023 were identified (earliest PD code in this time deemed "PD date"); those who began SSP within 30 days before or any time after PD date formed the SSP cohort (initiation date of SSP deemed "index date"). Duration between PD date and index date was used to assign "pseudo index date" to comparators who did not use SSP. All patients were required to have 6 months of continuous enrollment prior to index. Demographic and clinical characteristics were compared between groups as of PD date. Procedure codes related to RPM or at-home dialysis-related services ±30 days of index date were assessed.
Results
A total of 63,824 patients in Norstella met selection criteria for PD; 3,835 (6.0%) initiated SSP; 2,476 (64.6% of SSP patients) had 6 months of continuous enrollment. SSP patients (vs. those without SSP) were more likely to be White (41.8% vs. 33.5%), commercially insured (40.8% vs. 22.9%), and to have hyperlipidemia (47.9% vs. 33.5%); they were less likely to have hyperparathyroidism (42.4% vs. 61.1%) (all p<0.001). Few (<2%) SSP patients had codes specific to RPM (any type) ±30 days of index date; most (73.5%) had non-specific codes for at-home dialysis services in this interval.
Conclusion
Without the ability to leverage tokenization when specific codes are unavailable, potentially important findings on access challenges—such as is suggested in our data related to SSP—may go unidentified. Development of codes (and corresponding billing and reimbursement guidelines) specific to SSP and other RPM modalities would improve abilities to leverage relatively accurate and "timely" sources of real-world data such as healthcare claims databases to address questions of their access and impact.
Funding
- Commercial Support – Vantive