OM1 Lupus Data & Research

Unlock the Lupus Patient Journey with OM1

Data Depth & Breadth

Longitudinal Patient Journeys

Evidence Generation

Preferred Research Partner
Prospective Studies to supplement Real-World Data

AI & Machine Learning

Patient Identification
Endpoint Amplification

Clinically Rich Data to Power Your Research

OM1 SLE Specialized Data

OM1 Evidence Generation Networks

A Better Approach for Prospective Research

Accessing, processing, and harmonizing complex and extensive RWD to support clinical study needs is costly for sponsors and burdensome for providers.

Established RWD evidence generation networks are transforming prospective research and resulting in faster times to study launch, more complete and reliable data, and lower costs.

| Use Case

Understanding the role of race in response to hydroxychloroquine treatment in SLE

Challenge

Understanding the effectiveness of hydroxychloroquine in the real world across specific racial subtypes with Systemic Lupus Erythematosus (SLE).

Solution

Leverage OM1 clinical data along with SLEDAI and the OM1 generated eSLEDAI to understand response or differential effect based on race.

Results

After 3 months of continuous treatment with hydroxychloroquine, 7.6% of Caucasians and 7.6% of African Americans experienced a 2 point reduction in SLEDAI.
After 12 months of continuous treatment with hydroxychloroquine, 10.1% of whites and 12.8% of African Americans experienced a 2 point reduction in SLEDAI.

97.8% of Caucasians and 38.5% African Americans were on systemic treatment at presentation

| Use Case

Identify Patient Subtypes for Lupus Nephritis (LN)

Characteristics and Disease Management in a Large Real-World Lupus Nephritis Cohort in the U.S.

View full publication

Objectives

Some evidence suggests men with SLE more frequently develop LN and with greater severity. The heterogeneity of LN complicates development of new treatment strategies. The purpose of this study was to characterize LN patients compared to non-LN SLE patients in a large, longitudinal, diverse dataset.

Methods

OM1 SLE Registry patients were identified as having new onset or prevalent LN based on relevant diagnosis codes (e.g., glomerular disease in SLE, glomerulonephritis, chronic kidney disease, proteinuria; ≥ 2 codes at least 30 days apart) or documented dialysis or kidney transplant.

Findings

20% of SLE patients have characteristics of Lupus Nephritis. More patients with LN were black and men and cardiovascular (CVD) risk factors were more common in LN patients.

Understanding the heterogeneous characteristics and course of LN patients is critical to refining clinical trial populations and improving renal and CVD outcomes.

Harness deep clinical data to identify subtypes of patients for your cohort.

| Use Case

Calculation of Swollen to Tender Joint Count Ratio in a Real World SLE Cohort

Joint Involvement and Disease Activity in Systemic Lupus Erythematosus (SLE) Patients: Calculation of Swollen to Tender Joint Count Ratio in a Real World Cohort in the U.S.

View full poster

Objectives

To characterize SLE patients in a real-world cohort based on disease activity as defined by STR.

Methods

The OM1 SLE Registry provided patients for evaluation. Patients ≥16 years of age with swollen and tender joint counts based on 28 joints on the same encounter were included. STRs were calculated by inserting 1 if the denominator was 0.

Findings

Moderate and high STRs were more common among patients with SLE plus RA, psoriatic arthritis, or ankylosing spondylitis compared to patients with SLE alone.

Differences in treatments received were apparent between patients of varying disease activity groups with trends towards increased use among patients with higher disease activity.

Disease activity measures identify patients with increased burden of disease

| Use Case

Use AI for SLEDAI Endpoint Amplification

Validation of a Machine Learning Approach to Estimate SLE Disease Activity Index (SLEDAI) Score Categories and Application in a Real-World Dataset

View full publication

Objective

Use of SLEDAI scores in routine clinical practice is inconsistent and limited. We validated a machine learning model to estimate SLEDAI score categories using clinical notes and applied the model to a large, real-world dataset to generate estimated score categories.

Methods

OM1 employed one of the first uses of machine learning to estimate SLE disease activity score categories from unstructured clinical notes.

Conclusion

With a validation AUC of 0.91, the model performed superbly when estimating SLEDAI score categories using unstructured clinical notes.

Machine learning models can be successfully used to estimate disease activity scores. The OM1 Patient Finder^TM generated 30x more disease activity datapoints for analytics.