Session Information
Date: Monday, October 22, 2018
Title: Spondyloarthritis Including Psoriatic Arthritis – Clinical Poster II: Clinical/Epidemiology Studies
Session Type: ACR Poster Session B
Session Time: 9:00AM-11:00AM
Background/Purpose: Big data research is important for studying uncommon diseases in real-world settings. Most big data studies in axial spondyloarthritis (axSpA) have been limited to populations identified with billing codes for ankylosing spondylitis (AS). axSpA is a more inclusive concept, and reliance on AS codes does not produce a comprehensive axSpA study population. The first objective was to describe our process for establishing an appropriate sample of patients with and without axSpA for developing novel axSpA identification methods. The second objective was to determine the classification performance of AS billing codes against the chart-reviewed reference standard.
Methods: Veteran Health Affairs data, between January 2005 and June 2015, were used to randomly select patients with phenotypes determined by expert opinion to represent high, moderate, and low likelihoods of an axSpA diagnosis. A risk stratified sampling approach was applied to balance chart review feasibility(enrichment with high risk patients) with generalizability (inclusion of low risk patients). With chart review, the sampled patients were classified as Yes, No, or Uncertain axSpA status. These classification assignments were used as the reference standard for determining the positive predictive value (PPV) and sensitivity of AS ICD-9 codes for axSpA
Results: A higher percentage of patients were included from the high risk stratum than the moderate and low risk strata (0.83%, 0.25%, 0.01%, respectively) (Table 1). Six hundred patients were classified as Yes axSpA (26.8%), No axSpA (68.3%), and Uncertain axSpA (4.8%) (Table 2). The PPV of an AS ICD-9 code for axSpA was 83.3% and the sensitivity was 57.3% (Figure 1).
Conclusion: Standard methods of identifying axSpA patients with AS diagnosis codes lacked sensitivity. An appropriate sample of patients with and without axSpA was established for developing novel axSpA identification methods that are anticipated to enable previously impractical big data research
Table 2. AxSpA classification by chart review
|
|
All |
High risk for AxSpA |
Moderate risk for AxSpA |
Low risk for AxSpA |
|||
|
|
No. [%] (95% CI) n=600 |
AS No. (95% CI) n=100 |
B27+ No. (95% CI) n=100 |
Non-AS SpA subtype No. (95% CI) n=100 |
Sacroiliitis No. (95% CI) n=100 |
SpA mimics No. (95% CI) n=100 |
Chronic back pain No. (95% CI) n=100 |
|
Yes AxSpA |
162 [27.0] (23.5-30.6) |
87 (80.4-93.6) |
38 (28.5-47.5) |
27 (18.3-35.7) |
7 (2.0-12.0) |
2 (0.0-4.7) |
1 (0.0-3.0) |
|
No AxSpA |
409 [68.2] (64.4-71.9) |
4 (0.2-7.8) |
57 (47.3-66.7) |
63 (53.5-72.5) |
89 (82.9-95.1) |
97 (93.7-100.0) |
99 (97.0-100.0) |
|
Uncertain AxSpA |
29 [4.8] (3.1-6.5) |
9 (3.4-14.6) |
5 (0.7-9.3) |
10 (4.1-15.9) |
4 (0.2-7.8) |
1 (0.0-3.0) |
0 (0.0-0.0) |
No. = number. CI = Confidence interval
Table 1. Selection of patients sampled for the chart review population
|
Subgroups |
Subgroup Criteria (ICD-9 or laboratory data) |
No. of Veterans |
No. of Veterans selected to chart review population |
% from each risk stratum selected to the chart review population |
|
High risk for AxSpA |
|
|
|
|
|
Ankylosing spondylitis |
720.0 |
15,862 |
100 |
0.83 |
|
HLA-B27 positivity |
positive B27 test result |
8,168 |
100 |
|
|
Moderate risk for AxSpA |
|
|||
|
Sacroiliitis |
720.2 |
50,603 |
100 |
0.25 |
|
SpA subtype other than AS |
|
100* |
||
|
Spondyloarthritis NOS |
720.8x and/or 720.9x |
6,319 |
||
|
Reactive arthritis |
711.x and/or 99.3 |
1,072 |
||
|
Psoriatic arthritis |
696.0 |
22,625 |
||
|
Enteropathic arthritis |
713.1 AND either 555.x OR 556.x |
521 |
||
|
Low risk of AxSpA |
|
|||
|
Chronic back pain |
(≥2 ICD-9 codes for back pain ≥ 3 months apart [724.1, 724.2, 724.5]) |
2,069,644 |
100 |
0.01 |
|
Non-SpA rheumatologic disease |
|
|
100* |
|
|
DISH |
721.6 |
2,963 |
||
|
Crystal arthritis |
274.x and/or 712.x |
675,799 |
||
|
Rheumatoid arthritis |
714.x |
143,620 |
||
|
Other inflammatory arthritis |
CTD (710.x), vasculitis (273.2, 446.0, 446.4, 446.5, 446.7), PMR (725), Paget’s (731.0), sarcoidosis (135) |
135,608 |
*25 patients from each subcategory of spondyloarthritis NOS, reactive arthritis, psoriatic arthritis, enteropathic arthritis, DISH, crystal arthritis, rheumatoid arthritis, and other inflammatory arthritis.
To cite this abstract in AMA style:
Penmetsa G, Walsh J. Cohort Identification of Axial Spondyloarthritis in a Large Healthcare Dataset: Current and Future Methods [abstract]. Arthritis Rheumatol. 2018; 70 (suppl 9). https://acrabstracts.org/abstract/cohort-identification-of-axial-spondyloarthritis-in-a-large-healthcare-dataset-current-and-future-methods/. Accessed .« Back to 2018 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/cohort-identification-of-axial-spondyloarthritis-in-a-large-healthcare-dataset-current-and-future-methods/