Session Information
Date: Tuesday, November 10, 2015
Title: Vasculitis Poster III
Session Type: ACR Poster Session C
Session Time: 9:00AM-11:00AM
Background/Purpose: To facilitate clinical care and
research, validated algorithms are needed to accurately identify patients with
Takayasu’s arteritis (TAK). This study sought to evaluate and validate case-finding
algorithms for TAK in 2 large healthcare administrative databases.
Methods: All adult patients with the International
Classification of Diseases version 9 (ICD9) code for TAK (446.7) were identified
from 2 large healthcare systems (102 from site 1 and 85 from site 2). 26
case-finding algorithms were constructed using a combination of ICD9 code,
encounter type (1 inpatient ICD9 code on 3 consecutive days or 2 outpatient ICD9
codes 3 months apart), physician specialty (Rheumatology, Cardiology, or
Vascular Surgery), use of immunosuppressive medications, age, and sex. The
diagnosis was confirmed by chart review using the ACR classification criteria
or the Chapel Hill Consensus Conference definitions for TAK.
Results: 102 patients from the first healthcare
system (site 1) and 85 patients from the second system (site 2) were included
in the analysis. 47/102 (46%) patients had a confirmed diagnosis of TAK at site
1 and 35/85 (42%) patients at site 2. Table 1 shows the positive and negative
predictive values of the studied algorithms in each healthcare system. An
algorithm including the encounter type, physician specialty, age, and immunosuppressive
medications had the highest average positive predictive value (PPV: 76.9% and
88.2 % respectively). An algorithm including only the physician specialty had
the highest average negative predictive value (NPV: 90.2% and 100%
respectively).
Conclusion: Case-finding algorithms can accurately
identify patients with TAK using large administrative databases. A simple
algorithm including the encounter type, physician specialty, age, and immunosuppressive
medications had the highest positive predictive value. Similarly, an algorithm
including only the physician specialty had the highest negative predictive
value. These algorithms can be used to assemble a population-based cohort of
patients with TAK and facilitate future research in healthcare use, outcomes,
and comparative effectiveness.
|
Table 1. Test characteristics of Algorithms for Takayasu’s Arteritis |
||||||
|
Algorithms for Takayasu’s Arteritis |
Site 1 (n=102) |
|
Site 2 (n=85) |
|||
|
|
PPV |
NPV |
|
PPV |
NPV |
|
|
ICD9+ |
|
|
|
|
|
|
|
Sex |
45.7 |
52.4 |
|
42.4 |
63.2 |
|
|
Age |
58.5 |
75.7 |
|
64.3 |
70.2 |
|
|
Encounter |
51.3 |
69.2 |
|
74.4 |
92.9 |
|
|
Specialty** |
70.5 |
90.2 |
|
52.2 |
100.0 |
|
|
Medications |
53.8 |
56.6 |
|
47.8 |
83.3 |
|
|
|
|
|
|
|
|
|
|
ICD9+Sex+ |
|
|
|
|
|
|
|
Age |
|
59.2 |
66.0 |
|
73.7 |
68.2 |
|
Encounter |
50.8 |
61.5 |
|
72.2 |
81.6 |
|
|
Specialty |
|
68.0 |
75.0 |
|
53.8 |
78.8 |
|
Medications |
50.0 |
54.9 |
|
46.3 |
67.7 |
|
|
|
|
|
|
|
|
|
|
ICD9+Age+ |
|
|
|
|
|
|
|
Encounter |
60.8 |
68.6 |
|
83.3 |
70.1 |
|
|
Specialty |
|
72.9 |
77.8 |
|
75.0 |
72.1 |
|
Medications |
64.7 |
57.6 |
|
75.0 |
72.1 |
|
|
Specialty+Medications |
76.9 |
58.4 |
|
81.8 |
73.0 |
|
|
Encounter +Sex |
62.5 |
64.5 |
|
85.7 |
67.6 |
|
|
Specialty+Sex |
71.1 |
68.8 |
|
77.8 |
68.7 |
|
|
Medications+Sex |
66.7 |
56.7 |
|
82.4 |
69.1 |
|
|
|
|
|
|
|
|
|
|
ICD9+Encounter+ |
|
|
|
|
|
|
|
Specialty+Medications |
76.5 |
60.0 |
|
74.4 |
87.0 |
|
|
Specialty+Age |
|
74.4 |
71.4 |
|
83.3 |
70.1 |
|
Medications+Age |
|
64.7 |
57.6 |
|
88.2 |
70.6 |
|
Specialty+Sex |
|
52.6 |
55.4 |
|
72.2 |
81.6 |
|
Specialty+Medications+Age* |
76.9 |
58.4 |
|
88.2 |
70.6 |
|
|
Specialty+Medications+Sex |
|
75.0 |
57.8 |
|
69.7 |
76.9 |
|
Specialty+Age+Sex |
|
71.9 |
65.7 |
|
69.7 |
76.9 |
|
Medications+Age+Sex |
|
66.7 |
56.7 |
|
85.7 |
67.6 |
|
Specialty+Medications+Age+Sex |
|
77.8 |
57.0 |
|
85.7 |
67.6 |
|
|
|
|
|
|
|
|
|
ICD9+ Specialty+Medications+Age+Sex |
|
77.8 |
57.0 |
|
82.4 |
69.1 |
|
* Algorithm with the highest average PPV ** Algorithm with the highest average NPV. PPV: positive predictive value. NPV: negative predictive value. ICD9: ICD9 code 446.7. ENCOUNTER: 1 inpatient ICD9 code on 3 consecutive days or 2 ICD9 codes 3 months apart. SPECIALTY: a Rheumatologist, Cardiologist, or a Vascular Surgeon involved in the care of the patient. MEDICATIONS: an immunosuppressive medication used. AGE: current age < 50. Sex: = female. |
||||||
To cite this abstract in AMA style:
Annapureddy N, Sreih AG, Byram K, Casey G, Frangiosa V, George M, Sharim R, Sangani S, Merkel PA. Evaluation and Validation of Case-Finding Algorithms for the Identification of Patients with Takayasu’s Arteritis in Large Healthcare Administrative Databases [abstract]. Arthritis Rheumatol. 2015; 67 (suppl 10). https://acrabstracts.org/abstract/evaluation-and-validation-of-case-finding-algorithms-for-the-identification-of-patients-with-takayasus-arteritis-in-large-healthcare-administrative-databases/. Accessed .« Back to 2015 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/evaluation-and-validation-of-case-finding-algorithms-for-the-identification-of-patients-with-takayasus-arteritis-in-large-healthcare-administrative-databases/