Back to Annual Meeting

Back to Annual Meeting

APHA Scientific Session and Event Listing

Recorded presentation

Jasvinder A. Singh, MBBS, MPH, Medicine, Minneapolis VA Medical Center, One Veterans Drive, Rheumatology (111 R), Minneapolis, MN 55417, 612-467-4190, jasvinder.singh@va.gov and Siamak Noorbaloochi, PhD, Center for Chronic Diseases Outcomes Research, Minneapolis VA MEdical Center, One Veterans Drive, Minneapolis, MN 55417.

Objective: To study the accuracy of diagnoses of Spondyloarthropathies in the administrative databases at Minneapolis VA Medical Center (MVAMC).

Method: A random sample was selected from a cohort of all patients at MVAMC rheumatology clinic that had at least one clinic visit between 1/1/1 and 7/31/2 and data were collected from clinical and administrative databases and patients' clinical charts at MVAMC. We drew a random sample of 184 patients from the cohort of 787 patients. Two blinded reviewers with moderate to excellent agreement in record abstraction (kappa statistic of 0.69 for any rheumatic diagnoses and 1 for spondyloarthropathies), who were independent of the rheumatologists that made the diagnoses examined the medical charts. A chart diagnosis of Ankylosing Spondylitis (AS), Psoriatic Arthritis (PsA) and Reactive arthritis (ReA) by a rheumatologist on at least two visits more than 6 weeks apart was considered the gold standard. We compared the presence of an International Classification of Diseases (ICD) diagnostic code for diagnoses of AS (720.0), PsA (696.0) or ReA coded as Reiter's Syndrome/Reiter's arthritis (099.3, 711.11-711.19) in the local Decentralized Hospital Computer Program (DHCP) database to the gold standard. The accuracy of ICD diagnoses was assessed by calculating sensitivity, specificity, positive and negative predictive values.

Results: Of 184 randomly selected rheumatology clinic patients, 6% (11/184) had AS, 9% (17/184) had PsA and 4% (7/184) had ReA as per the gold standard. Mean age of the patients with AS was 60±13 years (mean ± standard deviation), PsA, 66±13 years, and ReA 49±6 years. All patients were men.

ICD-codes for AS, PsA and ReA were very specific (98-100%) with excellent negative predictive values (99-100%) as compared to the gold standard of a rheumatologist's diagnosis in the chart. Sensitivity of ICD-codes was 91% for a diagnosis of AS, 100% for PsA and 71% for ReA. Positive predictive values of ICD-codes were 83% for AS, 85% for PsA and 100% for ReA.

Conclusions: The ICD-codes for AS, PsA and ReA in Minneapolis VA administrative database have excellent specificity and good sensitivity and predictive value. This finding in conjunction with our previous finding of poor accuracy of ICD-code for RA implies that the accuracy of ICD-code for rheumatic conditions may vary by disease.

Learning Objectives:

• Describe the accuracy of administrative diagnoses of spondyloarthropathies in Veterans Affairs Databases
• Recognize the benefits and limitations of using VA administrative databases to identify patient cohorts of Spondyloarthropathies

Keywords: Disease Data, Veterans

Presenting author's disclosure statement:

Any relevant financial relationships? No

Recorded presentation