Background/Purpose: The ANA test is nonspecific and can be difficult to interpret without understanding the limitations of the test. Referrals to Rheumatology for positive ANA is a common practice that may lead to unnecessary resource utilization and contribute to delays in patients seeing a rheumatologist. At the University of Chicago, the current wait time for a new patient visit is over 60 days. The aim of this project is to evaluate the quality of our ANA referrals by analyzing how often these patients are diagnosed with an ANA-associated rheumatic disease (AARD) and the associated healthcare expenditures.

Methods: Charts were reviewed for patients referred to the University of Chicago outpatient rheumatology clinic over a 6-month period (April 2017-September 2017) with “positive ANA” as the reason for referral, or if “positive ANA” was a listed problem in the initial Rheumatology note’s plan. Demographic data, referral information, relevant rheumatologic labs and imaging were recorded. Patients’ final diagnoses were organized into the following categories: AARD, possible AARD (ongoing work-up at time of review) or no AARD.  Positive predictive value for AARD was calculated using the total number of ANA referrals and the number of referrals diagnosed with AARD. Costs of lab tests and imaging were estimated using the Healthcare Bluebook and lab testing websites.

Results: Eighty-three patients referred for positive ANA were evaluated. The majority of patients were female (64/83, 77%). Most patients were either Caucasian (36/83, 43%) or African American (36/83, 43%) with an average age of 40 years (range 19-85 years).

The majority of referrals were internal from University of Chicago providers (58/83, 70%). Of these referrals, 50% came from primary care clinics. The six most common reasons for ordering an ANA were: joint pain, fatigue, neuropathy, rash, headache and interstitial lung disease. During initial rheumatologic visit, a total of 234 lab tests were ordered, with an estimated cost of $23,592. A total of 41 joint x-rays were ordered, with an estimated cost of $3,731. Table 1 shows the most common lab tests ordered and the number of positive tests. The majority of tests were unremarkable.

AARD was diagnosed in 5 patients (2 SS, 1 overlap CTD, 2 UCTD). Possible AARD included 7/83 (8%) patients and 71/83 (85%) had no systemic AARD. The positive predictive value (PPV) for AARD was 5.7%.

Conclusion:  Our review shows that the PPV of diagnosing AARD in our system is very low and ANA referrals are contributing to unnecessary cost, resource use, and longer wait times for our clinic. As many referrals are internal, the next step of this project will be a multifold intervention including an educational component for primary providers and creating a decision-making support tool in our electronic medical record with the goal of improving the quality of referrals and creating more space for urgent patients to be seen.