A137: Appropriateness of Rheumatology Lab Studies and Referrals among General Pediatricians

Background/Purpose:Rheumatology lab tests are often of low diagnostic utility when ordered by primary care physicians, and may lead to erroneous referrals. No prior studies have investigated pediatricians' reasons for ordering these tests. We surveyed pediatricians in three states of the Upper Midwest regarding their ordering of rheumatology lab tests and reasons for referral to pediatric rheumatology (PR).Methods:A web-based survey was emailed to general pediatricians practicing in Minnesota, North Dakota, and South Dakota. The rheumatology tests queried included antinuclear antibody (ANA), anti-double-stranded DNA antibody (anti-dsDNA), rheumatoid factor (RF), anti-citrullinated peptide antibodies (anti-CCP), and HLA-B27. Basic descriptive analysis was performed. Subgroup analysis was performed to assess the relationship between the pediatrician's experience with PR (defined as years since completion of residency training, whether or not the training program had a PR division, time spent in a PR clinic during training and completion of continued medical education focused on PR in the last three years) and the appropriateness of lab tests and PR referrals.Results:The response rate was 15.3% (93/609). Eighty-nine percent (83/93) of the respondents had ordered ANA, for reasons including abnormal joint exam (72%), musculoskeletal (MSK) pain with a normal exam (52%), concern for juvenile idiopathic arthritis (JIA) (47%), rash (41%), concern for systemic lupus erythematosus (39%), fatigue (35%), and fever (35%). Fifty-nine percent (54/92) had ordered anti-dsDNA, and the distribution of reasons paralleled that of ANA. Seventy-three percent (67/92) had ever ordered RF, most commonly for abnormal joint exam (67%), joint pain with a normal exam (60%), concern for JIA (42%), and fever (30%). Only 8% (7/92) had ordered anti-CCP. Forty-two percent (39/92) had ordered HLA-B27, primarily for back pain (59%). Of those who had ever referred a patient to a pediatric rheumatologist, 44% (37/85) did so for a positive ANA of unclear significance, 14% (12/85) for positive RF of unclear significance, and 9% (8/85) for anti-dsDNA of unclear significance. Positive HLA-B27 and anti-CCP were less common referral reasons. Subgroup analysis did not show significant differences between experience level and appropriateness of lab tests and referrals to PR.Conclusion:Inappropriate ordering of rheumatology lab tests with subsequent referral to PR is common among pediatricians in the Upper Midwest and likely elsewhere. The correctness of lab test utilization and referral to PR did not obviously improve with increased experience in PR. New educational interventions are needed to improve general pediatricians' understanding and utilization of rheumatology lab tests and to decrease unnecessary referrals to PR.