[Skip to Navigation]
October 25, 1993

Serum Complements: Inappropriate Use in Patients With Suspected Rheumatic Disease

Author Affiliations

From the Department of Medicine, Santa Clara Valley Medical Center, San Jose, Calif. Dr Shlotzhauer is now with the Park Medical Group, Rochester, NY.

Arch Intern Med. 1993;153(20):2363-2366. doi:10.1001/archinte.1993.00410200089010

Background:  The diagnostic value of serum complement testing is well established in inherited complement deficiencies and glomerulonephritis. Their utility is less certain in diagnosing rheumatic diseases. We noted that complement tests were frequently ordered for patients who were referred to our rheumatology clinic. We sought to determine the clinical rationale for ordering complement tests in our hospital and the effect of the test results in patients with rheumatic diseases.

Methods:  We conducted a retrospective medical chart review of patients who had serum complement tests ordered at our hospital. We determined whether the test was ordered as a diagnostic tool in a patient with a suspected rheumatic disease. In these cases, we attempted to correlate the results of the complement tests with the patient's eventual diagnosis.

Results:  We obtained the medical charts of 130 patients who had 307 complement assays (C3, C4, or total hemolytic complement) performed between October 1988 and July 1989. The tests were ordered for diagnostic reasons in 68% of the patients; 54% of these were ordered by nonspecialists. The complement tests were ordered on 28 patients with suspected rheumatic diseases. The three patients with hypocomplementemia did not have a connective tissue disease. The 10 patients who eventually were diagnosed as having rheumatic disease all had normal serum complement levels. Additionally, we found that 77% of patients had more than one complement assay ordered. The test results were discordant in only 24% of these cases.

Conclusion:  Complement screening is not a useful diagnostic test in most patients with suspected rheumatic disease. Despite their lack of established diagnostic value, these tests were frequently performed in our hospital. Judicious use of complement testing would provide substantial cost savings without a loss of clinically relevant information. When the complement testing is clinically indicated, clinicians should consider using a single C3 assay initially rather than multiple assays unless a hereditary deficiency is suspected.(Arch Intern Med. 1993;153:2363-2366)