Acute arthritis in the hospital: comparison of rheumatologic with nonrheumatologic care

Abstract

Certain health care reform proposals emphasize "primary" rather than specialty care, so it is important to document whether these changes might affect patients with rheumatic diseases. We therefore assessed outcome and costs of patients who were hospitalized with acute arthritis, comparing management by rheumatologists with nonrheumatologists. We reviewed charts retrospectively from 1991 to 1993 at our community medical center. Twenty patients with acute arthritis were managed by rheumatologists and 35 were managed by nonrheumatologists. Demographic, clinical, and rheumatologic features of patients were comparable.Rheumatologists ordered joint radiographs (65%) and performed diagnostic arthrocentesis (75%) significantly more often than nonrheumatologists (31 and 34%, respectively; p < 0.05). Rheumatologists' initial recorded diagnostic impressions were usually confirmed at discharge, whereas nonrheumatologists' more often were not (p < 0.05). Rheumatologists established definite diagnoses by American College of Rheumatology criteria significantly more often (75%) than nonrheumatologists (34%;p < 0.05). Nonrheumatologists selected antibiotics, systemic corticosteroids, nonsteroidal anti-inflammatory drugs, and allopurinol more often and intra-articular steroids (p < 0.05) and adrenocorticotropic hormone less often for a similar case mix than rheumatologists. Rheumatologists' clinical evaluations (p < 0.001), selection of diagnostic studies (p < 0.001), and therapeutic decisions (p < 0.005) were significantly more complete. Trends strongly favored rheumatologists' patients improving more rapidly (3.5 vs. 6.6 days; p = 0.06) and being hospitalized for shorter stays (7.4 vs. 14.7 days; p = 0.08) and favored rheumatologists' patients as managed at considerably less hospital cost ($8756 vs. $14,750).These limited observations suggest caution about equating nonspecialty with rheumatologic care and are consistent with suggestions that, at least for certain patients and as defined by our criteria above, rheumatologists offer more complete evaluation, more accurate diagnosis, more rational therapy, and better clinical outcome at lower cost than do nonrheumatologists.