Comparison of Pediatric and Internist Providers of Pediatric Rheumatology Care
| Pediatric Rheumatologists Only (n = 20) | Internists Involved in Care of Children (n = 78) | Unadjusted P Value* Pediatric vs Internists Who Treat Children | |
|---|---|---|---|
| Age distribution | |||
| 0–5 y | 20.0 | 0.0 | <.001 |
| 6–11 y | 30.0 | 4.5 | <.001 |
| 12–15 y | 32.5 (9.84) | 33.6 (22.59) | NS |
| 16–17 y | 16.4 (9.77) | 52.4 (27.73) | <.001 |
| Disease distribution among pediatric patients | |||
| JRA | 41.5 (17.5) | 44.1 (31.3) | NS |
| SLE | 19.3 (13.6) | 23.0 (25.7) | NS |
| Pain syndromes† | 7.5 | 1.0 | NS |
| Other rheumatic disease† | 11.5 | 5.0 | NS |
| Rheumatic compliant with no diagnosis† | 5.0 | 0.0 | NS |
| Comfortable treating | |||
| Dermatomyositis | 95.0 | 56.2 | .001 |
| JRA, systemic onset | 95.0 | 67.1 | .01 |
| JRA, pauciarticular | 100.0 | 80.8 | .04 |
| JRA, polyarticular | 100.0 | 80.8 | .04 |
| Fibromyalgia | 100.0 | 64.4 | .002 |
| Kawasaki disease | 100.0 | 24.7 | <.001 |
| Polyarteritis Nodosa | 85.0 | 43.8 | .001 |
| Psoratic arthritis | 90.0 | 80.8 | NS |
| Reflex sympathethic dystropy | 90.0 | 34.3 | <.001 |
| Rheumatic fever | 100.0 | 49.3 | <.001 |
| Psychogenic rheumatism | 90.0 | 30.1 | <.001 |
| Anterior knee syndrome | 85.0 | 50.7 | .005 |
| SLE | 95.0 | 72.6 | <.001 |
| Spondylarthropathy | 100.0 | 79.5 | .02 |
| Wegener’s granulomatosis | 85.0 | 42.5 | .001 |
| Henoch-Schonlein purpura | 100.0 | 61.6 | .001 |
| Fever of unknown origin | 100.0 | 28.8 | <.001 |
| Back pain | 85.0 | 60.3 | .06 |
| Reason for treating | |||
| Personal expertise | N/A | 78.4 | N/A |
| Patient preference | N/A | 73.6 | N/A |
| Distance to nearest pediatric rheumatologist | N/A | 65.8 | N/A |
| Lack of available appointment | N/A | 38.4 | N/A |
| Inability to refer because of insurance | N/A | 31.9 | N/A |
| Practice income considerations | N/A | 4.4 | N/A |
| Other | N/A | 42.1 | N/A |
| Interested in obtaining advice from pediatric rheumatologist | N/A | 71.4 | N/A |
SLE indicates systemic lupus erythematosus.
↵* For dichotomous variables, statistical tests were performed using χ2 when all cell sizes were >5. When ≥1 cells were <5, a 2-sided Fisher exact test was used. For continuous variables, 2-sided t tests were performed without assuming equal variances. For nonnormally distributed variables (Shapiro-Wilks test < .05), Wilcoxon rank sum tests were used.
↵† Medians are presented for variables with highly skewed distributions for one or both groups.