This study is designed to test the helpfulness of a diagnostic tool, SimulConsult, when clinicians are diagnosing pediatric rheumatic diseases. Trainees and nurse practitioners will use the software or not use it, and their differential diagnoses will be compared to those of attending physicians in the same clinical encounter, and then to definitive diagnoses weeks later after testing, if one is reached.

The study aims to assess improvement in the clinical performance of trainees evaluating patients with a suspected rheumatologic disorder. The Intervention is the use of Diagnostic Decision Support Software (DDSS) already available as a Boston Children's Hospital resource.

• Main objective: To assess if the use of the DDSS by trainees and nurse practitioners can improve their diagnostic performance in developing a differential diagnosis and plan in real clinical use. (Studies have shown in prior research that the use of the DDSS does so for case vignettes, lowering diagnostic errors of trainees by 75%). The study will compare how closely the differential diagnosis and plan of the trainees and nurse practitioners approximate those of their senior Pediatric Rheumatology attending physician colleagues.
• Secondary objective: To assess how frequently the differential diagnosis of the trainees and nurse practitioners includes the definitive diagnosis by doing a manual chart review weeks after the visit. (Studies have shown in prior research that if the definitive diagnosis is in the initial differential diagnosis, the right tests get ordered, and diagnostic errors are avoided.)

The design is a parallel one, comparing subjects (trainees and nurse practitioners) using the DDSS (Intervention Arm) to subjects who do not use the DDSS (Control Arm). In both study arms, subjects may use all other traditional resources. Arm assignment will be by patient case rather than trainee / nurse practitioner to reduce variation due to each subject's different capabilities (years of training and familiarity with pediatric rheumatology).

The subjects are trainees (rheumatology and rotating fellows) and nurse practitioners, not patients. The Attendings participate in their supervisory role, and their clinical judgments are recorded as a preliminary "gold standard", but they are not research subjects.

• Trainees seeing new pediatric patients with a potential rheumatologic disorder are eligible. They will be drawn from the pool of Rheumatology fellows (3), Allergy/Immunology fellows (6), rotating pediatric and occasional adult residents (1 - 2) per month)
• Nurse practitioners in Rheumatology (2).
• Rheumatology Attendings for these new patient visits will also be asked to participate but will be supervisory participants in the study, not subjects.

We anticipate 15 trainees plus nurse practitioners will participate. There are no exclusion criteria, and we do not anticipate screen failures; however, trainees and nurse practitioners may decline to participate for individual patients or the entire study.

The results will be shared.

2021-2022

No masking is possible for subjects, since they will know whether they are or are not using the software.