Abstract Rationale Sarcoidosis is commonly managed with long-term oral corticosteroids (OCS), which increase risk for hyperlipidemia, diabetes, and osteoporosis. In other rheumatologic diseases, guidelines recommend routine monitoring for these complications, but sarcoidosis guidelines provide no specific screening schedule. Clinicians prescribing steroid-sparing agents (SSAs) may demonstrate greater adherence to best-practice monitoring. We hypothesized that patients receiving both OCS and SSAs would have higher rates of screening for OCS-related comorbidities than those managed with steroid monotherapy (SM). Methods We conducted a retrospective electronic medical record study across an urban safety-net hospital and a tertiary referral center serving ∼3.1 million residents. Adults with ≥2 ICD-10 codes for sarcoidosis (D86.xx) and ≥12 months of continuous OCS therapy were included. Data extracted included demographics, insurance type, comorbidities (hypertension, diabetes, osteoporosis), and immunosuppressive medication use (OCS and SSAs: methotrexate, azathioprine, leflunomide, mycophenolate, hydroxychloroquine, infliximab, adalimumab). Random Forest models evaluated predictors of screening for OCS-related comorbidities (hemoglobin A1c, lipid panel, and DEXA scan) using demographic, socioeconomic, and clinical variables. Results We identified 4002 patients who received OCS for 12 months; 1373 (34%) also received an SSA. Screening rates for steroid-related comorbidities were low overall: 29% had a lipid panel, 24% had a hemoglobin A1c, and 19% had a DEXA scan documented within the study period. By comparison, 95% of SSA-treated patients had a white blood cell count within 3 months of therapy initiation, suggesting substantially higher routine monitoring. Across all Random Forest models, SSA use was consistently associated with a higher likelihood of appropriate screening. For lipid panel screening, the predicted probability of not being screened was 0.63 (95% CI, 0.60-0.66) among SSA users versus 0.73 (0.70-0.75) among SM patients. Similar trends were observed for hemoglobin A1c (0.69 vs. 0.77) and DEXA screening (0.75 vs. 0.87). No other demographic or socioeconomic variable demonstrated a comparably strong or consistent association with screening behavior. Conclusions Among patients with sarcoidosis receiving prolonged OCS therapy, screening for steroid-related comorbidities was infrequent. Concurrent SSA use was associated with a predicted 8-12% absolute improvements LH1 in screening likelihood across all domains, suggesting that multidisciplinary or specialty-based care may improve adherence to monitoring standards. Defining and implementing measurable indicators of high-quality sarcoidosis care is urgently needed. This abstract is funded by: NIMHD
Harper et al. (Fri,) studied this question.