Patient, examination, and system factors associated with completed follow-up for probably benign (BI-RADS 3) breast findings in a large, complex health care system

Catherine S Giess,Elyse Lynch,Ronilda Lacson,Pamela J DiPiro,Fatima Salah,Stephanie H Chung,Gary X Wang,Ramin Khorasani
DOI: https://doi.org/10.1016/j.acra.2024.09.037
2024-10-14
Abstract:Rationale and objectives: To investigate patient, examination, and system factors associated with follow-up completion for probably benign breast (BI-RADS 3) findings. Materials and methods: Retrospective review identified all BI-RADS 3 mammography, ultrasound, and MRI reports (index studies) for unique patient encounters at a large multi-institution Health Care System Jan 1-Dec 31, 2021. Electronic health record supplied patient demographics and examination type; Radiology Information System supplied ordering and scheduling information. University of Wisconsin's Neighborhood Atlas was used to map patient home addresses to determine area deprivation index (ADI). Univariable and multivariable analyses identified variables associated with noncompleted BI-RADS 3 follow-up. Results: Among 8834 BI-RADS 3 examinations, 6778 (76.7%) had follow-up imaging completed within 15 months. Factors associated with incomplete follow-up on multivariable analysis included: ultrasound (Odds Ratio [OR] 0.22; 95% Confidence Interval [95%CI] 0.19-0.25); MRI (0.10, 95%CI 0.08-0.12); Asian race (0R 0.77; 95%CI 0.61-0.98); age< 40 years (OR 0.22; 95%CI 0.18-0.26); non-married status (single, OR 0.68; 95% CI 0.59-0.79; divorced OR 0.77; 95% CI 0.61-0.97; widowed OR 0.61; 95% CI 0.44-0.85); public insurance (OR 0.84; 95% CI 0.71-0.98), self-pay (OR 0.59; 95% CI 0.39-0.96); orders placed > 8 months after index examination (OR 0.20; 95%CI 0.14-0.29); ordering by non-primary care (PCP) (OR 0.51; 95%CI 0.36-0.72); and non-same day scheduling. Socioeconomic disadvantage (ADI) was not associated with incomplete BI-RADS 3 follow-up. Conclusion: Non-completion of recommended BI-RADS 3 follow-up is associated with Asian race, age< 40 years, MRI or ultrasound (versus mammography), marital status, insurance coverage, delayed order placement or scheduling, and order placement by non-PCPs.
What problem does this paper attempt to address?