(P122) Streamlining Referring Physicians Orders With ‘Reflex Testing’ Significantly Decreases Time to Resolution for Abnormal Screening Mammograms

Meena S. Moran, MD, David Styren, BS, Joseph Colasanto, MD, Donna Slonski, RN, Stacy Spooner, MD; Yale University School of Medicine; Therapeutic Radiology, Backus Breast Center

Introduction: Patients with abnormal screening mammogram (abSM) often experience prolonged wait times for additional testing to resolve the cause of the abnormality. The purpose of this study was to initiate an intervention, termed ‘reflex testing’ (RefT), and analyze pre- and post-RefT timelines to determine if, where, and in whom RefT resulted in significant decreases in timelines for resolution of abSM of patients undergoing routine screening in an accredited community hospital-based breast center.

Methods: All patients with abSM (Breast Imaging Reporting and Data System [BIRADS] 0, IV, and V) from the diagnostic sites of our breast center program were eligible for analysis. The two cohorts consisted of women with abSM from January to June 2009 (preintervention) and January to June 2012 (postintervention), with timelines created for each patient from the date of abSM to date of: additional diagnostic tests (mammogram, ultrasound, magnetic resonance imaging [MRI]), upgrade to higher BIRADS score, biopsy, or resolution (BIRADS I/II). Based on analysis of the timelines, RefT was implemented to streamline physicians’ written orders as follows: All of the medical staff/referring physicians were notified that, as part of the initial written order, every abSM order would automatically result in additional diagnostic testing and/or biopsy reflexively (as deemed appropriate by the reading radiologist); thus, additional diagnostic test/biopsy-specific written orders would not be required by the referring physicians. Each abSM patient was contacted within 24 hours following the abnormal reading by the breast nurse navigation team to facilitate scheduling for additional testing. Data were collected from the hospital’s Meditech/PenRad electronic medical records. SPSS V.19 and Excel were utilized for statistical analysis.

Results: Analysis of the preintervention abSM timelines revealed that significant delays (defined as > 1 month) occurred only in patients starting with a BIRADS 0 assignment, and thus, the two cohorts that were analyzed consisted of 1,523 patients initially designated BIRADS 0 (n_pre-RefT = 647; n_post-RefT = 876). Comparison of the two cohorts demonstrated that the overall mean interval from abSM to return for additional testing decreased from 23.5 days pre-RefT (median 20, standard deviation [SD]: 21.825, range: 0–256 d) to 8.2 days post-RefT (median 7, SD: 6.987, range: 0–99 d) (P < .001). For patients not requiring a biopsy (n = 1,190), the RefT decreased the mean time from abSM to first diagnostic test/resolution from 29.7 days to 10.77 days (P < .010). For the 333 patients requiring biopsy (n_pre-RefT = 138; n_post-RefT = 195), RefT significantly decreased the mean time from abSM to first diagnostic test from 31.4 days (n = 138,) to 7.7 days (P < .001) and the interval from first diagnostic test to biopsy from 21 days to 17 days (P < .013). There was no difference in the median number of days from biopsy to pathology report (2.44 vs 2.30; P > .05) for pre- vs post-RF.

Conclusions: Reflex testing streamlines the resolution of abSM by allowing for supplementary diagnostic testing and/or biopsy reflexively (by the reading radiologist) without additional written orders from referring physicians. We have demonstrated that this simple initiative significantly decreases patient timelines to resolution of abSM and benefits both subsets of patients not requiring and those requiring biopsies. Implementation of this intervention should be considered in breast center programs wishing to further decrease scheduling times for resolution of abSMs.