S. Fehrs, B. Carr, K. Luca, B. Ghavidel, and M. Axente; Department of Radiation Oncology, Winship Cancer Institute of Emory University, Atlanta, GA
Purpose/Objective(s): To balance clinic load by triaging incoming patients between O-ring and C-arm machines based on treatment site, patient characteristics, and treatment platforms limitations. Materials/
Methods: 3 months of treatment records were sampled (231 patients) to detail clinic operating capacity for each used treatment machine (2 C-arm, 1 O-ring). Patients were categorized based on treatment site: Brain/HN (62), Breast (42), Extremities (8), Thorax (21), Abdomen (21), and Pelvis (77). Physics and dosimetry team members compiled treatment planning, machine, and patient specific limitations that could inform triaging rules. These included: C-arm to service all extremities, breast cases, hypo fractionated courses (SRS/SBRT), and treatments needing gating or couch rotations (O-ring lacks features). O-ring to service all standard fractionation Brain/HN patients that do not need couch kicks, and pelvis patients with <35cm lateral (LAT) separation (anecdotal cutoff). All other pelvis, thorax, abdomen, and brain sites are to be treated on the C-arm machines. These rules led to C-arm schedule overload, treating on average 82 patients per machine, vs. 67 for O-ring. To detect patient characteristics that may affect planning decisions, thus informing triage, a 14-patient study compared 10MV prostate plans on C-arm vs. 6MV FFF O-ring. For all thorax, abdomen, and pelvis categories (119 patients), the LAT and anterior-posterior (AP) separations at the treated isocenter were measured. Results: Separation data was not normally distributed (Shapiro-Wilk p<0.05). AP separations were not significantly different across categories (Kruskal-Wallis p>0.05). A post hoc Mann-Whitney test, showed that only LAT abdomen separation was significantly different (median±std: 32cm±5.2cm), while the rest were not (combined thorax and pelvis 35.8cm±4.6cm). All compared prostate plans were clinically equivalent. One significant difference, the 30% isodose line average volume conformity index (p<0.05), was highlighted between C-arm (8.4) and O-ring (9.9) plans. This difference was independent of any separations. Based on analysis, updated triage rules maintained prior definitions while imposing the median LAT separation value as cutoff for pelvis, thorax, and abdomen sites: assign C-arm if LAT separations >36cm (>32cm abdominal), O-ring otherwise. Application of triage rules increased the potential patient load for the O-ring machine (Table 1). Conclusion: Applying local triage rules has the potential to balance treatment load on each machine. By implementing standardized decision trees, we can facilitate operation efficiency thus increasing quality of patient facing flexibility scheduling and consistency.
