Diffusion and Health Impacts of New Breast Radiotherapy Technology
Methods. We used 2005-2008 data from the Chronic Condition Data Warehouse for female Medicare beneficiaries receiving RT after breast-conserving surgery merged with data from AMA, ARF (county), Medicare managed care, and Census tracts. We evaluated RT utilization (n=74254 patient-subjects; n=2291 physicians): 1) logistic regression, clustering on physician, of the probability a patient receives accelerated RT or not, and 2) logistic regression of the proportion of patients for whom the radiation oncologist uses partial breast RT. Using bivariate probit with instrumental variables, we assessed complications of whole breast RT compared with accelerated partial breast RT on a more homogeneous patient population (n=48850).
Results. There has been increased use of accelerated partial breast RT over time from 8% of cases in 2005 to 17% in 2008. Physician-level analysis indicates rural practices are much less likely to perform accelerated RT (reference 0-10% rural: 10-50% rural OR: 0.57;95%CI: 0.45-0.73; 50+% rural OR: 0.40;95%CI: 0.21-0.76); as are those licensed 20+yrs (OR:0.61;95%CI: 0.46-0.80); and physicians in Midwest, Northeast and West Census regions much less likely to use compared to South. High-poverty-area physicians are more likely to use it (>20%poverty: OR: 1.86;95%CI: 1.4-2.4). Medicare managed-care penetration and radiation oncologist density were insignificant. At the patient level, 11.7% received accelerated RT. Treatment after 2005 was associated with increasing odds of receiving accelerated RT (OR: 1.3, 1.8, 2.4 for 2006, 2007, 2008, respectively, p<0.0001). Older age was associated with lower odds of using accelerated RT (reference 66-69yo: 75-79yo and 80+yo OR:0.90, p<0.006), as was Black (OR:0.73;95%CI:0.6-0.8) or Other race (OR: 0.78;95%CI: 0.65-0.99), living in rural areas (OR:0.8, p<0.0001), or seeing an older physician (20+yrs post-graduation (OR:0.7;95%CI:0.5-0.9)). Patients living in counties with 1+ hospital/100K population with advanced RT facilities (IMRT or shaped-beam) were much more likely to undergo accelerated RT (OR:1.4;95%CI:1.1-1.8). Patients in the South were much more likely to have accelerated RT (p<0.009). Comorbidity or area-based poverty and high-school-graduate rates were insignificant. Use of accelerated partial breast RT comes with higher complication rates, with 14% of women having at least one RT complication compared to 9% for whole breast RT. Specific complication rates were predicted significantly higher for patients receiving accelerated partial breast RT (infection/wound complication: 6% v. 3%, p=0.002; seroma: 1% vs. 0.2%, p=0.002; breast pain: 8% vs. 3%, p<0.0001); only fibrosis was higher in whole breast RT (0.2% vs. 0.7%, p<0.05).
Discussion. Although this new technology lacks long-term clinical effectiveness evidence, it is diffusing rapidly and comes with higher complication rates. Populations that might benefit more from accelerated RT as a simpler, intermediate RT modality – rural or older patients – are not tending to undergo the treatment. Higher complication rates must be weighed against convenience of accelerated RT compared to standard RT protocols.