Fig 1.
(A) Percent of stage IIIA NSCLC patients receiving hfIGRT in free-standing and hospital based practices by year. There was a trend toward increased utilization among free-standing radiation oncology facilities, but this did not reach significance (Fig 1a). (B) Percent of stage IIIA NSCLC patients who received hfIGRT in regions with permissive LCDs, restrictive LCDs, or no LCDs governing IMRT use. Although IMRT is distinct from hfIGRT, the two are highly correlated. Restrictive LCDs were associated with a significant reduction if hfIGRT use (p < 0.01). Abbreviations: NSCLC = non-small cell lung cancer, hfIGRT = high frequency image guided radiation therapy, LCD = local coverage determination, IMRT = intensity-modulated radiation therapy.
Table 1.
Logistic regression using state for location.
Fig 2.
Cumulative incidence of acute toxicities calculated via claims for hfIGRT (dashed line) and other (solid line).
No statistically significant difference was observed for any of the endpoints except esophageal toxicity. Although hfIGRT was associated with increased esophageal toxicity among all patients, this association did not withstand an analysis using propensity score matching. Abbreviations: hfIGRT = high frequency image guided radiation therapy.
Table 2.
Association of hfIGRT with toxicities and survival, grouped by analytic technique.
Fig 3.
Percentage increase in mean radiation-related treatment cost.
Costs include all Medicare reimbursements to providers for radiation therapy-related claims from 15 days before diagnosis to 6 months after diagnosis, are normalized to 2017 dollars, and are adjusted for geographic variation in Medicare reimbursement. Abbreviations: hfIGRT = high frequency image guided radiation therapy.