Shifts in Medicare spending for patients with cancer undergoing chemotherapy following implementation of the Maryland Global Budget Revenue program.

Abstract

11079 Background: In January 2014, the statewide Maryland Global Budget Revenue (GBR) model was implemented to control the growth in total hospital spending and improve the quality of care. Specific impacts of GBR on cancer-related spending are not fully understood. This study aimed to 1) quantify the impact of GBR on Medicare spending for beneficiaries undergoing chemotherapy for cancer and 2) investigate any shift in such spending during GBR to non-hospital settings. Methods: Using 2011-2018 Medicare claims from Maryland and a control set of 11 comparable states, we constructed 6-month chemotherapy episodes. Propensity matching was used to identify appropriate comparison episodes based on treatment year, patient demographic, clinical, and area-level characteristics. Using a difference-in-differences (DiD) approach, we evaluated the impact of GBR on standardized total Medicare payments and non-hospital professional payments during the episode, after confirming the parallel trends assumption during GBR’s pre-period (2011-2013). Results: Among 42,206 and 708,486 chemotherapy episodes in Maryland and control states, respectively, we studied 42,199 episodes in Maryland matched to 42,199 episodes in control states. Our analysis showed that GBR’s implementation led to smaller increases in total episode payments over time relative to control states and larger increases in non-hospital professional payments (Table); these impacts notably varied with time. Conclusions: Our finding of smaller increases in total Medicare payment for a 6-month chemotherapy episode in Maryland versus control states indicates that GBR’s intended reductions on the spending growth in the context of cancer patients undergoing chemotherapy were actualized. Importantly, larger increases in non-hospital professional payments suggest these savings may have been attained via shifts in sites-of-care following GBR’s implementation. Further studies evaluating the effects of these shifts on cancer care quality are warranted. Adjusted mean standardized payments in 2018 dollars during 6-month chemotherapy episodes and DiD estimates versus 2013. Total payments Non-hospital professional payments Year Adjusted Mean, Maryland Adjusted Mean, Control states DiD Adjusted Mean, Maryland Adjusted Mean, Control states DiD 2013 $54,213 $52,659 Ref $22,876 $14,844 Ref 2014 $54,522 $53,476 -$759 $21,261 $14,327 $828 2015 $55,355 $55,426 -$2,001 $22,249 $14,186 $1,433* 2016 $58,599 $59,331 -$2,940* $24,314 $14,922 $1,286 2017 $61,141 $60,462 -$1,561 $27,954 $15,330 $3,612* 2018 $62,329 $64,756 -$4,853* $27,969 $15,356 $2,753* *p<0.05. The adjustment methodology accounted for patient demographic, clinical and area-level characteristics, and time-varying Hospital Service Area (HSA) level variables. For the DiD analysis, HSA fixed effects were also included.