-Author name in bold denotes the presenting author
-Asterisk * with author name denotes a Non-ASH member
Clinically Relevant Abstract denotes an abstract that is clinically relevant.

PhD Trainee denotes that this is a recommended PHD Trainee Session.

Ticketed Session denotes that this is a ticketed session.

307 Hospitalization Patterns and Medicare Spending for Non-Malignant Morbidity Among Older Survivors of Non-Hodgkin Lymphoma (NHL)

Program: Oral and Poster Abstracts
Type: Oral
Session: 902. Health Services Research—Malignant Conditions (Lymphoid Disease) I
Hematology Disease Topics & Pathways:
survivorship, Diseases, Elderly, Non-Hodgkin Lymphoma, Lymphoid Malignancies, Study Population, Quality Improvement
Saturday, December 5, 2020: 2:15 PM

Kelly Kenzik, PhD1*, Gaurav Goyal, MD2, Amitkumar Mehta, MD3 and Smita Bhatia, MD, MPH1

1Institute for Cancer Outcomes and Survivorship, University of Alabama at Birmingham, Birmingham, AL
2University of Alabama at Birmingham, Birmingham, AL
3University of Alabama School of Medicine, Birmingham, AL

Introduction: Growth in the number of older cancer survivors in the face of projected healthcare workforce shortages is challenging the US health-care system in delivering appropriate care. A risk-stratified approach based on burden of morbidity and healthcare expertise is urgently needed to provide efficient and cost-effective care. We examined the cumulative, non-malignant, condition-specific hospitalization rates and associated spending in older NHL survivors compared to non-cancer controls, using a population-based approach, to develop evidence for risk-stratified care.

Methods: Using SEER-Medicare data, we identified 14,533 patients diagnosed with NHL at age >65y (2008-2015). An age, race, sex, and follow-up time comparable non-cancer cohort (n=14,533) was also identified. Hospitalizations for 10 health conditions were used to examine cumulative utilization burden of non-malignant morbidity by time from diagnosis to 5y or date of bone marrow transplant (whichever came first), censoring at 6 mo prior to death or end-of-study (12/31/2016). We estimated the average number of hospitalizations per 100 individuals, up to 1y and 5y, accounting for the competing risk of leaving the cohort (BMT or 6 months prior to death). We calculated incident rate ratios [IRR] per person-time comparing NHL patients to controls controlling for pre-cancer comorbidity, Dual Medicaid-Medicare coverage, race/ethnicity, and pre-NHL hospitalization. We calculated Medicare spending per hospitalization and per person-year (adjusted to 2016 pricing), defining high hospitalization or high spending using the 90th percentile cutpoint for predicted hospitalization or spending.

Results: Hospitalization: In Fig 1, we summarize the average number of condition-specific hospitalizations/100 individuals by 1y and by 5y. Adjusting for demographics and pre-existing comorbidity, NHL patients were significantly more likely to be hospitalized when compared with controls (IRR1y=3.08, IRR5y=1.54, all p<0.001). Compared with controls, NHL patients were at higher risk for hospitalizations related to cardiovascular disease (IRR1y=3.08, IRR5y=1.54, all p<0.001), age-related conditions (IRR1y=7.14, IRR5y=2.53, all p<0.001), gastrointestinal (GI) disease (IRR1y=5.84, IRR5y=2.30, all p<0.001), and pulmonary conditions (IRR1y=3.06, IRR5y=1.36, p<0.001). Factors associated with high hospitalization rates included a diagnosis of diffuse large B cell lymphoma (RR=6.36, p<0.001) and ≥2 pre-existing comorbidities (RR=10.1, p<0.001), non-white race/ethnicity (Black: RR=1.74, p<0.001; Hispanic (RR=1.97, p<0.001), residence in low-education area (RR=1.49, p<0.001) or low-income area (RR=37, p<0.001), or rural residence (RR=1.38, p<0.001). The high-hospitalization subgroup (n=1,435) had 3.24 times the hospitalization rate when compared with the lowest-hospitalization subgroup (p<0.001). Spending: Average hospitalization spending per person-year was higher in the NHL cohort (Y1: $8,178 vs. $1,621; 5y: $3,805 vs. $1,543) when compared with non-cancer controls (Fig 2). The average spending per hospitalization over 5y was also higher for the NHL cohort ($16,950 vs. $13,474, Fig 3). Pulmonary conditions were associated with the largest per hospitalization spending differential between NHL and controls ($18k vs. $9k). Other per hospitalization spending disparities between NHL patients and controls included GI disease ($17.8k vs. $10.2k) and age-related conditions ($12.6k vs. $7.9k). The NHL high-spending group was similar to the high-hospitalization group (71% overlap). The highest spending group had 2.5 times the spending compared with the lowest spending (10th percentile) group (p<0.001).

Conclusions: Medicare beneficiaries diagnosed with NHL experience significantly greater rates of non-malignant hospitalizations and spending compared to their non-cancer counterparts. These hospitalizations are related to cardiovascular disease, gastrointestinal disease, pulmonary complications and aging-related conditions. Non-white DLBCL patients from lower SES with ≥2 comorbidities and living in rural areas were at greatest risk of high hospitalization and spending, providing evidence for a targeted risk-stratified care.

Disclosures: Mehta: Seattle Genetics: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Roche-Genentech: Research Funding; Merck: Research Funding; TG Therapeutics: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Affimed: Research Funding; Takeda: Research Funding; Incyte: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Innate Pharmaceuticals: Research Funding; fortyseven Inc/Gilead: Research Funding; Juno Parmaceuticals/BMS: Research Funding; Gelgene/BMS: Research Funding; Oncotartis: Research Funding; Kite/Gilead: Research Funding.

*signifies non-member of ASH