Social determinants of health shape treatment course and long term outcomes in patients with high-risk myeloproliferative neoplasms

Abstract Introduction Myeloproliferative neoplasms (MPNs) such as Essential Thrombocythemia (ET) and Polycythemia Vera (PV) are chronic hematological neoplasms that have a long disease course and clinical trajectory and carry a high risk of thromboembolism and hemorrhagic complications. Treatment aims to prevent vascular complications such as stroke, heart attack and blood clots that contribute to frailty and mortality. The extended disease course of MPNs allows for the investigation of the impact of social risk factors on frailty, long-term outcomes, and health care utilization patterns. Therefore, we conducted a retrospective cohort analysis to determine the impact of social risk factors on treatment and disease course in patients with high-risk MPNs. Methods Patients with high-risk ET or PV were identified through the UTHSC Research Enterprise Data Warehouse (rEDW), a multi-institutional data collaborative with great socioeconomic and geographic diversity. Inclusion criteria were patients aged 18-70 diagnosed with ET or PV between 2014 and 2019 with a documented thrombotic event. Healthcare utilization patterns, follow-ups, and clot burden were defined using hospital visits and their associated ICD-9/10 code classifications. Census tract data were obtained from PolicyMap (www.policymap.com) based on the location of residence and were used to assess social determinants of variability in outcomes and disease course. A modified deficit accumulation frailty index (DAFI) score was calculated using 18 common frailty-associated conditions and was compared among the cohort. Demographic and clinical variables were compared using Student's t-tests and multivariable logistic regressions. Results A total of 870 patients with high-risk ET or PV (mean age = 51.8 years) were included in the analysis. The cohort was ethnically diverse, consisting of 530 (62.4%) White and 289 (33.9%) African-American patients. In terms of outcomes, race, median household income, and rurality did not impact the age of diagnosis. Those in medically underserved areas (MUAs) were diagnosed, on average, 2.6 years later than those who were not (p = 0.0089). There were no significant differences in the setting of diagnosis (inpatient vs outpatient) between these demographic groups. Upon assessing whether social risk factors impacted morbidity and outcomes, poverty rate was found to be associated with an increased risk of MI and stroke, with an odds ratio of 1.019 and 1.025 per percent increase in poverty rate, respectively (p = 0.001; p = 0.004). Higher insurance coverage at the community level was associated with improved overall survival, with each 1% increase in census tract insurance coverage corresponding to a 5% reduction in the odds of death (OR 0.95; 95% CI, 0.91–0.99; p = 0.0087). Increased frailty was observed in males and those living in census tracts with a lower median household income (β = -0.042, p < 0.001; β = -5.18 × 10⁻⁷, p = 0.028). Unexpectedly, patients residing in non-MUAs experienced higher frailty scores, which could be attributed to cumulative disease burden or survivorship effects (β = 0.033, p = 0.007). Additionally, we identified disparities in healthcare utilization within the cohort. The number of clinical site visits after the initial diagnosis was 32.4 in African Americans versus 15.5 in White patients (p < 0.0001). Similarly, the number of hospitalizations in African Americans was significantly higher than in White patients (3.36 vs 1.93; p < 0.0001). Conclusions In this retrospective cohort study of patients with high-risk MPNs, we demonstrate that social risk factors significantly influenced both the timing of diagnosis and disease trajectory. Delayed diagnosis in underserved areas indicates that potential disparities exist that create barriers in accessing to early care. Notably, African-American patients had significantly higher healthcare utilization, indicating possible differences in disease monitoring and morbidity. Although frailty was more associated with male gender and lower income, patients in non-medically underserved areas had higher frailty scores, suggesting complexity in the factors determining disease burden and access to care that warrants further investigation. These findings highlight the need to incorporate social determinants of health into risk stratification and management models for MPNs and to address systemic barriers to equitable care in hematologic malignancies.