The Area Deprivation Index Identifies Survival Disparities in Osteosarcoma When Other Social Determinants of Health...

Large database studies have demonstrated an association between measures of affluence and survival outcomes for patients with osteosarcoma. However, some recent reports have suggested that these disparities may not be as present in the pediatric population. It is unknown which measure of social determinants of health (SDOH) best captures outcome disparities, such as geospatial measures like the Area Deprivation Index (ADI) or patient-level measures such as insurance type or poverty rates, for osteosarcoma.

(1) Are different measures of SDOH associated with differences in the presentation of osteosarcoma, (2) differences in the initiation and completion of treatment, and (3) differences in survival outcomes?

This was a single-institution, retrospective, comparative study of 88 children, adolescents, and young adults diagnosed with high-grade osteosarcoma between 2010 and 2024. The ADI and other measures of SDOH previously investigated by other authors, including the Childhood Opportunity Index (COI), insurance type, US Census Tract poverty level, and self-reported race or ethnicity, were determined for each patient. The ADI scores were grouped into low (< 75th percentile) and high (≥ 75th percentile) deprivation groups, and the COI was classified as low (≤ 25th percentile) and high (> 25th percentile) opportunity, based on prior National Cancer Database (NCDB) and Surveillance, Epidemiology, and End Results (SEER) studies, which showed worse outcomes in patients in the bottom quartile of income. Data on patient age, gender, primary site, presence of metastasis, treatment including chemotherapeutic regimen and surgical treatment, chemotherapeutic response, and outcomes (including overall survival, metastasis-free survival, local recurrence-free survival, and relapse-free survival) were gathered from the medical record. Whether the patient presented to an acute care setting (urgent care or emergency department) or outpatient clinic was recorded. Patients with Medicaid and those with no insurance were compared with those with commercial insurance. Symptom duration before diagnosis and treatment timelines were compared against various measures of SDOH. Survival outcomes were also compared among SDOH groups. A multivariable Cox regression analysis including ADI group, tumor size, and presence of metastatic disease was performed to look for factors independently associated with survival.

Patients in the high deprivation group presented to an acute care setting more frequently than the low deprivation group (51% [20 of 39] versus 22% [10 of 45], OR 1.9 [95% confidence interval (CI) 1.2 to 2.9]; p = 0.006) and had greater median (IQR) symptom duration (13 weeks [6 to 22] versus 6 weeks [4 to 11]; p = 0.002). The COI, US Census Tract poverty level, or race and ethnicity groups did not differ in symptom duration or presentation. The time to begin and complete treatment after diagnosis did not differ among the ADI, COI, US Census Tract poverty level, or race and ethnicity groups. The total median (IQR) treatment time was greater in those with Medicaid and patients without insurance compared with those with commercial insurance who completed planned chemotherapy (131% [122% to 154%] versus 122% [113% to 132%]; p = 0.01). Overall survival in the high deprivation group was worse compared with the low deprivation group (5-year overall survival 37% [95% CI 21% to 53%] versus 65% [95% CI 49% to 78%]; p = 0.04) as was local recurrence-free survival in the high deprivation group (5-year local recurrence-free survival 81% [95% CI 63% to 91%] versus 96% [95% CI 76% to 99%]; p = 0.03). A multivariable Cox regression survival analysis including ADI group, tumor size, and metastatic disease at presentation found that ADI group (HR 2.0 [95% CI 1.1 to 3.9]) and metastatic disease at presentation (HR 3.2 [95% CI 1.7 to 6.4]) were independently associated with an increased risk of death. There was no difference in survival when stratified by COI, US Census Tract poverty level, insurance type, or race and ethnicity.

Stratifying pediatric and young adult patients with osteosarcoma by the ADI, which heavily favors income, identified patients with worse survival outcomes and agrees with large NCDB and SEER database studies showing survival disparities among income groups. Importantly, we show how stratifying patient groups by differing SDOH measures can give conflicting results and highlight the importance of careful SDOH measure selection. The ADI, or other measures that heavily weigh income, should be used when examining disparities that are found among income groups in the NCDB or SEER studies. Large database studies have demonstrated disparities in many orthopaedic diseases and procedures, and investigators who wish to study these further at an institutional level should choose a concordant measure of SDOH.

Level III, prognostic study.