To our knowledge, this is the first study investigating the impact of SVI on post-transplant outcomes for patients with MM. This study determined that patients living in high SVI counties who underwent ASCT had significantly lower odds of PFS and OS as compared to patients living in low SVI counties. Our SVI subanalyses further identified that the socioeconomic status, household characteristics, and racial and ethnic minority status themes were most strongly associated with poorer odds of PFS. However, each individual theme’s association with PFS was less pronounced than composite SVI, which suggests that the variables comprising SVI have a synergistic effect.

These results are consistent with literature studying individual social determinants of health and their associations with MM outcomes. In the present study, Non-Hispanic Whites had a significantly lower odds of PFS and OS in comparison to other racial/ethnic groups (Black, Asian, other). Saraf et al. conducted a retrospective analysis investigating outcomes of MM patients post-ASCT and demonstrated that Black patients had a significantly increased median event-free survival compared to non-Black patients (21 vs 12 months, p = 0.02) [14]. Ailawadhi et al. similarly found that Black patients had significantly longer median multiple MM specific survival (MSS) in comparison to Whites (5.4 years vs 4.5 years, respectively; p < 0.05) and comparable median MSS for Hispanics and Whites (4.9 vs 4.5 years, respectively; p = 0.41) [15]. More recently, Dong et al. found that Black patients had significantly longer 5-year OS than Non-Hispanic White patients, when they were treated similarly (absolute difference =3.8%, p = 0.003) [16]. It has been thought that this improved survival in Black patients with MM may be due to more biologically indolent disease subtypes present in this population [17].

Our study found that patients living in urban areas had a significantly higher odds of PFS compared to patients from rural areas, even in the modern era, consistent with previous studies [18, 19]. Our cancer center has a uniquely high rural catchment area, allowing us to study rural/urban differences. Previously, Rao et al. concluded that patients from rural areas who received ASCT had a higher relative risk of death (RR = 1.18, p = 0.016) compared to patients from urban areas [18]. Survival at 1 year (73% vs 78%, p < 0.04) and 5 years (48% vs 54%; p = 0.12) were also lower for patients from rural areas versus urban areas, respectively [18]. This disparate outcome is believed to be due in part to rural patients having to travel greater distances to access quality health care, delay in seeking treatments, and modifiable risk factors such as tobacco use and obesity [20, 21].

The ATSDR in conjunction with the CDC, created the CDC/ATSDR SVI to assist public health officials and emergency response planners identify communities that will most likely need support, during, and after a hazardous event [6]. However, SVI is being increasingly utilized in medical health outcomes research, particularly in the field of surgery [22]. For example, Azap et al. demonstrated that patients with high SVI had decreased odds (OR = 0.89, 95% CI 0.82–0.97) of achieving superior outcomes after pancreatic surgery [23]. Hyer et al. similarly reported that patients from high SVI areas undergoing common oncologic surgical procedures had a higher probability of 90-day mortality [24]. As a result, SVI is becoming increasingly recognized as helpful in government allocation of resources to cancer patients in hopes of reducing excess cancer morbidity and mortality in specific communities [25].

SVI is not the only measure of community/neighborhood level vulnerability as other indices have been studied in medical health outcomes research [26,27,28,29,30]. Carmichael and colleagues (2020) determined that SVI performed similarly to other indices of neighborhood vulnerability (Area Deprivation Index, Community Needs Index (CNI), and Distressed Communities Index (DCI)) in predicting emergent surgical presentations but had several advantages [31]. Unlike other measures of neighborhood vulnerability, SVI is maintained by the CDC/ATSDR (as opposed to a private organization or university) and is frequently updated. SVI is additionally available at the census tract level, offering a more granular view of a patient’s social vulnerability. In contrast, CNI and DCI capture data at the ZIP code level.

The current study is additionally strengthened by the analysis of the four subthemes of SVI and their impact on a patient’s probability of achieving PFS and OS post-ASCT. We demonstrated that patients living in areas with higher vulnerability scores in the socioeconomic status, household characteristics, and racial and ethnic minority status subthemes had significantly lower odds of PFS post-ASCT. Higher vulnerability values in the socioeconomic and household characteristics subthemes were also associated with lower OS post-ASCT. Interestingly, increases in the housing type and transportation subtheme (which encompasses transportation, housing structure, etc.) were not associated with lower odds of PFS or OS. This may be due to our institution’s ability to provide subsidized housing and transportation.

While data are scarce investigating social vulnerability and outcomes following HCT, Bhandari et al. studied outcomes in patients who underwent allogeneic HCT [10]. Their group similarly found that the highest values in the socioeconomic status, household characteristics, and racial and ethnic minority status subthemes were associated with a higher odds of 1-year non-relapse mortality [10].

Our study has several limitations. We performed a single-institution retrospective study investigating patients who live in the Commonwealth of Virginia. While our findings may be unique to the demographics of Virginia, Virginia demographics are similar to the US (Virginia 60% White, 19% Black, 11% Hispanic, 7% Asian and US 62% White, 12% Black, 19% Hispanic, 6% Asian) [32]. An important feature of our Cancer Center is to increase access to cancer care for rural populations; hence, 35% of our patients were from rural areas, including Appalachia. Appalachia is associated with a rural population, and a high prevalence of poverty, obesity, and substance use, which may explain the high SVI seen in rural areas in our study [33]. In contrast, urban areas in our study include wealthy suburbs of Washington D.C., which likely accounts for the lower SVI in these counties. Also, our SVI scores were relative values at the state level and thus, these values will change when compared to counties across states. An additional limitation is our limited sample size and study population consisting of predominantly Non-Hispanic White patients. Furthermore, since SVI was studied at the county level it may not reflect patient-specific risks. Lastly, because SVI is a composite score, it can be difficult to discern which of the discrete individual factors have the most profound effect on one’s PFS or OS. However, our study serves as a launching point for future investigations in answering this as we have shown which of the 4 SVI subthemes have a significant association with PFS and OS post-ASCT.

In conclusion, this study presents important results in the field of HCT. Our institution utilized county-level SVI as a comprehensive measure of patients’ vulnerability and determined that patients living in areas of high social vulnerability have worse odds of PFS and OS post-ASCT. The subtheme analysis also reveals key variables that future studies can highlight. Specifically, the study results illustrated that the socioeconomic status, household characteristics, and racial and ethnic minority status themes were significantly associated with lower odds of PFS. Therefore, further research will be focused on these subthemes and the variables that make up these groups. For example, for patients living in areas with higher household vulnerability, we would plan to investigate the impact of assistance with family/caretaker support in hopes that this may improve survival outcomes. This study highlights the importance of social vulnerability in post-transplant outcomes and provides a framework for future interventions to improve post-transplant outcomes.