Urban relatives ameliorate survival disparities for genitourinary cancer in rural patients

Abstract Introduction Patients living in rural areas have worse cancer‐specific outcomes. This study examines the effect of family‐based social capital on genitourinary cancer survival. We hypothesized that rural patients with urban relatives have improved survival relative to rural patients without urban family. Methods We examined rural and urban based Utah individuals diagnosed with genitourinary cancers between 1968 and 2018. Familial networks were determined using the Utah Population Database. Patients and relatives were classified as rural or urban based on 2010 rural–urban commuting area codes. Overall survival was analyzed using Cox proportional hazards models. Results We identified 24,746 patients with genitourinary cancer with a median follow‐up of 8.72 years. Rural cancer patients without an urban relative had the worst outcomes with cancer‐specific survival hazard ratios (HRs) at 5 and 10 years of 1.33 (95% CI 1.10–1.62) and 1.46 (95% CI 1.24–1.73), respectively relative to urban patients. Rural patients with urban first‐degree relatives had improved survival with 5‐ and 10‐year survival HRs of 1.21 (95% CI 1.06–1.40) and 1.16 (95% CI 1.03–1.31), respectively. Conclusions Our findings suggest rural patients who have been diagnosed with a genitourinary cancer have improved survival when having relatives in urban centers relative to rural patients without urban relatives. Further research is needed to better understand the mechanisms through which having an urban family member contributes to improved cancer outcomes for rural patients. Better characterization of this affect may help inform policies to reduce urban–rural cancer disparities.


| INTRODUCTION
2][3] Patients in rural areas have been shown to experience a higher incidence of preventable cancers, are more likely to receive a diagnosis at a later stage and have higher cancer-specific mortality than patients living in urban areas. 1,4Several possible explanations have been explored, though none have completely accounted for the breadth of the disparity.
Screening and preventive care rates have been shown to be lower for patients living in rural areas, however rural patients who receive similar screening and treatment to urban patients have similar outcomes. 5][8] Rural cancer patients are also more likely to have lower economic and social capital than urban patients.This can result in barriers to care including financial limitations, transportation difficulties, and lack of insurance coverage. 9tudies have found that social support is an important aspect of cancer care in rural patients. 10The type of social support that is most beneficial to rural patients has not been well explored.In this study, we seek to better understand how rural patients with relatives in urban areas who live nearer to cancer centers and cancer resources may contribute to better outcomes for rural patients.
We leveraged a unique data resource, the Utah Population Database (UPDB), that contains information about cancer, healthcare utilization, and residential histories for both patients and their family members to better characterize the impact of rural patients with urban relatives.We compared cancer outcomes between three groups of cancer patients: (a) rural without urban relatives, (b) rural with urban relatives, and (c) urban.We hypothesized that rural patients who have family members in urban areas have better overall survival (OS) compared to rural patients who lack relatives in urban areas.

| Study design
We performed a retrospective cohort analysis of adults diagnosed with genitourinary cancers, including: prostate, bladder, kidney, penile, and testicular cancers between 1990 and 2018.Due to potential concerns about privacy and misuse of this unique data resource, access to UPDB by researchers requires selection of specific cancer subsets from the entire data.We selected the included genitourinary cancers both because of our research group's expertise and to represent a variety of cancer types that affect different populations, incidences, affected age groups, and mortality risk.

| Data setting
Patients were classified as living in either rural or urban areas based on the 2010 rural-urban commuting area (RUCA) codes associated with their zip code of residence at the time of cancer diagnosis. 11RUCA codes include whole numbers ranging from 1 to 10, with 1 indicating a metropolitan area core with primary flow within an urbanized area and 10 indicating rural areas with primary flow to tracts outside of urbanized areas or urban clusters.Whole integers between these two extremes represent a range of metropolitan, micropolitan, and small towns with various flows.Secondary numbers range from 1.10 to 10.3 and further breakdown whole integer categories by secondary flows to different urbanized areas and urban clusters.If a patient lived in an area classified by RUCA codes 4, 5, 7, 8, 10, 10.3, they were considered rural for this study.
Adult first-and second-degree relatives were identified and classified as urban or rural based on the zip code or county of residence at the time of the patients' diagnosis or, when unavailable, the county or zip code of residence before or after diagnosis.Family members were classified as living in an urban area if their residence at the time of patient diagnosis was classified by RUCA codes 1, 2, 3, 4.1, and 7.1.

| Statistical analysis
All statistical analyses were performed using R software, with a p value <0.05 denoting statistical significance.OS was analyzed using Cox proportional hazards models after adjusting for sex, age, race, cancer type, SEER cancer stage, year of cancer diagnosis, health improvement index (HII), 12 and Simpson's diversity index. 13OS hazard ratios (HR) were then used to calculate the probability of survival for patients grouped as urban, rural without urban relatives, and rural with urban relatives.The results were then plotted as Kaplan-Meier curves.
Sensitivity analyses further examined the robustness of results by exploring whether subjects with larger family networks had a different outcome to account for the potential explanatory effect of having larger families.These were demonstrated in the form of OS.Patients diagnosed with genitourinary cancer living within urban regions described by the RUCA were utilized as the reference group.

| Main outcomes
There were minimal differences between groups when evaluating any number of first-degree relatives versus those with a minimum of five first-degree relatives (Table S1).We also stratified cancers into subgroups, including prostate and non-prostate GU cancer, finding minimal differences between groups (Table S2).

| RESULTS
We identified 22,766 individuals diagnosed with genitourinary cancer between 1990 and 2018 in the UPDB.Median follow-up was 8.78 years.The population predominantly represented men (81%) and white individuals (86.7%).Most patients had prostate cancer, which accounted for 71.9% of all the studied cancers (Table 1).
We found urban patients (reference group) had a significantly higher OS at 5 and 10 years compared to rural patients with and without an urban first-degree relative (Figure 1).Rural patients without an urban first-degree relative had a 41% and 46% greater mortality risk than urban patients, as seen with OS HRs at 5 and 10 years (5year HR 1.41, 95% CI 1.22-1.63,p < 0.001) and (10-year HR 1.46, 95% CI 1.30-1.65,p < 0.001) respectively (Table 2).The significantly higher OS of urban patients compared to rural patients without urban relatives is further supported by higher cancer-specific survival curves at both 5 and 10 year (Figures S1 and S2).

| DISCUSSION
In this study, we confirmed work supported by other researchers that rural cancer patients have worse outcomes.However, we were able to take a deeper dive and show that familial ties also affect this relationship.Having urban first-degree relatives appear to improve survival outcomes for rural patients.
There are several possible explanations for this observation.First, when patients have family members in urban areas, they may have greater family-based social capital that facilitates care in urban centers.Families may be able to provide improved access to care via temporary lodging, transportation or other resources that facilitate care.This may help patients overcome common barriers to accessing high-quality tertiary cancer care in rural patients. 14Second, having a family network near centralized cancer care may permit rural patients to have increased levels of social support, including caregiver support and knowledge about local high-quality resources, needed to achieve better outcomes.Third, rural patients may have access to additional financial resources through family networks living in urban areas similar to immigrant populations who are able to share income with family that is abroad to help them attain a better living standard.
These potential explanations are supported by prior studies highlighting the impact of family and social support on access to care and patient health.Strong family support was shown to improve emotional and physical health in the postoperative period in patients diagnosed with digestive cancers. 15On the other hand, lower levels of family support for young breast cancer patients was significantly associated with cost-related lack of access to care. 16Another study evaluating a collaborative care model to improve social support for disadvantaged patients, found improvement in depression severity when patients received enhanced social support. 17Our study is the first to specifically evaluate the impact of having an urban relative for patients physically living in rural areas.In the context of existing literature, it is a reasonable presumption that enhanced family support paired with enhanced social resources from urban relatives living physically nearer to cancer centers contributes to improved outcomes for rural patients.More research is needed to determine which characteristics of rural patients and urban relatives lead to the greatest improvements in survival outcomes.
Various policies and initiatives have attempted to narrow the gap in outcomes between rural and urban patients in recent years.Awareness of the issue and efforts to improve upon these disparities have recently hastened.The American Society of Clinical Oncology and the National Institute of Health have both targeted this issue with separate campaigns, taking unprecedented steps to address lagging quality of care and outcomes. 18,19The results of our study suggest that policies to provide resources and support to rural patients may have real impacts on cancer outcomes.Possible interventions may include providing expanded insurance access, transportation and lodging, support for caregivers, or cash for cancer related care.Policies may further attempt to simulate familial networks by appointing case workers with access to resources to help patients overcome additional barriers that may arise from traveling in unfamiliar locations.
Our study has five main limitations.First, our data is limited by the retrospective nature of the study.Second, our study population consisted of patients residing in Utah, limiting generalizability to other states and more diverse populations.Third, due to the selection of genitourinary cancer patients, this cohort was predominantly male.Fourth, our database did not contain comorbidity data.Fifth, our database did not contain extensive information which would have allowed us to extrapolate the specific social supports that reduce mortality for rural cancer patients with urban relatives.More research is warranted to include an increased percentage of female patients.Nevertheless, this study represents a novel and unique opportunity to understand the role of family in supporting patients with cancer to improve outcomes.

| CONCLUSIONS
Individuals diagnosed with cancer who live in rural areas have worse survival as compared to their urban counterparts, but this relationship appears to be improved by the presence of family who live in urban areas.Further research is needed to better understand the mechanisms through which having an urban family member may contribute to improved cancer outcomes for rural patients.

1
Overall-specific survival.T A B L E 2 Five-and ten-year overall survival comparing rural patients with and without urban relatives to urban patients.
Baseline patient demographics stratified by rural or urban status.
T A B L E 1

Overall Urban Rural with Urban Relative Rural without Urban Relative p-value
Abbreviation: OS, overall survival.