Effect of Geospatial and Sociodemographic Factors on Live Birth After Fertility-Sparing Treatment for Breast and Gynecologic Cancers.

OBJECTIVE

To evaluate whether geospatial or sociodemographic characteristics are associated with live birth after fertility-sparing treatment for early-stage gynecologic cancer and stages I-III breast cancer.

STUDY

DESIGN

Retrospective matched case-control study using linked data from the California Cancer Registry, the California Office of Statewide Health Planning and Development (now known as the California Department of Health Care Access and Information), and Society for Assisted Reproductive Technology. We included patients aged 18-45 years who were diagnosed with stages I-III of breast cancer or stage I cervical, endometrial, or ovarian cancer between 2000 and 2012 and underwent fertility-sparing treatment. Patients in the case group had live births after treatment; patients in the control group did not. Propensity score matching was performed in a 1:2 ratio. Generalized linear mixed models were used to estimate associations between odds of live birth and geospatial and sociodemographic exposures such as the California Healthy Places Index, clinic proximity, race and ethnicity, marital status, insurance, and socioeconomic status.

RESULTS

Our study included 254 individuals in the case group and 455 individuals in the matched control group. On multivariable analysis, no significant associations were observed between geospatial factors and live birth. Compared with individuals living in ZIP codes within the lowest California Healthy Places Index quartile (Q, 0-25%), those in higher quartiles had similar odds of live birth (Q2: odds ratio [OR] 1.16; 95% CI, 0.45-2.96, P=.76; Q3: OR 1.29; 95% CI, 0.46-3.59, P=.62; Q4 or unknown: OR 1.19; 95% CI, 0.39-3.62, P=.75). Neither the number of nor the distance to the nearest gynecologic oncology or in vitro fertilization clinic was associated with outcome. Sociodemographic characteristics also were not significantly associated with odds of live birth. Subanalyses by treatment type and cancer type also demonstrated no significant associations variables of interest and odds of live birth.

CONCLUSION

Among patients who accessed fertility-sparing treatment, geospatial and sociodemographic factors were not associated with live-birth outcomes. These findings provide reassurance that once patients access fertility-sparing care, birth outcomes are not influenced by sociodemographic or geographic disadvantage.