End-stage kidney disease (ESKD) is a grave public health problem in the US affecting over 808,000 Americans in 2017. Options for treatment of ESKD include renal transplantation and peritoneal dialysis. However, the most common treatment for ESKD in the US is hemodialysis, with over 468,000 Americans being hemodialysis dependent in 2017. For hemodialysis-dependent patients, vascular access is required, for which there are two options that are considered permanent, typically created in the upper extremity: 1) fistula, created by surgically connecting a patient’s own vein to an artery, and 2) graft, using a prosthetic tube as a conduit between artery and vein. However, evidence to support the choice of fistula vs graft in an individual patient is lacking. Frailty is a measure of patient global functional status and overall health that could be used in the decision-making progress to determine the optimal vascular access type for an individual patient. The objective of this dissertation is to investigate the association of frailty with probability of short-term vascular access failure to assess the usefulness of incorporating measurement of frailty into the vascular access decision-making process.I performed a quantitative analysis of the United States Renal Data Systems Registry which includes all hemodialysis dependent patients in the US. I determined the presence of frailty for each patient using a validated claims-based frailty indicator. I then assessed the association of frailty with time to fistula and graft functional use (the process of becoming usable for hemodialysis) using multivariable models. The analysis demonstrated that patients with increasing claims-based frailty indicator values have increasing length of time to fistula functional use. There was no association of claims-based frailty indicator value with time to graft functional use. The association between frailty and longer fistula functional use times has significant implications for clinical decision-making and guideline creation. Given that this study suggests frailty could also be associated with longer time to fistula functional use, frailty could serve as a strong indicator that fistula is not an appropriate choice of vascular access for that patient. Further, since increasing degrees of frailty, as measured by the claims-based frailty indicator, are associated with the highest risk of longer time to fistula m functional use in this study, it may be appropriate to consider patients with the highest levels of frailty as candidates for permanent access through a tunneled catheter.

The population of cancer survivors in the United States is currently estimated to be close to 14 million. Cancer survivors are at risk for a variety of physical and mental health deficits related to their disease and treatments. Most cancer survivors receive excellent care during their active cancer treatment (surgery, radiation, chemotherapy) but can feel lost and overwhelmed when entering the post-treatment phase of care. Research has shown that cancer survivors do not always receive optimal post-treatment care and preventative services.

This dissertation consists of three studies that explore issues in cancer survivorship care: 1) the use of guideline-recommended post-treatment health services in breast cancer survivors; 2) provider expectations and perceptions of breast cancer post-treatment care delivery; and 3) the prevalence of post-traumatic stress symptoms in cancer survivors and associated risk factors. This research used multiple data sources to explore these issues, including health insurance claims data, medical record abstraction, qualitative interview data, and patient-reported outcomes. Several methodologies were employed to analyze these data, including qualitative data analysis, multi-level modeling, elaboration models, multivariate linear and logistic regression models, and Kaplan-Meier estimates.

This research found that there are persistent gaps in survivorship care delivery. First, breast cancer survivors treated and followed at an academic medical center did not consistently receive guideline-recommended post-treatment care, and patients in this sample commonly received non-recommended care that has not been shown to provide benefit and could potentially be harmful. Second, oncology and primary care providers perceive many barriers to providing high-quality care in the post-treatment care period, most importantly lack of care coordination within oncology and across specialties. Finally, a sub-set of cancer survivors followed in a clinical survivorship program was found to have persistent post-traumatic stress symptoms.

These results demonstrate the importance of organized survivorship care delivery programs that would ensure high-quality patient care for this unique population. Findings suggest that post-treatment care delivery is a complex, multi-level process with many potential targets for improvement in quality of care. These studies demonstrate the pressing need for improving survivorship care coordination in order to deliver guideline concordant care. In addition, the persistent psychological effects of cancer and cancer treatment require continued research into effectively identifying and treating those at risk for ongoing distress. Future research should focus on multi-level interventions targeting system-, provider-, and patient-level factors.