Evaluating the impact of delayed study startup on accrual in cancer studies.

Background: Drug development in cancer medicine depends on high-quality clinical trials, but these require large investments of time to design, operationalize, and complete; for oncology drugs, this can take 8-10 years. Long timelines are expensive and delay innovative therapies from reaching patients. Delays often arise from study startup, a process that can take 6 months or more. We assessed how study-specific factors affected the study startup duration and the resulting overall success of the study. Method: Data from The University of Kansas Cancer Center (KUCC) were used to analyze studies initiated from 2018 to 2022. Accrual percentage was computed based on the number of enrolled participants and the desired enrollment goal. Accrual success was determined by comparing the percentage of enrollments to predetermined threshold values (50%, 70%, or 90%). Results: Studies that achieve or surpass the 70% activation threshold typically exhibit a median activation time of 140.5 days. In contrast, studies that fall short of the accrual goal tend to have a median activation time of 187 days, demonstrating the shorter median activation times associated with successful studies. Wilcoxon rank-sum test conducted for the study phase (W=13607, p-value=0.001) indicates that late-phase projects took longer to activate compared to early-stage projects. We also conducted the study with 50% and 90% accrual thresholds; our findings remained consistent. Conclusions: Longer activation times are linked to reduced project success, and early-phase studies tend to have higher success than late-phase studies. Therefore, by reducing impediments to the approval process, we can facilitate quicker approvals, increasing the success of studies regardless of phase.

PerSEveML: a web-based tool to identify persistent biomarker structure for rare events using an integrative machine learning approach.

Omics data sets often pose a computational challenge due to their high dimensionality, large size, and non-linear structures. Analyzing these data sets becomes especially daunting in the presence of rare events. Machine learning (ML) methods have gained traction for analyzing rare events, yet there has been limited exploration of bioinformatics tools that integrate ML techniques to comprehend the underlying biology. Expanding upon our previously developed computational framework of an integrative machine learning approach, we introduce PerSEveML, an interactive web-based tool that uses crowd-sourced intelligence to predict rare events and determine feature selection structures. PerSEveML provides a comprehensive overview of the integrative approach through evaluation metrics that help users understand the contribution of individual ML methods to the prediction process. Additionally, PerSEveML calculates entropy and rank scores, which visually organize input features into a persistent structure of selected, unselected, and fluctuating categories that help researchers uncover meaningful hypotheses regarding the underlying biology. We have evaluated PerSEveML on three diverse biologically complex data sets with extremely rare events from small to large scale and have demonstrated its ability to generate valid hypotheses. PerSEveML is available at https://biostats-shinyr.kumc.edu/PerSEveML/ and https://github.com/sreejatadutta/PerSEveML.

Predictors of cytomegalovirus reactivation after allogeneic hematopoietic cell transplantation: Insights from a real-world experience.

BACKGROUND: We aimed to investigate factors associated with cytomegalovirus (CMV) viremia and CMV disease and its impact on post-transplant outcomes including overall survival (OS) following allogeneic hematopoietic stem cell transplantation (Allo-SCT). METHODS: We conducted a single-center retrospective study including 452 Allo-SCT recipients (matched unrelated donor, MUD 61%; haploidentical, haplo 39%) from 2016 to 2021. Data were analyzed using SPSS v28. Descriptive (chi-square and t-test), Kaplan-Meier and regression analyses were conducted. RESULTS: The median age was 57 years. Sixty-one percent were males and 84.3% were Caucasians. CMV serostatus was positive in 59.1% of recipients. The median follow-up was 24.4 months. CMV viremia and CMV disease were observed in 181 (40%) and 32 (7%) patients, respectively. Among CMV seropositive recipients, 65% developed CMV viremia and 11% were noted to have CMV disease compared to 4% and 1% in seronegative recipients, respectively (p < 0.001). Patients with CMV disease had significantly lower OS than those without CMV disease (median 14.1 months vs. not reached, p = 0.024); however, OS was not associated with CMV viremia (median not reached in both groups, p = 0.640). Letermovir prophylaxis was used in 66% (n = 176/267) of CMV seropositive recipients, but no impact was observed on the incidence of CMV viremia or CMV disease and OS. CONCLUSIONS: CMV disease leads to significantly inferior survival after an allogeneic hematopoietic cell transplantation. Recipient CMV seropositive status was associated with the risk of CMV viremia and CMV disease, and this was not abrogated with the use of Letermovir prophylaxis.

Using Bayesian hierarchical modeling for performance evaluation of clinical trial accrual for a cancer center.

Introduction: Slow patient accrual in cancer clinical trials is always a concern. In 2021, the University of Kansas Comprehensive Cancer Center (KUCC), an NCI-designated comprehensive cancer center, implemented the Curated Cancer Clinical Outcomes Database (C3OD) to perform trial feasibility analyses using real-time electronic medical record data. In this study, we proposed a Bayesian hierarchical model to evaluate annual cancer clinical trial accrual performance. Methods: The Bayesian hierarchical model uses Poisson models to describe the accrual performance of individual cancer clinical trials and a hierarchical component to describe the variation in performance across studies. Additionally, this model evaluates the impacts of the C3OD and the COVID-19 pandemic using posterior probabilities across evaluation years. The performance metric is the ratio of the observed accrual rate to the target accrual rate. Results: Posterior medians of the annual accrual performance at the KUCC from 2018 to 2023 are 0.233, 0.246, 0.197, 0.150, 0.254, and 0.340. The COVID-19 pandemic partly explains the drop in performance in 2020 and 2021. The posterior probability that annual accrual performance is better with C3OD in 2023 than pre-pandemic (2019) is 0.935. Conclusions: This study comprehensively evaluates the annual performance of clinical trial accrual at the KUCC, revealing a negative impact of COVID-19 and an ongoing positive impact of C3OD implementation. Two sensitivity analyses further validate the robustness of our model. Evaluating annual accrual performance across clinical trials is essential for a cancer center. The performance evaluation tools described in this paper are highly recommended for monitoring clinical trial accrual.

An R Shiny Application (SDOH) for Predictive Modeling Using Regional Social Determinants of Health Survey Responses.

Social determinants of health (SDoH) surveys are data sets that provide useful health-related information about individuals and communities. This study aims to develop a user-friendly web application that allows clinicians to get a predictive insight into the social needs of their patients before their in-patient visits using SDoH survey data to provide an improved and personalized service. The study used a longitudinal survey that consisted of 108,563 patient responses to 12 questions. Questions were designed to have a binary outcome as the response and the patient's most recent responses for each of these questions were modeled independently by incorporating explanatory variables. Multiple classification and regression techniques were used, including logistic regression, Bayesian generalized linear model, extreme gradient boosting, gradient boosting, neural networks, and random forests. Based on the area under the curve values, gradient boosting models provided the highest precision values. Finally, the models were incorporated into an R Shiny application, enabling users to predict and compare the impact of SDoH on patients' lives. The tool is freely hosted online by the University of Kansas Medical Center's Department of Biostatistics and Data Science. The supporting materials for the application are publicly accessible on GitHub.

Conducting a bayesian multi-armed trial with response adaptive randomization for comparative effectiveness of medications for CSPN.

Background: Response adaptive randomization is popular in adaptive trial designs, but the literature detailing its execution is lacking. These designs are desirable for patients/stakeholders, particularly in comparative effectiveness research, due to the potential benefits including improving participant buy-in by providing more participants with better treatment during the trial. Frequentist approaches have often been used, but adaptive designs naturally fit the Bayesian methodology; it was developed to deal with data as they come in by updating prior information. Methods: PAIN-CONTRoLS was a comparative-effectiveness trial utilizing Bayesian response adaptive randomization to four drugs, nortriptyline, duloxetine, pregabalin, or mexiline, for cryptogenic sensory polyneuropathy (CSPN) patients. The aim was to determine which treatment was most tolerable and effective in reducing pain. Quit and efficacy rates were combined into a utility function to develop a single outcome, which with treatment sample size, drove the adaptive randomization. Prespecified interim analyses allowed the study to stop for early success or update the randomization probabilities to the better-performing treatments. Results: Seven adaptations to the randomization occurred before the trial ended due to reaching the maximum sample size, with more participants receiving nortriptyline and duloxetine. At the end of the follow-up, nortriptyline and duloxetine had lower probabilities of participants that had stopped taking the study medication and higher probabilities were efficacious. Mexiletine had the highest quit rate, but had an efficacy rate higher than pregabalin. Conclusions: Response adaptive randomization has become a popular trial tool, especially for those utilizing Bayesian methods for analyses. By illustrating the execution of a Bayesian adaptive design, using the PAIN-CONTRoLS trial data, this paper continues the work to provide literature for conducting Bayesian response adaptive randomized trials.

An open-source system for efficient clinical trial support: The COMET study experience.

Exercise clinical trials are complex, logistically burdensome, and require a well-coordinated multi-disciplinary approach. Challenges include managing, curating, and reporting on many disparate information sources, while remaining responsive to a variety of stakeholders. The Combined Exercise Trial (COMET, NCT04848038) is a one-year comparison of three exercise modalities delivered in the community. Target enrollment is 280 individuals over 4 years. To support rigorous execution of COMET, the study team has developed a suite of scripts and dashboards to assist study stakeholders in each of their various functions. The result is a highly automated study system that preserves rigor, increases communication, and reduces staff burden. This manuscript describes system considerations and the COMET approach to data management and use, with a goal of encouraging further development and adaptation by other study teams in various fields.

The rural mortality penalty in U.S. hospital patients with COVID-19.

Background: The COVID-19 pandemic brought greater focus to the rural mortality penalty in the U.S., which describes the greater mortality rate in rural compared to urban areas. Although it is understood that issues such as access to care, age structure of the population, and differences in behavior are likely drivers of the rural mortality penalty, it is critical to try and understand these factors to enable more effective public health policy. Methods: We performed a cross-sectional analysis of a population of patients with COVID-19 who were admitted to hospitals in the United States between 3/1/2020 and 2/26/2023 to better understand factors leading to outcome disparities amongst groups that all had some level of access to hospital care, hypothesizing that deteriorated patient condition at admission likely explained some of the observed difference in mortality between rural and urban populations. Results: Our results supported our hypothesis, showing that the rural mortality penalty persists in this population and that by multiple measures, rural patients were likely to be admitted in worse condition, had worse overall health, and were older. Conclusions: Although the pandemic threw the rural mortality penalty into sharp relief, it is important to remember that it existed prior to the pandemic and will continue to exist until effective interventions are implemented. This study demonstrates the critical need to address the underlying factors that resulted in rural-dwelling patients being admitted to the hospital in worse condition than their urban-dwelling counterparts during the COVID-19 pandemic, which likely affected other healthcare outcomes as well.

Patient Polypharmacy use Following a Multi-Disciplinary Dementia Care Program in a Memory Clinic: A Retrospective Cohort Study.

Introduction: Dementia increases the risk of polypharmacy. Timely detection and optimal care can stabilize or delay the progression of dementia symptoms, which may in turn reduce polypharmacy. We aimed to evaluate the change in polypharmacy use among memory clinic patients living with dementia who participated in a dementia care program compared to those who did not. We hypothesized that patients in the dementia care program would reduce their use of polypharmacy compared to those who were not in standard care. Methods: We retrospectively analyzed data extracted from electronic medical records from a university memory clinic. Data from a total of 381 patients were included in the study: 107 in the program and 274 matched patients in standard care. We used adjusted odds ratios to assess the association between enrollment in the program and polypharmacy use at follow-up (five or more concurrent medications), controlling for baseline polypharmacy use and stratified polypharmacy use by prescription and over-the-counter (OTC). Results: The two groups did not differ in the use of five or more overall and prescription medications at follow-up, controlling for the use of five or more of the respective medications at baseline and covariates. Being in the program was associated with a three-fold lower odds of using five or more OTC medications at follow-up (adjusted odds ratio = 0.30; p <0.001; 95% Confidence interval = 0.15-0.58) after controlling for using five or more OTC medications at baseline and covariates. Conclusions: Dementia care may reduce polypharmacy of OTC medications, potentially reducing risky drug-drug interactions. More research is needed to infer causality and understand how to reduce prescription medication polypharmacy.

DHA, nutrient intake, and maternal characteristics as predictors of pregnancy outcomes in a randomised clinical trial of DHA supplementation.

PURPOSE: To investigate the relationships among docosahexaenoic acid (DHA) intake, nutrient intake, and maternal characteristics on pregnancy outcomes in a phase III randomised clinical trial designed to determine the effect of a DHA dose of 1000 mg/day compared to 200 mg/day on early preterm birth (<34 weeks gestation). METHODS: A secondary aim of the phase III randomised trial was to explore the relationships among pregnancy outcomes (maternal red blood cell phospholipid (RBC-PL) DHA at delivery, preterm birth, gestational age at delivery, labor type, birth anthropometric measures, low birth weight, gestational diabetes, pre-eclampsia, and admission to a neonatal intensive care unit) in participants (n = 1100). We used Bayesian multiple imputation and linear and logistic regression models to conduct an analysis of five general classes of predictor variables collected during the trial: a) DHA intake, b) nutrient intake from food and supplements, c) environmental exposure to tobacco and alcohol, d) maternal demographics, and e) maternal medical history. RESULTS: DHA supplementation lowered the risk of preterm birth and NICU admission, and increased gestation and birth weight as observed in the primary analysis. Higher maternal RBC-PL-DHA at delivery was associated with DHA supplementation and formal education of a bachelor's degree or higher. DHA supplementation and maternal age were associated with a higher risk of gestational diabetes. Total vitamin A intake was associated with longer gestation, while fructose and intake of the long chain omega-6 fatty acid, arachidonic acid, were associated with shorter gestation. Risk of preterm birth was associated with a history of low birth weight, preterm birth, pre-eclampsia, and NICU admission. CONCLUSION: Bayesian models provide a comprehensive approach to relationships among DHA intake, nutrient intake, maternal characteristics, and pregnancy outcomes. We observed previously unreported relationships between gestation duration and fructose, vitamin A, and arachidonic acid that could be the basis for future research. TRIAL REGISTRATION NUMBER AND DATE: ClinicalTrials.gov (NCT02626299); December 10, 2015.

Evaluating Work Impairment as a Source of Financial Toxicity in Cancer Healthcare and Negative Impacts on Health Status.

How the socioeconomic factors intersect for a particular patient can determine their susceptibility to financial toxicity, what costs they will encounter during treatment, the type and quality of their care, and the potential work impairments they face. The primary goal of this study was to evaluate financial factors leading to worsening health outcomes by the cancer subtype. A logistic model predicting worsening health outcomes while assessing the most influential economic factors was constructed by the University of Michigan Health and Retirement Study. A forward stepwise regression procedure was implemented to identify the social risk factors that impact health status. Stepwise regression was done on data subsets based on the cancer types of lung, breast, prostate, and colon cancer to determine whether significant predictors of worsening health status were different or the same across cancer types. Independent covariate analysis was also conducted to cross-validate our model. On the basis of the model fit statistics, the two-factor model has the best fit, that is, the lowest AIC among potential models of 3270.56, percent concordance of 64.7, and a C-statistics of 0.65. The two-factor model used work impairment and out-of-pocket costs, significantly contributing to worsening health outcomes. Covariate analysis demonstrated that younger patients with cancer experienced more financial burdens leading to worsening health outcomes than elderly patients aged 65 years and above. Work impairment and high out-of-pocket costs were significantly associated with worsening health outcomes among cancer patients. Matching the participants who need the most financial help with appropriate resources is essential to mitigate the financial burden. Significance: Among patients with cancer, work impairment and out-of-pocket are the two primary factors contributing to adverse health outcomes. Women, African American or other races, the Hispanic population, and younger individuals have encountered higher work impairment and out-of-pocket costs due to cancer than their counterparts.

Rurality Impacts Mammography Screening Adherence Among Mid-life Women in the Kansas Region.

OBJECTIVE: The gold standard for breast cancer screening and prevention is regular mammography; thus, understanding what impacts adherence to this standard is essential in limiting cancer-associated costs. We assessed the impact of various understudied sociodemographic factors of interest on adherence to the receipt of regular mammograms. METHODS: A total Nc = 14,553 mammography-related claims from Nw = 6,336 female Kansas aged between 45 and 54 were utilized from insurance claim databases furnished by multiple providers. Adherence to regular mammography was quantified continuously via a compliance ratio, used to capture the number of eligible years in which at least one mammogram was received, as well as categorically. The relationship between race, ethnicity, rurality, insurance (public/private), screening facility type, and distance to nearest screening facility with both continuous and categorically defined compliance were individually assessed via Kruskal-Wallis one-way ANOVAs, chi-squared tests, multiple linear regression models, and multiple logistic regression, as appropriate. Findings from these individual models were used to inform the construction of a basic, multifaceted prediction model. RESULTS: Model results demonstrated that all factors race and ethnicity had at least some bearing on compliance with screening guidelines among mid-life female Kansans. The strongest signal was observed in the rurality variable, which demonstrated a significant relationship with compliance regardless of how it was defined. CONCLUSION: Understudied factors that are associated with regular mammography adherence, such as rurality and distance to nearest facility, may serve as important considerations when developing intervention strategies for ensuring that female patients stick to prescribed screening regimens.

Career disruption and limitation of financial earnings due to cancer.

PURPOSE: This study investigated how cancer diagnosis and treatment lead to career disruption and, consequently, loss of income and depletion of savings. DESIGN: This study followed a qualitative descriptive design that allowed us to understand the characteristics and trends of the participants. METHOD: Patients recruited (n = 20) for this study were part of the University of Kansas Cancer Center patient advocacy research group (Patient and Investigator Voices Organizing Together). The inclusion criteria were that participants must be cancer survivors or co-survivors, be aged 18 years or older, be either employed or a student at the time of cancer diagnosis, have completed their cancer treatment, and be in remission. The responses were transcribed and coded inductively to identify themes. A thematic network was constructed based on those themes, allowing us to explore and describe the intricacies of the various themes and their impacts. RESULTS: Most patients had to quit their jobs or take extended absences from work to handle treatment challenges. Patients employed by the same employer for longer durations had the most flexibility to balance their time between cancer treatment and work. Essential, actionable items suggested by the cancer survivors included disseminating information about coping with financial burdens and ensuring that a nurse and financial navigator were assigned to every cancer patient. CONCLUSIONS: Career disruption is common among cancer patients, and the financial burden due to their career trajectory is irreparable. The financial burden is more prominent in younger cancer patients and creates a cascading effect that financially affects close family members.

The Effects of COVID-19 Pandemic Policy on Social Needs Across the State of Kansas and Western Missouri: Paired Survey Response Testing.

BACKGROUND: Studying patients' social needs is critical to the understanding of health conditions and disparities, and to inform strategies for improving health outcomes. Studies have shown that people of color, low-income families, and those with lower educational attainment experience greater hardships related to social needs. The COVID-19 pandemic represents an event that severely impacted people's social needs. This pandemic was declared by the World Health Organization on March 11, 2020, and contributed to food and housing insecurity, while highlighting weaknesses in the health care system surrounding access to care. To combat these issues, legislators implemented unique policies and procedures to help alleviate worsening social needs throughout the pandemic, which had not previously been exerted to this degree. We believe that improvements related to COVID-19 legislature and policy have positively impacted people's social needs in Kansas and Missouri, United States. In particular, Wyandotte County is of interest as it suffers greatly from issues related to social needs that many of these COVID-19-related policies aimed to improve. OBJECTIVE: The research objective of this study was to evaluate the change in social needs before and after the COVID-19 pandemic declaration based on responses to a survey from The University of Kansas Health System (TUKHS). We further aimed to compare the social needs of respondents from Wyandotte County from those of respondents in other counties in the Kansas City metropolitan area. METHODS: Social needs survey data from 2016 to 2022 were collected from a 12-question patient-administered survey distributed by TUKHS during a patient visit. This provided a longitudinal data set with 248,582 observations, which was narrowed down into a paired-response data set for 50,441 individuals who had provided at least one response before and after March 11, 2020. These data were then bucketed by county into Cass (Missouri), Clay (Missouri), Jackson (Missouri), Johnson (Kansas), Leavenworth (Kansas), Platte (Missouri), Wyandotte (Kansas), and Other counties, creating groupings with at least 1000 responses in each category. A pre-post composite score was calculated for each individual by adding their coded responses (yes=1, no=0) across the 12 questions. The Stuart-Maxwell marginal homogeneity test was used to compare the pre and post composite scores across all counties. Additionally, McNemar tests were performed to compare responses before and after March 11, 2020, for each of the 12 questions across all counties. Finally, McNemar tests were performed for questions 1, 7, 8, 9, and 10 for each of the bucketed counties. Significance was assessed at P<.05 for all tests. RESULTS: The Stuart-Maxwell test for marginal homogeneity was significant (P<.001), indicating that respondents were overall less likely to identify an unmet social need after the COVID-19 pandemic. McNemar tests for individual questions indicated that after the COVID-19 pandemic, respondents across all counties were less likely to identify unmet social needs related to food availability (odds ratio [OR]=0.4073, P<.001), home utilities (OR=0.4538, P<.001), housing (OR=0.7143, P<.001), safety among cohabitants (OR=0.6148, P<.001), safety in their residential location (OR=0.6172, P<.001), child care (OR=0.7410, P<0.01), health care access (OR=0.3895, P<.001), medication adherence (OR=0.5449, P<.001), health care adherence (OR=0.6378, P<.001), and health care literacy (0.8729, P=.02), and were also less likely to request help with these unmet needs (OR=0.7368, P<.001) compared with prepandemic responses. Responses from individual counties were consistent with the overall results for the most part. Notably, no individual county demonstrated a significant reduction in social needs relating to a lack of companionship. CONCLUSIONS: Post-COVID-19 responses showed improvement across almost all social needs-related questions, indicating that the federal policy response possibly had a positive impact on social needs across the populations of Kansas and western Missouri. Some counties were impacted more than others and positive outcomes were not limited to urban counties. The availability of resources, safety net services, access to health care, and educational opportunities could play a role in this change. Future research should focus on improving survey response rates from rural counties to increase their sample size, and to evaluate other explanatory variables such as food pantry access, educational status, employment opportunities, and access to community resources. Government policies should be an area of focused research as they may affect the social needs and health of the individuals considered in this analysis.

Financial burden among cancer patients: A national-level perspective.

BACKGROUND: This research study aimed to evaluate the financial burden among older cancer patients and its corresponding risk factors. Factors such as increasing treatment costs and work limitations often lead cancer patients to bankruptcy and poor quality of life. These consequences, in turn, can cause higher mortality rates among these patients. METHODS: This retrospective cohort study utilized data from the Health Retirement Study (HRS), conducted by the University of Michigan (N = 18,109). Eligible participants had responses captured from years 2002 to 2016. Participants were classified according to any self-reported cancer diagnosis (yes or no) and were compared on the basis of financial, work, and health-related outcomes. Propensity score (PS) matching was applied to reduce the effects of potential confounding factors. Also only, individuals with an age ≥50 and ≤85 during Wave 6 were retained. RESULTS: Multivariate analysis with random effects revealed several indicators of financial burden when comparing participants with a cancer diagnosis to those with no history of cancer. Mean out-of-pocket costs associated with a cancer diagnosis were $1058 higher when compared to participants with no history of cancer, suggesting that even cancer patients with insurance coverage faced out-of-pocket costs. Respondents with cancer patients had higher odds of encountering financial hardship if they are facing Work Limitations (OR = 2.714), Regular use of Medications (OR = 2.518), Hospital Stays (OR = 2.858), Declining Health (OR = 2.349), or were being covered under government health insurance (OR = 5.803) than respondents who did not have cancer, or suffered from mental health issues such as Depression (OR = 0.901). CONCLUSION: Cancer patients contend with increasing financial costs during their treatment. However, most newly diagnosed patients are not aware of these costs and are given few resources to handle them.

The value of a two-armed Bayesian response adaptive randomization trial.

We investigate the value of a two-armed Bayesian response adaptive randomization (RAR) design to investigate early preterm birth rates of high versus low dose of docosahexaenoic acid during pregnancy. Unexpectedly, the COVID-19 pandemic forced recruitment to pause at 1100 participants rather than the planned 1355. The difference in power between number of participants at the pause and planned was 87% and 90% respectively. We decided to stop the study. This paper describes how the RAR was used to execute the study. The value of RAR in two-armed studies is quite high and their use in the future is promising.

PerSEveML: A Web-Based Tool to Identify Persistent Biomarker Structure for Rare Events Using Integrative Machine Learning Approach.

Omics datasets often pose a computational challenge due to their high dimensionality, large size, and non-linear structures. Analyzing these datasets becomes especially daunting in the presence of rare events. Machine learning (ML) methods have gained traction for analyzing rare events, yet there remains a limited exploration of bioinformatics tools that integrate ML techniques to comprehend the underlying biology. Expanding upon our previously developed computational framework of an integrative machine learning approach1, we introduce PerSEveML, an interactive web-based that uses crowd-sourced intelligence to predict rare events and determine feature selection structures. PerSEveML provides a comprehensive overview of the integrative approach through evaluation metrics that help users understand the contribution of individual ML methods to the prediction process. Additionally, PerSEveML calculates entropy and rank scores, which visually organize input features into a persistent structure of selected, unselected, and fluctuating categories that help researchers uncover meaningful hypotheses regarding the underlying biology. We have evaluated PerSEveML on three diverse biologically complex data sets with extremely rare events from small to large scale and have demonstrated its ability to generate valid hypotheses. PerSEveML is available at https://biostats-shinyr.kumc.edu/PerSEveML/ and https://github.com/sreejatadutta/PerSEveML.

A novel Bayesian adaptive design incorporating both primary and secondary endpoints for randomized IIB chemoprevention study of women at increased risk for breast cancer.

BACKGROUND: Our randomized controlled clinical trial will explore the potential of bazedoxifene plus conjugated estrogen to modulate breast tissue-based risk biomarkers as a surrogate for breast cancer risk reduction. This paper investigates the statistical design features of the trial and the rationale for the final choice of its design. Group sequential designs are a popular design approach to allow a trial to stop early for success or futility, potentially saving time and money over a fixed trial design. While Bayesian adaptive designs enjoy the same properties as group sequential designs, they have the added benefit of using prior information as well as inferential interpretation conditional on the data. Whether a frequentist or Bayesian trial, most adaptive designs have interim analyses that allow for early stopping, typically utilizing only the primary endpoint. A drawback to this approach is that the study may not have enough data for adequate comparisons of a single, key secondary endpoint. This can happen, for example, if the secondary endpoint has a smaller effect than the primary endpoint. METHODS: In this paper, we investigate a trial design called two-endpoint adaptive, which stops early only if a criterion is met for primary and secondary endpoints. The approach focuses the final analysis on the primary endpoint but ensures adequate data for the secondary analysis. Our study has two arms with a primary (change in mammographic fibroglandular volume) and secondary endpoint (change in mammary tissue Ki-67). RESULTS: We present operating characteristics including power, trial duration, and type I error rate and discuss the value and risks of modeling Bayesian group sequential designs with primary and secondary endpoints, comparing against alternative designs. The results indicate that the two-endpoint adaptive design has better operating characteristics than competing designs if one is concerned about having adequate information for a key secondary endpoint. DISCUSSION: Our approach balances trial speed and the need for information on the single, key secondary endpoint.

Bolstering the complex study start-up process at NCI cancer centers using technology.

Background: The study startup process for interventional clinical trials is a complex process that involves the efforts of many different teams. Each team is responsible for their startup checklist in which they verify that the necessary tasks are done before a study can move on to the next team. This regulatory process provides quality assurance and is vital for ensuring patient safety [10]. However, without having this startup process centralized and optimized, study approval can take longer than necessary as time is lost when it passes through many different hands. Objective: This manuscript highlights the process and the systems that were developed at The University of Kansas Comprehensive Cancer Center regarding the study startup process. To facilitate this process the regulatory management, site development, cancer center administration, and the Biostatistics & Informatics Shared Resources (BISR) teams came together to build a platform aimed at streamlining the startup process and providing a transparent view of where a study is in the startup process. Process: Ensuring the guidelines are clearly articulated for the review criteria of each of the three review boards, i.e., Disease Working Group (DWG), Executive Resourcing Committee (ERC), and Protocol Review and Monitoring Committee (PRMC) along with a system that can track every step and its history throughout the review process. Results: Well-defined processes and tracking methodologies have allowed the operations teams to track each study closely and ensure the 90-day and 120-day deadlines are met, this allows the operational team to dynamically prioritize their work daily. It also provides Principal investigators a transparent view of where their study stands within the study startup process and allows them to prepare for the next steps accordingly. Conclusion/future work: The current process and technology deployment has been a significant improvement to expedite the review process and minimize study startup delays. There are still a few opportunities to fine-tune the study startup process; an example of which includes automatically informing the operational managers or the study teams to act upon deadlines regarding study review rather than the current manual communication process which involves them looking it up in the system which can add delays.