I am a Professor in the Department of Biostatistics and Medical Informatics and the Department of Population Health Sciences in the School of Medicine and Public Health at the University of Wisconsin-Madison. I have an affiliate appointment in the Department of Statistics.
I grew up in Duluth, Minnesota. I graduated from East High School in 1988. I received a BA in Mathematics and Economics in 1992 from the University of Minnesota-Duluth and an MS in Statistics in 1994 and a PhD in Statistics with emphasis in Biostatistics in 1998 from the University of Wisconsin-Madison. My PhD advisor was Murray Clayton.
I was a research scientist in the Statistical Data Analysis Center(SDAC) in the Department of Biostatistics and Medical Informatics at the University of Wisconsin-Madison, 1998-2005. I joined the faculty with a joint appointment in the Department of Biostatistics and Medical Informatics and the Department of Population Health Sciences in 2005.
I am an applied biostatistician focusing on problems in clinical and epidemiologic research. Current methodologic areas of interest include (1) multi-state models for incidence, progression and regression of ocular (and other) diseases, (2) small area estimation problems, particularly ranking, (3) spatial and spatio-temporal modeling, particularly cluster detection and high-dimensional variable selection and (4) age-period-cohort modeling.
Outside of the office, I’m an avid cyclist. You can check out my recent rides on Strava. I’m also a big movie fan. You’ll definitely see me at the Wisconsin Film Festival, one of my favorite events every year, and you can see lists of my favorite films and what I’ve been watching recently on Letterboxd.
PhD in Statistics (emphasis in Biostatistics), 1998
University of Wisconsin-Madison
MS in Statistics, 1994
University of Wisconsin-Madison
BA in Mathematics and Economics, 1992
University of Minnesota-Duluth
Context: Population health rankings can be a catalyst for the improvement of health by drawing attention to areas in need of relative improvement and summarizing complex information in a manner understood by almost everyone. However, ranks also have unintended consequences, such as being interpreted as “hard truths”, where variations may not be significant. There is a need to improve communication about uncertainty in ranks, with accurate interpretation. The most common solutions discussed in the literature have included modeling approaches to minimize statistical noise or borrow strength from covariates. However, the use of complex models can limit communication and implementation, especially for broad audiences. Objectives: Explore data-informed grouping (cluster analysis) as an easier-to-understand, empirical technique to account for rank imprecision that can be effectively communicated both numerically and visually. Design: Cluster analysis, specifically k-means clustering with Wasserstein (earth mover’s) distance, was explored as an approach to identify natural and meaningful groupings and gaps in the data distribution for the County Health Rankings’ (CHR) health outcomes ranks. Setting: County-level health outcomes from the 2022 CHR Participants: 3,082 counties that were ranked in the 2022 CHR Main Outcome Measure: Data-informed health groups Results: Cluster analysis identified 30 health groupings among counties nationwide, with cluster size ranging from nine to 184 counties. On average, states had 16 identified clusters, ranging from 3 in Delaware and Hawaii to 27 in Virginia. Number of clusters per state was associated with number of counties per state and population of the state. The method helped address many of the issues that arise from providing rank estimates alone. Conclusions: Public health practitioners can use this information to understand uncertainty in ranks, visualize distances between county ranks, have context around which counties are not meaningfully different from one another, and compare county performance to peer counties.
Objective Investigate the contributions of Area Deprivation Index (ADI) and geographic location on age-matched birth weight z-scores. Methods A retrospective cohort study of all singleton births >22 weeks’ gestation in Dane County, WI from 1/2016-6/2018. Generalized additive models were adjusted for race/ethnicity, cigarette use, delivery route, pregnancy-related or chronic hypertension, pregestational and gestational diabetes, number of prenatal visits, maternal age, total weight gain, and pre-pregnancy BMI. Results We find evidence of an association between birth weight z-score and spatial location (median p-value: 0.006). With area deprivation, we find no evidence of an association with birth weight z-score (-0.01, 95% CI: -0.03-0.01, p= 0.109). Mean birth weight z-scores were lowest (-0.72) in the urban center of Madison, while mean birth weight z-score was highest (0.18) in rural areas near the northeast, southeast, and southwest county borders. We find an effect of race/ethnicity on birth weight. Conclusion We identified geographic variations in birth weight at a granular level using census block groups and a holistic measure of deprivation, which can inform targeted public health interventions. The lack of evidence of area deprivation on birth outcomes but significant spatial trends demonstrated continued geographic disparities in our healthcare systems.
Prior research has shown that cancer risk varies by geography, but scan statistic methods for identifying cancer clusters in case-control studies have been limited in their ability to identify multiple clusters and adjust for participant-level risk factors. We develop a method to identify geographic patterns of breast cancer odds using the Wisconsin Women’s Health Study, a series of 5 population-based case-control studies of female Wisconsin residents aged 20-79 enrolled in 1988-2004 (cases=16,076, controls=16,795). We create sets of potential clusters by overlaying a 1 km grid over each county-neighborhood and enumerating a series of overlapping circles. Using a two-step approach, we fi=t a penalized binomial regression model to the number of cases and trials in each grid cell, penalizing all potential clusters by the least absolute shrinkage and selection operator (Lasso). We use BIC to select the number of clusters, which are included in a participant-level logistic regression model. We identify 15 geographic clusters, resulting in 23 areas of unique geographic odds ratios. After adjustment for known risk factors, confidence intervals narrowed but breast cancer odds ratios did not meaningfully change; one additional hotspot was identified. By considering multiple overlapping spatial clusters simultaneously, we discern gradients of spatial odds across Wisconsin.
Spatial and spatio-temporal cluster detection are important tools in public health and many other areas of application. Cluster detection can be approached as a multiple testing problem, typically using a space and time scan statistic. We recast the spatial and spatio-temporal cluster detection problem in a high-dimensional data analytical framework with Poisson or quasi-Poisson regression with the Lasso penalty. We adopt a fast and computationally-efficient method using a novel sparse matrix representation of the effects of potential clusters. The number of clusters and tuning parameters are selected based on (quasi-)information criteria. We evaluate the performance of our proposed method including the false positive detection rate and power using a simulation study. Application of the method is illustrated using breast cancer incidence data from three prefectures in Japan.
Background: Recent reports suggest that racial differences in breast cancer incidence rates have decreased. We examined whether these findings apply to breast cancer mortality while considering age, period, and cohort influences on both absolute and relative measures of breast cancer mortality. Methods: Using publicly available datasets (CDC Wonder, Human Mortality Database), we developed an age-period-cohort model of breast cancer mortality and breast cancer deaths as a proportion of all deaths during 1968-2019 among all women and by five race/ethnicity groups with sufficient numbers for estimation: Hispanic (all races), American Indian/Alaska Native and Asian/Pacific Islanders (regardless of ethnicity), non-Hispanic Black, and non-Hispanic White. Results: Initially increasing after 1968, age-adjusted breast cancer mortality rates have decreased among all racial/ethnic groups since 1988. The age-adjusted percent of all deaths due to breast cancer also has been declining for non-Hispanic White women since about 1990 while increasing or holding steady for other race/ethnic groups. In 2019, the age-adjusted percent of deaths due to breast cancer for women was highest for Asian/Pacific Islanders (5.6%) followed by non-Hispanic Black (4.5%), Hispanic (4.4%), non-Hispanic White (4.1%), and American Indian/Alaska Native women (2.6%). Conclusions: Breast cancer mortality disparities are now greater on both relative and absolute scales for non-Hispanic Black women, and using the relative scale for Asian/Pacific Islanders and Hispanic women, compared with non-Hispanic White women for the first time in 50 years.
Context: Population health rankings can be a catalyst for the improvement of health by drawing attention to areas in need of relative improvement and summarizing complex information in a manner understood by almost everyone. However, ranks also have unintended consequences, such as being interpreted as “hard truths”, where variations may not be significant. There is a need to improve communication about uncertainty in ranks, with accurate interpretation. The most common solutions discussed in the literature have included modeling approaches to minimize statistical noise or borrow strength from covariates. However, the use of complex models can limit communication and implementation, especially for broad audiences. Objectives: Explore data-informed grouping (cluster analysis) as an easier-to-understand, empirical technique to account for rank imprecision that can be effectively communicated both numerically and visually. Design: Cluster analysis, specifically k-means clustering with Wasserstein (earth mover’s) distance, was explored as an approach to identify natural and meaningful groupings and gaps in the data distribution for the County Health Rankings’ (CHR) health outcomes ranks. Setting: County-level health outcomes from the 2022 CHR Participants: 3,082 counties that were ranked in the 2022 CHR Main Outcome Measure: Data-informed health groups Results: Cluster analysis identified 30 health groupings among counties nationwide, with cluster size ranging from nine to 184 counties. On average, states had 16 identified clusters, ranging from 3 in Delaware and Hawaii to 27 in Virginia. Number of clusters per state was associated with number of counties per state and population of the state. The method helped address many of the issues that arise from providing rank estimates alone. Conclusions: Public health practitioners can use this information to understand uncertainty in ranks, visualize distances between county ranks, have context around which counties are not meaningfully different from one another, and compare county performance to peer counties.
Objective: We examined associations between older drivers’ social and environmental characteristics and odds of using non-driving transportation modes. Methods: Using 2015 National Health and Aging Trends Study data for community-dwelling drivers (n = 5102), we estimated logistic regression models of associations between social characteristics, environmental characteristics, and odds of using non-driving transportation modes three years later. Results: Drivers had 20% increase in odds of getting rides three years later for each additional confidante (adjusted odds ratio [aOR] = 1.20, 95% confidence interval [CI]: 1.11–1.30). Drivers living in more walkable neighborhoods were more likely to walk to get places (National Walkability Index [NWI] score of 18 vs. 2 aOR = 1.71, 95% CI: 1.02–2.90) and take public transit three years later (NWI 18 vs. 2 aOR = 7.47, 95% CI: 1.69–33.0). Discussion: Identifying modifiable social and environmental characteristics can inform future interventions supporting older adults’ health during the transition to non-driving.
IMPORTANCE The effects of breast cancer incidence changes and advances in screening and treatment on outcomes of different screening strategies are not well known. OBJECTIVE To estimate outcomes of various mammography screening strategies. DESIGN, SETTING, AND POPULATION Comparison of outcomes using 6 Cancer Intervention and Surveillance Modeling Network (CISNET) models and national data on breast cancer incidence, mammography performance, treatment effects, and other-cause mortality in US women without previous cancer diagnoses. EXPOSURES Thirty-six screening strategies with varying start ages (40, 45, 50 years) and stop ages (74, 79 years) with digital mammography or digital breast tomosynthesis (DBT) annually, biennially, or a combination of intervals. Strategies were evaluated for all women and for Black women, assuming 100% screening adherence and “real-world” treatment. MAIN OUTCOMES AND MEASURES Estimated lifetime benefits (breast cancer deaths averted, percent reduction in breast cancer mortality, life-years gained), harms (false-positive recalls, benign biopsies, overdiagnosis), and number of mammograms per 1000 women. RESULTS Biennial screening with DBT starting at age 40, 45, or 50 years until age 74 years averted a median of 8.2, 7.5, or 6.7 breast cancer deaths per 1000 women screened, respectively, vs no screening. Biennial DBT screening at age 40 to 74 years (vs no screening) was associated with a 30.0%breast cancer mortality reduction, 1376 false-positive recalls, and 14 overdiagnosed cases per 1000 women screened. Digital mammography screening benefits were similar to those for DBT but had more false-positive recalls. Annual screening increased benefits but resulted in more false-positive recalls and overdiagnosed cases. Benefit-to-harm ratios of continuing screening until age 79 years were similar or superior to stopping at age 74. In all strategies, women with higher-than-average breast cancer risk, higher breast density, and lower comorbidity level experienced greater screening benefits than other groups. Annual screening of Black women from age 40 to 49 years with biennial screening thereafter reduced breast cancer mortality disparities while maintaining similar benefit-to-harm trade-offs as for all women. CONCLUSIONS This modeling analysis suggests that biennial mammography screening starting at age 40 years reduces breast cancer mortality and increases life-years gained per mammogram. More intensive screening forwomen with greater risk of breast cancer diagnosis or death can maintain similar benefit-to-harm trade-offs and reduce mortality disparities.
Objective To test our initial hypotheses that the COVID-19 pandemic was associated with: (1) decreases in adaptive behavior and increases in behavioral and emotional problems of children with autism; (2) greater impacts for those who lost specialty services; and (3) greater behavioral and emotional problems for children with autism versus control participants. Method Eligible participants (N=1,158) enrolled in Phase 3 of the multi-site, case-control Study to Explore Early Development (SEED) prior to March 31, 2020, between 2-5 years old and completed follow-up assessments between January-July 2021. Caregivers completed a COVID-19 Impact Assessment Questionnaire, Vineland Adaptive Behavior Scales (VABS), and Child Behavior Checklist (CBCL) for 274 children with autism and 385 control participants. Results Mean VABS communication scores of children with autism decreased significantly (-4.2; standard deviation [SD], 10.5) between pre-pandemic and pandemic periods, while VABS composite (+2.0; SD, 9.0), daily living (+5.5; SD, 11.4), socialization (+2.3; SD, 10.0), and CBCL scores (-3.2; SD, 8.4) improved. In contrast, CBCL scores worsened in population control participants (+3.4; SD, 8.8). Children with autism who missed specialty appointments scored significantly lower on the VABS during the pandemic versus those who did not (VABS Composite 70.6; 95% confidence interval [CI]: 68.8-72.4 vs. 74.5; 95% CI: 71.8-77.2). Conclusion While stay-at-home policies of the pandemic may have beneficially impacted daily living skills, socialization, and behavioral and emotional wellbeing of children with autism, benefits may have occurred at the cost of communication skills. These findings indicate the need for strategies to maintain therapeutic services in future emergency settings.