Elizabeth Handorf, PhD

Elizabeth Handorf, PhD

Associate Professor

Educational Background

  • PhD, Biostatistics, University of Pennsylvania, Philadelphia, PA, 2012
  • BSE, Systems Engineering, University of Pennsylvania, Philadelphia, PA, 2005


  • International Biometrics Society
  • American Statistical Association
  • Society for Medical Decision Making

Honors & Awards

  • Student Travel Award, International Conference in Health Policy Statistics, 2011
  • Jonathan Raz Award, 2010
  • GlaxoSmithKline Scholar Award, 2010

Research Interests

  • Statistical methods for observational data
  • Cost-effectiveness and comparative effectiveness studies
  • Propensity score methods
  • Machine learning methods for prediction and variable selection
  • Assessing sensitivity to unmeasured confounding variables
  • Interventional studies
  • Tissue microarray analysis
  • Tumor growth modeling

Lab Overview

My research focuses on statistical analysis of observational data, particularly applied to comparative-effectiveness, cost-effectiveness, and population health. Many important healthcare questions are infeasible or impractical to test via randomized controlled trials. Often, the most rigorous way to evaluate hypotheses in the absence of a randomized trial is by extracting and analyzing information from large observational datasets, such as cancer registries, claims files, electronic medical records (EMR), or census data. I develop innovative methods to analyze such data, often using computationally intensive solutions.

Lab Description

  • Collaborative work with clinicians and scientists
    I provide statistical expertise for a wide variety of studies led by Fox Chase Cancer Center investigators, including efficacy analysis of interventions, biomarker analysis and non-therapeutic randomized trials. Often, these studies require advanced statistical analyses or novel applications of existing methods. For example, I developed a novel hierarchical clustering and latent class analysis procedure to evaluate quality of survey responses from an anonymous web panel. (Handorf et al, 2018)
  • Development of sensitivity analysis methods for cost and cost-effectiveness data
    One major drawback of non-randomized data is that groups of patients with different exposures of interest are often systematically different. Although most secondary data allows us to adjust for basic risk factors, such as gender, age and disease characteristics, it is unlikely that all important variables will be available in an observational dataset. These omitted variables, also called unmeasured confounders, may bias conclusions of study results. I developed a comprehensive framework for evaluating the effect of unmeasured confounders of cost and cost-effectiveness outcomes. (Handorf et al, 2013, 2018)
  • Computationally intensive and machine learning methods
    I apply and extend cutting-edge machine learning methods for high-dimensional data.For example, I conducted a comprehensive evaluation of sparse regressions and tree ensemble methods for variable selection when analyzing U.S. Census data.(Handorf 2020) I also use sophisticated ensemble models (e.g. SuperLearner, Bayesian Additive Regression Trees) for estimation of propensity score weights in observational data analysis.
  • Application of methods for observational data to clinically relevant questions
    I have conducted many analyses of large datasets based on questions from clinical collaborators, using methods appropriate for the observational nature of the data and the outcome of interest. For example, I used the National Cancer Database to demonstrate that delays in treatment for head and neck cancers are associated with poorer survival. (Murphy, 2016) I have conducted many studies using data from the SEER registry, linked SEER-Medicare data, and EMR-derived databases.

Selected Publications

Vapiwala N., Wong J.K., Handorf E., Paly J., Grewal A., Tendulkar R., Godfrey D., Carpenter D., Mendenhall N.P., Henderson R.H., Stish B.J., Vargas C., Salama J.K., Davis B.J.,Horwitz E.M., A pooled toxicity analysis of moderately hypofractionated proton beam therapy and intensity modulated radiation therapy in early-stage prostate cancer patients. Int J Radiat Oncol Biol Phys. 110(4): 1082-1089, 2021. https://www.ncbi.nlm.nih.gov/pubmed/33539968.

Winer A., Handorf E.,Dotan E., Dosing schedules of gemcitabine and nab-paclitaxel for older adults with metastatic pancreatic cancer. Jnci Cancer Spectrum. 5(5): pkab074, 2021. PMC8438244. https://www.ncbi.nlm.nih.gov/pubmed/34532641.

Wong J.K., Handorf E., Lee D., Jain R., Zhang E., Cooper H.S., Farma J.M., Dotan E.,Meyer J.E., Toxicity and outcomes in older versus younger patients treated with trimodality therapy for locally advanced rectal cancer. J Geriatr Oncol. 11(8): 1331-1334, 2020. PMC7606716. https://www.ncbi.nlm.nih.gov/pubmed/32381438.

Handorf E, Yin Y, Slifker M, Lynch S. Variable selection in social-environmental data: Sparse regression and tree ensemble machine learning approaches. BMC Medical Research Methodology.  2020 Dec 10;20(1):302.

Handorf EA, Heckman CJ, Darlow S, Slifker M, Ritterband L. A hierarchical clustering approach to identify repeated enrollments in web survey data. PLOS One. 2018;13(9):e0204394.

Handorf EA, Bekelman JE, Heitjan DF, Mitra N. Evaluating costs with unmeasured confounding: a sensitivity analysis for the treatment effect. The annals of applied statistics. 2013;7(4):2062-80.

Handorf EA, Heitjan DF, Bekelman JE, Mitra N. Estimating cost-effectiveness from claims and registry data with measured and unmeasured confounders. Statistical methods in medical research. 2018;28(7):2227-42.

Ramamurthy C, Handorf EA, Correa AF, Beck JR, Geynisman DM. Cost-effectiveness of abiraterone versus docetaxel in the treatment of metastatic hormone naive prostate cancer. Urologic oncology. 2019;37(10):688-95.

Gabitova-Cornell L, Surumbayeva A, Peri S, Franco-Barraza J, Restifo D, Weitz N, Ogier C, Goldman AR, Hartman TR, Francescone R, Tan Y, Nicolas E, Shah N, Handorf EA, Cai KQ, O'Reilly AM, Sloma I, Chiaverelli R, Moffitt RA, Khazak V, Fang CY, Golemis EA, Cukierman E, Astsaturov I. Cholesterol Pathway Inhibition Induces TGF-β Signaling to Promote Basal Differentiation in Pancreatic Cancer. Cancer Cell. 2020 Oct 12;38(4):567-583.e11.

Lynch SM, Sorice K, Tagai EK, Handorf EA. Use of empiric methods to inform prostate cancer health disparities: Comparison of neighborhood-wide association study "hits" in black and white men. Cancer. 2020 Jan 1;126(9):1949-1957... Expand

Additional Publications

The following ratings and reviews are based on verified feedback collected from independently administered patient experience surveys. The ratings and comments submitted by patients reflect their own views and opinions. Patient identities are withheld to ensure confidentiality and privacy. Learn more about our Patient Experience Ratings.

Ratings Breakdown

Loading ...

Patient comments

Loading ...