Faculty Profile: Joseph Hogan, Sc.D., Harvard University, 1995

Joseph Hogan
Joseph Hogan, Sc.D., Harvard University, 1995
Work: +1 401-863-9243
My research concerns statistical methods for missing data, causal inference, and sensitivity analysis, especially in HIV/AIDS. Much of my recent work focuses on HIV in sub-Saharan Africa. Current projects include development of methods for analysis of large-scale observational data, optimizing use of limited information for clinical monitoring of patients on antiviral therapy, and analysis of HIV sequence data in studies of drug resistance.


I am a faculty member in the Department of Biostatistics and the Center for Statistical Sciences, and have been on the Brown faculty since 1995. I completed my doctoral work in biostatistics at Harvard under the direction of Nan Laird. Prior to that I taught high school mathematics in Los Angeles.

My research concerns the development of statistical methods for missing data, causal inference, and sensitivity analysis, with focus on applications in HIV and behavioral sciences. I have a long-standing interest in HIV and AIDS, which still exerts an enormous burden on society, particularly the most vulnerable; I am inspired by the contributions statisticians have made to understanding, treating and preventing HIV, and am motivated by what remains to be done.

As part of my collaborations in HIV, I serve as Co-Director of the Biostatistics Program for AMPATH, an international consortium of universities in the US, Canada and Kenya focused on treatment and prevention of HIV in Kenya; co-Director of the Biostatistics Core for the Lifespan/Tufts/Brown Center for AIDS Research; and PI for the Biostatistics Core for the Brown Alcohol Research Center for HIV. My basic and collaborative research is funded by the NIH and USAID.

I teach both introductory and advanced courses in biostatistics, and supervise 1-2 PhD students at any given time. Between 2003 and 2012 I directed the Biostatistics Graduate Program.



Research Description

Current projects include:

- Statistical methods for observational event history data. Here we are developing methods for sampling and then analyzing data from an electronic medical record system in order to determine the optimal timing of initiation of antiviral therapy for those who are coinfected with HIV and TB.

- Optimal use of limited resources for clinical monitoring. Monitoring individuals on first-line HIV therapy requires use of viral load testing; however, this is a limited resource in much of sub-Saharan Africa. We are developing methods to combine low-cost clinical markers with limited use of more expensive viral load testing to maximize diagnostic accuracy subject to cost constraints.

- Statistical methods for representing uncertainty attributable to untestable assumptions. Many statistical models, such as those designed to draw causal inference from nonexperimental data, or those fit to incomplete data, rely on assumptions that cannot be tested (e.g., 'missing at random', or 'no unmeasured confounding'). This line of work is particularly important to analysis of clinical trials data (which forms the basis of most regulatory evaluations of new drugs and devices), and to 'patient-centered outcomes research' (which relies on analysis of large non-experimental datasets like electronic health records).

I am interested in developing models that clearly encode these assumptions, and then properly reflect the uncertainty about whether they actually hold in specific settings. My current work in this topic concerns the use prior distributions to represent untestable assumptions through one or more easily interpreted parameters that represent untestable assumptions.

Grants and Awards

Manning Assistant Professor, Brown University (2000-2003)
Fellow of the American Statistical Association (Elected 2008)


Editorial board
Journal of the American Statistical Association

International Consortium
AMPATH Partnership
(Kenya, US, Canada)

Professional Societies
American Statistical Association
International Biometric Society
Institute of Mathematical Statistics
American Academy for the Advancement of Science

Funded Research

Principal Investigator

New Approaches to Mediation Analysis Using Causal Inference Methods
Agency: NIH/NIAAA (RC1 AA 019186)
Period: 09/30/09 – 02/28/13

Core or Program Director

Biostatistics Core
Lifespan/Tufts/Brown Center for AIDS Research
Agency: NIH/NIAID (P30 AI 42853)
Period: 09/01/98 – 06/30/17

Biostatistics Core
Alcohol Research Center for HIV
Agency: NIH/NIAAA (P01 AA 019072)
Period: 9/30/10 – 8/31/15

Biostatistics Program
AMPATH: Controlling and Preventing HIV/AIDS in Western Kenya
Agency: USAID (623-A-00-0-08-00003-00)
Period: 10/01/08 – 09/30/17


HIV-1 Drug Resistance in Different Subtypes
PI: Rami Kantor, MD, Brown University/Miriam Hospital
Agency: NIH/NIAID (R01 AI 066922)
Period: 09/01/08 – 08/31/13

A Stage-2 Cognitive-Behavioral Trial: Reduce Alcohol First in Kenya Intervention
PI: Rebecca Papas, PhD, Brown University
Agency: NIH/NIAAA (R01 AA 020805)
Period: 09/01/11 – 08/31/16

Optimizing Linkage and Retention to Hypertension Care in Rural Kenya
PI: Valentin Fuster, MD, New York University
Agency: NIH/NHLBI (U01 HL 114200)

Orphaned & Separated Children's Assessment Related to their Health & Well-Being
PI: Paula Braitstein, PhD, Indiana University
Agency: NIH/NICHD (R01 HD 060478)
Period: 9/15/09 – 07/31/14

Improving Linkage to HIV Care Following Release from Incarceration (LINCS)
PI: Josiah Rich, MD, Brown University / Miriam Hospital
Agency: NIH/NIDA (R01 DA 030778)
Period: 11/01/12 – 06/30/13

ASANTE Cardiovascular Pulmonary Disease Center of Excellence
PI: Sylvester Kimaiyo, MD, Moi University School of Medicine, Eldoret, Kenya
Agency: NHLBI/NIH (N01-HV-98221)
Period: 07/01/09 – 06/30/14

Neuromarkers of Age-Related Cognitive Decline
PI: Robert Paul, MD, University of Missouri – St. Louis
Agency: NIH/NINDS (R01 NS 052470)
Period: 07/01/07 – 06/30/12

Teaching Experience

I teach both introductory and advanced courses in biostatistics. Recently I developed a seminar on research methods in biostatistics, focusing on developing comparative reviews of statistical literature, design of numerical simulation studies, delivery of oral presentations, and writing of manuscripts and grant applications.

I have also developed and presented several short coures on methods for longitudinal data and missing data, delivered at scientific conferences, universities, and private companies.

Courses Taught

  • Applied Regression Analysis (PHP 2511)
  • Causal Inference and Missing Data (PHP 2610)
  • Research Methods in Biostatistics (PHP 2680)

Selected Publications

  • Liu T, Hogan JW, Wang L, Zhang S, Kantor R (2013). Optimizing diagnosis of HIV treatment failure in resource limited settings with selective gold standard viral load testing. Journal of the American Statistical Association, in press. (2013)
  • Little RJ, D'Agostino R, Cohen ML, Dickersin K, Emerson SS, Farrar JT, Frangakis C, Hogan JW, Molenberghs G, Murphy SA, Neaton JD, Rotnitzky A, Scharfstein DO, Shih W, Siegel JP, Stern H (2012). The prevention and treatment of missing data in clinical trials. New England Journal of Medicine 367, 1355-1360. [PMID 23024025] [DOI] (2012)
  • Scharfstein DO, Hogan JW, Herman A (2012). On the prevention and analysis of missing data in randomized clinical trials: The state of the art. Journal of Bone and Joint Surgery 94, 80-84. [PMID 22810454] [PMC 3393113] (2012)
  • Huang A, Hogan JW, Istrail S, DeLong A, Kantor R (2012). Global analysis of sequence diversity within HIV-1 subtypes across geographic regions. Future Virology 7, 505-517. [PMID 22822410] [PMC3400699] (2012)
  • Braitstein P, Siika A, Hogan JW, Kosgei R, Sang E, Sidle J, Wools-Kaloustian K, Keter A, Mamlin J, Kimaiyo S (2012). A novel care model to reduce early mortality and increase clinic retention among high-risk HIV-infected patients initiating combination antiretroviral treatment. Journal of the International AIDS Society 15, e7. [PMID 22340703] [PMC3297518] [DOI] (2012)
  • Daniels MJ, Roy J, Kim C, Hogan JW, Perri M (2012). Bayesian inference for the causal effect of mediation. Biometrics 68, 1028-1036. [PMID 23005030] (2012)
  • Bloomfield GS, Hogan JW, Keter A, Sang E, Carter EJ, Velazquez EJ, Kimaiyo S (2011). Hypertension and obesity as cardiovascular risk factors among HIV seropositive patients in western Kenya. PLoS ONE 6(7), e22288. (2011)
  • Tate DF, Sampat M, Harezlak J, *Fiecas M, Hogan JW, Dewey J, McCaffrey D, Branson D, Russell T, Conley C, Taylor M, Schifitto G, Zhong J, Daar ES, Alger J, Brown M, Singer E, Campbell T, McMahon D, Tso Y, Matesan J, Letendre S, Paulose S, Gaugh M, Tripoli C, Yiannoutsos C, Bigler ED, Cohen RA, Guttmann CRG, Navia B for the HIV Neuroimaging Consortium (2011). Regional areas and widths of the midsaggital corpus callosum among HIV-infected patients on stable antiretroviral therapies. Journal of NeuroVirology 17, 368-379. [DOI] (2011)
  • Gezmu M, DeGruttola V, Dixon D, Essex M, Halloran E, Hogan JW, Grobler A, Kim S, McDermott J, McKaig R, Neaton J (2011). Strengthening biostatistics resources in Sub-Saharan Africa: Research collaborations through US partnerships. Statistics in Medicine 30, 695-708. [PMID 21394746] (2011)
  • Su L, Hogan JW (2011). HIV dynamics and natural history studies: Joint modeling with interval-censored event times and infrequent longitudinal data. Annals of Applied Statistics 5, 400-426. (2011)
  • Tate DF, DeLong A, McCaffrey DE, Kertesz K, Paul RH, Conley J, Russell T, Coop K, Gillani F, Flanigan T, Tashima K, Hogan JW (2011). Recent clinical history and cognitive dysfunction for attention and executive function among human immunodeficiency virus-infected patients. Archives of Clinical Neuropsychology 26, 614-623. [PMID 21873325] [PMC 3243921] (2011)
  • Su L, Hogan JW (2010). Mixtures of varying coefficient models for longitudinal processes with continuous-time informative dropout. Biostatistics 11, 93-110. [PMID 19837655] [PMC2800163] (2010)
  • Kantor R, DeLong A, Kamle L, Muyonga S, Mambo F, Emonyi W, Chan P, Carter EJ, Hogan JW, Buziba N (2009). Misclassification of first line antiretroviral treatment failure based on immunological monitoring of HIV infection in resource-limited settings. Clinical Infectious Diseases 49, 454-462. [PMID 19569972] (2009)
  • Hogan JW (2009). Bringing causal inference into the mainstream (invited commentary). Epidemiology 20, 431-432. [PMID 19363354] Comment/Rejoinder 20, 931-932. (2009)
  • Hogan JW (2009). Considerations for sensitivity analysis with likelihood-based models (invited discussion of 'Missing data methods in longitudinal studies: a review' by JG Ibrahim and G Molenberghs). TEST 18, 59-64. Corrections 18, 607. (2009)
  • Hogan JW, Liu T (2008). Mediation analysis for intervention trials: objectives, models and inference (invited editorial). Health Services Outcomes Research Methodology 8, 77-79. (2008)
  • Su L, Hogan JW (2008). Bayesian semiparametric regression for longitudinal binary processes with missing data. Statistics in Medicine 27, 3247-3268. [PMC2581820] [PMID 18351709] (2008)
  • Roy J, Hogan JW, Marcus BH (2008). Principal stratification with predictors of compliance for randomized trials with two active treatments. Biostatistics 9, 277-289. [PMID 17681993] (2008)
  • Daniels MJ, Hogan JW (2008). Missing Data in Longitudinal Studies: Strategies for Bayesian Modeling and Sensitivity Analysis. Chapman & Hall. (2008)
  • Macalino G, Hogan JW, Mitty JA, Bazerman LB, DeLong AK, Loewenthal HG, Caliendo AM, Flanigan TP (2007). A randomized clinical trial of community based directly observed therapy (MDOT) as an adherence intervention for highly active antiretroviral therapy (HAART) among substance users. AIDS 21, 1473-1477. [PMID 17589194] (2007)
  • Rich JD, Hogan JW, DeLong AK, Mehrotra M, Reinert SE, Wolf F (2007). Low risk syringe sharing and re-use after syringe legalization in Rhode Island. Drug and Alcohol Dependence 89, 292-297. [PMID 17386980] (2007)
  • Hogan JW (2007). Discussion of 'Analysis of longitudinal data with drop-out: Objectives, assumptions and a proposal' by Diggle, Farewell and Henderson (invited). Applied Statistics 56, 530-531. (2007)