Doctor of Philosophy in Health Services Research
The doctoral program in Health Services Research is designed to provide advanced training in research methods and the substantive fields of health outcomes and quality research or health economics, with students concentrating in one of these fields. Qualified students will hold a master’s or bachelor’s degree in a field related to health services research, such as social or behavioral sciences, epidemiology, management, biological sciences, or the health professions. Students who complete the program will be prepared to assume leadership positions in academic and applied research settings and to pursue careers as independent investigators.
The program focuses on developing independent research scientists and scholars with excellent methodological skills, strong substantive knowledge of health care settings and policies, and an understanding of diverse interdisciplinary perspectives. The desired program outcome is for students to develop the necessary skills to conceive, design, and execute innovative research projects of high quality within the peer-review process. The program’s practical context will enable students to produce research findings that address challenging problems in clinical and management settings and to apply them in those settings.
Doctoral students in this program develop the ability to create theoretical and conceptual models that incorporate elements of frameworks from across fields. Conceptual frameworks available for study include social sciences (e.g., economics, sociology, anthropology); the management sciences (e.g., organizational theory, operations research); epidemiology and clinical sciences; and law and political science.
The PhD program engages the individual in a comprehensive, hands-on experience of health services research. This program ties classroom education closely to practical experience in research. Students are assisted, encouraged, and expected to begin dissertation research projects early in their studies. Students also have the opportunity to collaborate with senior faculty mentors in innovative research crucial to the improvement of health care delivery, treatment outcomes, and government policies.
Faculty are actively engaged in diverse areas of research, and the department is closely affiliated with two centers for health services research in the US Department of Veterans Affairs.
Find out: What is health services research?
In addition to the MS in Health Services Research competencies, upon completion of the PhD in Health Services Research, the graduate is able to:
- acquire knowledge of the context of health and health care systems, institutions and actors, and environment
- apply or develop theoretical and conceptual models relevant to health services research
- pose relevant and important research questions, evaluate them, and formulate solutions to health problems, practice, and policy
- use or develop a conceptual model to specify study constructs for a health services research question, and develop variables that reliably and validly measure these constructs
- describe the strengths and weaknesses of study designs to appropriately address specific health services research questions
- sample and collect primary health and health care data and/or assemble and manage existing data from public and private sources
- execute and document procedures that ensure the reproducibility of the science, the responsible use of resources, and the ethical treatment of research subjects
- demonstrate proficiency in the appropriate application of analytical techniques to evaluate health services research questions
- work collaboratively in teams within disciplines, across disciplines, and/or with stakeholders
- effectively communicate the process, findings, and implications of health services research through multiple modalities with stakeholders
- effectively translate knowledge to policy and practice
The core courses required for the PhD are an extension of the core courses required for the MS. The PhD includes 69 credits of coursework, approximately half in core courses and half in field courses and electives. Course listings are available in the SPH Bulletin and the PhD Guide.
Successful applicants hold a master’s degree in a field related to health services research (i.e., social/behavioral science, epidemiology, management, biological sciences, or health professions), or should have completed substantial coursework in these fields. Exceptional candidates holding a bachelor’s degree may be considered for an articulated MS/PhD program and should indicate their intention to enroll in an articulated program in an application to the MS program.
- A high level of past academic performance is expected, particularly in courses within the applicant’s undergraduate major/master’s field or his/her course work related to health services: recommended GPA > 3.5/4.0.
- Applicants must have completed a minimum of one semester of calculus and one semester of college-level statistics, each with a passing grade of B or better.
- All candidates will be required to submit GRE scores. Competitive applicants will have scores above the 50th percentile, with substantially higher scores in at least one area.
- Applicants must demonstrate competence in English. Candidates from countries where English is not the language of instruction must submit official results of either the TOEFL or IELTS in their SOPHAS application. Scores of 100 and above on the internet-based TOEFL, 600 and above on the paper-based TOEFL, or 7.0 total band score on the IELTS are required. Test scores may be up to two years old, and certain international applicants may request a waiver of the testing requirement; however, provision of current TOEFL or IELTS scores is preferred.
- At least 1–2 years of research experience is strongly encouraged.
- Students are required to have on admission, or to obtain through additional study, an understanding of the US health care system and its financing structure.
- Documentation of academic skills and interests from three references. The Admissions Committee will be especially interested in references from former undergraduate or recent graduate program professors and former supervisors of health- or research-related employment.
- Applicants must provide evidence of their preparedness, interest, and elementary understanding of the health services research field in the form of a concise personal statement. This statement should include the following elements:
- the candidate’s reason for studying health services research
- his/her desired particular area of study within the department
- the anticipated value of health services research training at Boston University to the candidate’s personal career plans
- the relevance of his/her prior education and experience
In order to be accepted, applicants must go through an interview, in person or via telephone, with the PhD Admissions Committee.
Students will be eligible to work as research assistants on funded research projects in the Department of Health Policy & Management and we expect that all students will do so in the course of their studies unless they are employed in health services research environments in other settings.
In addition to research within the department, many of the faculty members conduct research and have funding opportunities at our affiliated US Department of Veterans Affairs research centers: the Center for Health Quality, Outcomes, and Economic Research (CHQOER); the Center for Organization, Leadership, and Management Research (COLMR); and the Office for Productivity, Efficiency and Staffing (OPES).
The program director, department faculty, and program manager will actively assist students in obtaining funding and identifying employment opportunities on funded faculty projects. Some of these projects may take place outside of the department and include placement at the Boston University School of Medicine or Boston Medical Center.
In addition, some advanced students, with assistance from program faculty, have sought and obtained dissertation research grants from federal agencies and other funding sources.
Recent PhD Dissertation Abstracts
- Prostate Cancer Post-Treatment Quality of Life for Gay Men (pdf)
(Research supported by a grant from the National Cancer Institute)
- Sexual Healthcare for Vulnerable Populations (pdf)
- Impact of Genetic Counseling Communication Patterns (pdf)
(Research supported by an AHRQ dissertation grant)
- Social Disparities in Epilepsy Care and Outcomes (pdf)
- Developing and Testing Tools to Improve Patient Safety in Surgery (pdf)
- PhD Guide
- MS Thesis and PhD Dissertation Titles in Health Services Research (pdf)
- Selected Publications and Awards—HS Research Students and Graduates (pdf)
Technical Reports: Statistical Methods for Health Services Research
In Spring 2013 five doctoral students in Health Services Research researched, wrote, presented, and reviewed their work and others on specific statistical methods for health policy and management research. The final products are five technical reports, the first for our department. The students became ‘classroom experts’ on these topics and developed research, writing, and teaching skills from their work for this class. The reports contain examples and categorical references (indicating material for beginning learners and for those with prior knowledge of the topics).
Classification and Regression Tree Analysis
Classification and Regression Tree Analysis, CART, is a simple yet powerful analytic tool that helps determine the most \important” (based on explanatory power) variables in a particular dataset, and can help researchers craft a potent explanatory model. This technical report will outline a brief background of CART followed by practical applications and suggested implementation strategies in R. A short example will demonstrate the potential usefulness of CART in a research setting.
Meng-Yun Lin, MPH
This report is a brief introduction of Bayesian statistics. The first section describes the basic concepts of Bayesian approach and how they are applied to statistical estimation and hypothesis testing. The next section presents the statistical modeling using Bayesian approach. It first explains the main components of Bayes model including prior, likelihood function, and posterior. Then, it introduces informative and non-informative Bayes models. The last section provides an example of fitting Bayesian logistic regression in SAS. It illustrates how to program Bayes model and how to check model convergence.
Kyung Min Lee
Marginal structural modeling in health services research
Statistical inferences from observational studies are often subject to confounding caused by both observed and unobserved confounding variables. Conventional methods for controlling for confounding include applying statistical techniques such as stratification and multivariable regression analysis. In the presence of time-dependent confounders, however, such techniques may still lead to biased estimates. Marginal structural modeling (MSM) uses a multi-step estimation strategy to separate confounding control from the estimation of the parameters of interest, allowing the investigator to obtain unbiased estimates. Given that there are no unmeasured confounders and the probability of treatment is positive, the estimates of a marginal structural model can be interpreted as causal. This report serves as a starting point for researchers who wish to use MSM in their studies, providing an overview of the theory behind MSM and guidance for its implementation.
This tech report presents the basic concepts and methods used to deal with missing data. After explaining the missing data mechanisms and the patterns of missingness, the main conventional methodologies are reviewed, including Listwise deletion, Imputation methods, Multiple Imputation, Maximum Likelihood and Bayesian methods. Advantages and limitations are specified so that the reader is able to identify the main trade-offs when using each method. The report also summarizes how to carry out Multiple Imputation and Maximum Likelihood using SAS and STATA.
A Bayesian network is a representation of a joint probability distribution of a set of random variables with a possible mutual causal relationship. The network consists of nodes representing the random variables, edges between pairs of nodes representing the causal relationship of these nodes, and a conditional probability distribution in each of the nodes. The main objective of the method is to model the posterior conditional probability distribution of outcome (often causal) variable(s) after observing new evidence. Bayesian networks may be constructed either manually with knowledge of the underlying domain, or automatically from a large dataset by appropriate software.