Selective
2
Critical Review of the Literature
A critical review of the literature typically poses an unresolved question and attempts to answer that question from evidence published in the medical literature. A critical review can take other forms as well, such as the analysis of an issue in health policy or biomedical ethics. The research can be initiated by you or by the sponsoring faculty member, as long as you make an intellectual contribution to the project.
- Identify a specific unresolved question relevant to the practice of clinical medicine, its scientific base, or the administration, regulation, or financing of medical care.
- Conduct a comprehensive, systematic search to identify the existing literature.
- Critically review this literature with particular attention to methodological strengths and weaknesses in the publications.
- Summarize the current status of this question with particular attention to areas of uncertainty.
- Recommend a logical next step for research. A hypothesis to test and suggested experimental approaches would be appropriate.
- Faculty sponsor. You will work on your research
with guidance from a faculty sponsor. Any regular or clinical faculty member
in any department at any WWAMI university is eligible to be a faculty sponsor.
The sponsor’s role is to help you plan your study, to meet with you
as necessary during the execution of the project, and to read and provide
feedback on your final paper. Your sponsor must sign and approve your research
proposal, your third year progress report, and your final paper.
The sponsor you choose and the relationship you build will be among the most important considerations in making this experience successful, enjoyable and valuable. Sponsors need to be:
- Interested in your topic (though not necessarily expert in it)
- Familiar with the methods you are likely to use in your study
- Available to you through phone, email or scheduled meetings.
The ideal sponsor will also be:
- A role model for qualities you seek to emulate in your professional development.
- Enthusiastic about working with you.
- Experienced with the clinical problem, scientific question or policy issue you will be studying.
- Someone with skills and knowledge that complement those you bring to the project.
In your search for your sponsor, start first with people you know: professors, guest lecturers, and preceptors. Talk to other trainees, co-workers from prior jobs, old professors and young residents. This is professional networking at work. You might also consult departmental Web sites and faculty interest databases such as the Community of Science (COS). When you first contact a potential sponsor, be prepared to explain a little about the III requirement; do not assume faculty members know what it is.
- Research proposal. A written proposal outlining
your research plan (download PDF form [92K])
must be submitted for review to the III Committee. This review is primarily
for feasibility and secondarily for scientific soundness. Proposals are reviewed
on the first Thursday of each month throughout the school year, and you will
receive written notice about your proposal shortly after it is reviewed. The
Committee will either approve your proposal, ask for further information,
or ask that you meet individually with a Committee member.
A successful proposal (and a successful study) begins with a simple, clear purpose. This purpose should be reflected in each of the components of the study described below. The purpose will dictate which databases to search, which articles to select and what information to use from the articles you read.
The proposal should be brief, generally 1-2 typed pages, but should provide enough information to give the committee a good idea of what you plan to do. The III Committee includes members from a variety of clinical and basic science departments, so write your proposal for a broad audience. If information can be best presented in non-narrative form (graph, bulleted list, flow diagram, etc.), by all means do so.
Below are some guidelines for what to include. Because each study is different, not all items will be pertinent to every study.
Background and rationale. Provide a brief introduction to the problem you are investigating. This might include:
- What is the research problem?
- Why is the problem important?
- How will your study contribute to this field of knowledge?
Question or hypothesis to be investigated. A hypothesis is a testable assertion about the relationship between variables in your study. The specificity of the research question should be tailored to the quantity of literature available, as described below under “Selection criteria”. Provide a preliminary estimate of the number of articles available that address your research question.
Search strategy. What databases will be searched? What key words will be used? How else will pertinent studies be located?
Selection criteria. Describe the types of studies you will include in terms of population, study design, dates of publication, which outcomes were studied and how they were defined and measured, etc. For example, if your question concerns the quality of life following two surgical treatments for a certain cancer, will you only include studies comparing the procedures head-to-head, or will you also include case series that describe outcomes for each procedure alone? Will you only include studies that used the same quality of life questionnaire? What range of follow-up times will you include? Will you only include studies published after 1994 when there was a sea change in how one of the procedures was performed? If there were pertinent animal studies, would they be included? Will studies be included that don’t take into consideration confounding factors such as patient age or disease severity?
Your answers to such questions will depend, in part, on the amount of literature available. If there is a huge literature, you should focus your question to a subset of studies homogeneous with respect to population, study design, and other research methods. If the body of literature is small, you’ll have to use studies that are more heterogeneous.
Variables. Indicate how, for purposes of the review, you will define the exposures and outcomes of interest and how outcomes are measured. If there are important confounding variables, list these as well.
Strategy for data synthesis. Describe how you will integrate information across studies to answer the research question. You do not need to perform a meta-analysis or mathematically pool data across studies.
Timetable. As best you can, lay out a realistic timetable for completing the key steps of the project.
- Deadlines
E2003 Dates
E2004 Dates
- Final paper.
The title should be brief and narrowly focused. It will become a permanent part of your curriculum vitae, so give it considerable thought. It does not need to be identical to the title on your proposal.
The abstract is a succinct summary of the paper's methods and results.
The introduction provides a rationale for why the study was done. Think of the introduction as a funnel. It can begin with a broad introduction to the issues, but quickly narrows its focus to the specific research problem being investigated. It should convince the reader that there's an important research problem that has been addressed in the literature which now calls for systematic review. By the end of the introduction the reader should understand what your study will be about and why it's an important study to do.
The methods section ought to contain enough detail to enable another investigator to replicate your study. This should include how articles were selected (inclusion and exclusion criteria), what measurements were abstracted directly or calculated from available data, and, how these data were synthesized.
The results section is the meat of the paper. Typically, the first results presented describe the sample of articles on which the remaining results are based. These might include authors, year of publication, population studied, etc. presented in the form of a table. Following this, report the data that bear most directly on the primary hypothesis of the study. This, too, can often be summarized in a table which shows the key results of each study.
The text should refer to tables and graphs but should not reiterate the information contained in them. The text can, however, guide the reader toward the message contained in the table or graph: "Table 2 shows that the treatment and control groups had similar hospital lengths of stay within studies, although length of stay varied considerably among the studies."
The discussion should be an interpretation of the results. Begin by providing an answer to the research question posed earlier. Include the weaknesses of the studies reviewed and how those weaknesses could influence the results of the individual studies and the result of your review. Comment on the validity and generalizability of your review. After taking the weaknesses into consideration, what is the meaning of the study for the field of medicine? What questions has your review resolved? What questions or directions for future research has your review generated?
The following guidelines are given to reviewers:
Section Criteria of Excellent Required Question Important and interesting/creative, clearly focused in terms of population, intervention/exposure and outcome. Important or interesting/creative, somewhat focused in terms of population, intervention/exposure and outcome; no fatal flaws in question. Method Clearly describes excellent strategy for literature search, study inclusion criteria and methods of data synthesis. Describes strategy for literature search, study inclusion criteria and methods of data synthesis. Results Demonstrates clear understanding of relevant concepts and thorough literature review. Well-articulated and makes interesting or creative points. Results well-articulated, appropriate use of tables and figures. Appropriate synthesis of results across studies, with attention to methodological heterogeneity and quality. Demonstrates general understanding of relevant concepts. Results described appropriately (i.e., reviewed studies relevant to research question; not missing important studies; not drawing inappropriate conclusions or going beyond the data, enough detail provided to understand conclusions without just listing studies, results are related specifically back to question) Discussion/Conclusions Draws interesting implications, strong understanding of the results in relation to the literature, clearly articulates the limitations of the review and future directions suggested by it. Demonstrates adequate understanding of the results in relation to the question. Articulates limitations of the review. Presentation Well-organized, readable, clear, style appropriate for refereed medical journal, almost no spelling or grammatical errors. Reasonable organization and readability, formatted in style for refereed journal, few spelling or grammatical errors.