Secondary Logo

Journal Logo

Conducting a successful systematic review of the literature, part 1

Riesenberg, Lee Ann PhD, RN; Justice, Ellen M. MLIS, AHIP

doi: 10.1097/01.NURSE.0000444728.68018.ac
Department: RESEARCH CORNER
Free

Lee Ann Riesenberg is director of medical education outcomes and research in the department of anesthesiology and associate professor of anesthesiology at the University of Alabama at Birmingham. Ellen M. Justice is a community health librarian at Christiana Care Health System in Newark, Del.

The authors have disclosed that they have no financial relationships related to this article.

THIS, THE FIRST OF two articles, is aimed at helping nurses conduct and publish a systematic review of the literature. Nurses know that the volume of published research is growing exponentially. What's missing is high-quality systematic reviews of this research. We believe that if nurses follow our proposed 10-step process, they'll be on the path to successful publication.

Back to Top | Article Outline

Understanding literature reviews

Systematic reviews of the literature have replaced expert commentaries and narrative reviews as the gold standard for summarizing the evidence found in research literature.1 Expert commentaries or narrative reviews may display the biases of the authors. In contrast, systematic reviews use explicit, rigorous methods to identify all relevant articles, critically appraise the articles, and synthesize the evidence. Following this rigorous process helps minimize bias and improve the reliability and accuracy of the conclusions. This makes it easier for healthcare professionals to trust the evidence presented, apply it to patient care, and write clinical practice guidelines.

According to a 2011 Institute of Medicine report, today's healthcare decision makers rely on systematic reviews to inform clinical decisions.2 Well-conducted systematic reviews “identify, select, assess, and synthesize” the evidence that makes clear what's known and not known about drugs, devices, education and training methods, and healthcare practices.2

As noted by Hemingway and Brereton,1 a systematic review of the literature is aimed at

  • identifying all relevant evidence
  • assessing the quality of each study or report
  • synthesizing the findings from individual studies or reports in an unbiased manner
  • interpreting the findings and presenting a balanced and impartial summary of the findings, with adequate consideration of any flaws.

Conducting a systematic review can be arduous and time-consuming, but done correctly, it can answer important questions and increase our understanding of a topic. All systematic reviews should start with a peer-reviewed protocol (or plan),1,3 which will be described in more detail later. The protocol will allow for future replication and transparency of the methods used, increasing the chances of publication.

A qualitative systematic review summarizes primary studies but doesn't combine the study results using statistical methods. A systematic review that uses quantitative statistical methods to combine the results of two or more primary studies is called a meta-analysis.4A mixed-method systematic review includes both qualitative and quantitative data in the final manuscript.1

In this article, we describe the first 5 of the 10 steps of conducting a qualitative systematic review that follows the latest guidelines, including the Institute of Medicine standards2 and the Preferred Reporting Items for Systematic Reviews and Meta-analysis statement.5 (See 10 steps to a systematic review.)

Back to Top | Article Outline

Step 1. Develop a clear question.

First, the question a systematic review aims to answer needs to be clearly stated. Without a clear question, it will be difficult to run a precise literature search. In addition, lack of clarity in the question may result in spending extra time clarifying the question and rerunning searches based on the evolving question.

Several frameworks can be used to develop a focused review question. One good framework is the SPICE model.6 SPICE stands for:

  • Setting. Where? (for example, in community hospitals)
  • Perspective. For whom? (for example, clinical nurses)
  • Intervention. What? (for example, bedside shift-to-shift handoffs)
  • Comparison. Compared with what? (for example, other types of shift-to-shift handoffs)
  • Evaluation. What conclusions can be drawn about the methods of handoffs?

Based on the examples given in the SPICE model above, a clear question would be: For clinical nurses (P) working in community hospitals (S), what's the evidence (E) to support the use of bedside shift-to-shift handoffs (I) compared with other types of shift-to-shift handoffs (C)?

Two other frameworks are the PICO (population, intervention, comparative intervention, outcomes) model and the PEO (population, exposure, outcomes) model for qualitative studies.3

Back to Top | Article Outline

Step 2. Conduct an initial literature search.

Once a clear question has been formulated but before assembling a complete team, the lead author should identify a collaborating librarian. This librarian can conduct the initial literature search to determine if a systematic review has already been published, or is in progress, on the chosen question.3 A preliminary search will be needed to answer this question; no one wants to do all the work required for a systematic review only to find that someone else has already published the answer. In this case, the question may be revised to address a different aspect of the topic.

One way to improve the quality of this and subsequent searches is for the lead author to provide the librarian with two to four relevant research articles on the topic of interest. This will help the librarian identify more search terms by looking at the indexing of these articles. Many databases can be searched. (See Search these databases to identify systematic reviews.)

Table

Table

Back to Top | Article Outline

Step 3. Assemble a team.

Once the initial search has been completed and the lead author and librarian both feel reasonably certain that the proposed systematic review addresses a unique question that hasn't already been addressed, it's time to identify and organize a complete systematic review team. The team should include people who have the following skills, attributes, and resources:

  • content expertise relevant to the topic or the ability to acquire this knowledge
  • enough protected or personal time to commit to the required work
  • detail-oriented personality
  • expertise in searching the literature in various databases.

Besides nurses, some people who may have these attributes include librarians, undergraduate and graduate students, allied health professionals, and residents and attending physicians interested in healthcare research and coauthoring duties.

In the authors' experience, including a librarian on a systematic research team has many advantages. Medical librarians are trained searchers and have knowledge of many databases and corresponding search techniques. They also appreciate the rigors of well thought-out search strategies and know how to save and document the details of the searches conducted. We believe that the librarian should be offered coauthorship because he or she is uniquely qualified to complete many of the tasks required.

As well as conducting expert searches and providing the corresponding documentation, the librarian may contribute in other ways, such as searching the grey literature, which includes government reports and other documents that aren't usually indexed in popular bibliographic databases; obtaining peer review by other librarians of the search strategies; writing the literature review methods section; and providing a substantive, critical review, and edit of the final manuscript.

Residents and attending physicians, nurses, and other healthcare professionals interested in research and authoring are an essential part of the team because they bring subject expertise. In addition, they can envision how the systematic review would benefit healthcare providers, patients, and those developing clinical practice guidelines.

Many undergraduate and graduate science and healthcare students have energy, enthusiasm, and interest, making them assets to the systematic review team. They may themselves benefit from involvement in such a rigorous process and, depending on their level of contribution, the opportunity to become coauthors.

The lead author should be prepared to organize team meetings. We recommend weekly meetings at the beginning, but over time, meetings can move to every other week and then monthly. Meetings can be conducted in person, by telephone conference, or through web-based meeting formats. One team member should take detailed notes and summarize the content of the meeting, especially noting tasks assigned and timeframes for due dates. Each systematic review team must be committed to meeting regularly, asking for clarification as needed, and performing tasks as assigned.

Back to Top | Article Outline

Step 4. Create a peer-reviewed protocol.

The gold standard is to use a peer-reviewed protocol.1,3 This protocol may be defined as a well thought-out document that describes in detail a plan that, if followed, increases the likelihood that “scientifically valid, transparent, and reproducible results” will be produced.2

The protocol is created after the initial search has been completed and the librarian has confirmed that no systematic reviews have been published on the selected topic. Equivalent to a standard methods section of a manuscript, the protocol:

  • includes the question
  • lists inclusion and exclusion criteria
  • details search strategies
  • outlines the process that will be used to select articles for inclusion
  • describes the process that will be used to develop the abstraction form and database
  • details how articles will be assessed for methodological quality.

The protocol should be proofread by at least one other healthcare professional who isn't part of the team to ensure it's complete and understandable. This colleague should review the protocol to make sure it's clearly written and includes enough detail so that others could replicate the systematic review.

Back to Top | Article Outline

Step 5. Conduct comprehensive literature searches.

High-quality comprehensive literature searches are the foundation upon which the systematic review will be built. Both the initial search and the other literature searches run to identify all relevant articles on the topic should be conducted by an experienced searcher, such as a librarian. For more details, see What a librarian needs.

The selected librarian should have advanced search skills and knowledge of various databases and platforms (such as Ovid and EBSCOhost). To help the searcher identify appropriate terms for the searches, the team should brainstorm as many terms as possible (that is, synonyms and other related terms), based on the review's central question. The librarian will use these terms for term harvesting. This is the process of using relevant terms and identifying appropriate controlled vocabulary that's used to index an article so that it can be retrieved easily from a database. Controlled vocabulary can be different from database to database (for example, from CINAHL to MEDLINE). Term harvesting also includes identifying appropriate keywords and keyword phrases. The searcher will take advantage of various search techniques, such as using controlled vocabulary (that is, Medical Subject Headings [MeSH] in MEDLINE), keywords, Boolean operators, and other advanced techniques such as truncation and adjacency.2,7 (See Coming to [research] terms.) The searcher needs to read the help sections of databases to learn about the specific advanced searching techniques available.

After term harvesting, the librarian will develop search strategies to optimize recall. According to Walters, recall describes how effective the search strategy is in identifying most of the relevant articles.8 On the other hand, precision takes into account the inclusion of relevant articles as well as the exclusion of nonrelevant articles.8 The searcher should be prepared to sacrifice precision to maximize recall because the goal of a systematic review is to identify all relevant articles for inclusion.9 This means that “search results will likely be large and contain a significant number of false positives”; however, this is preferable to missing relevant studies.9

Research has shown that no single database provides comprehensive coverage of healthcare topics.10–12 For this reason, multiple techniques must be used to identify relevant literature, including searching multiple databases and reviewing references in books and journal articles. At a minimum, the librarian should search two databases (such as MEDLINE, CINAHL, PsycINFO), but if appropriate, more databases should be searched to maximize recall and lessen the risk of bias in the authors' conclusions.

Back to Top | Article Outline

Forging ahead

Systematic reviews of the literature are a great way to make a significant contribution to the evidence around a specific topic and a perfect way to get published. In part 2 of this series, we'll present more about how to successfully conduct a systematic review.

Back to Top | Article Outline

10 steps to a systematic review

  1. Develop a clear question.
  2. Conduct an initial literature search to determine if a systematic review on the topic has already been published.
  3. Assemble a team.
  4. Create a peer-reviewed protocol.
  5. Conduct comprehensive literature searches.
  6. Review the search results and select articles to be included.
  7. Develop the abstraction database.
  8. Train abstractors and abstract study details from the included articles.
  9. Conduct quality assessment of included studies' methodologies.
  10. Prepare the final manuscript.
Back to Top | Article Outline

Coming to (research) terms

  • Boolean operators are word commands that allow a relationship between search terms to be created. Common operators are OR, AND, and ADJn.
    • - OR allows the retrieval of bibliographic records containing any of the terms.
    • - AND allows the retrieval of only those records that include all of the terms.
    • - Adjacency (ADJn) is a positional operator that finds records containing the terms in any order within a specified number (N) of words from each other.
  • Truncation lets a root word be used to find alternate suffixes.
Back to Top | Article Outline

What a librarian needs

To participate effectively in a systematic review team, a librarian may need the following:

  • initial detailed discussion of the purpose of the systematic review, the focused question that will be addressed, preliminary terms to consider for search strategies, and expected timeframes
  • inclusion in as many team meetings as possible
  • access to at least one primary investigator for questions and/or clarification
  • a reasonable time period in which to harvest terms and complete searches in multiple applicable databases and to document comprehensive search strategies
  • opportunity to write up the search strategies for the methods section and review manuscript before final submission
  • if applicable, inclusion in grants to help pay for interlibrary loan requests.
Back to Top | Article Outline

REFERENCES

1. Hemingway P, Brereton N. What is a systematic review? 2009. http://www.medicine.ox.ac.uk/bandolier/painres/download/whatis/Syst-review.pdf.
2. National Research Council. Finding What Works in Health Care: Standards for Systematic Reviews. Washington, DC: The National Academies Press; 2011. http://www.nap.edu/catalog.php?record_id=13059.
3. Bettany-Saltikov J. Learning how to undertake a systematic review: part 1. Nurs Stand. 2010;24(50):47–55.
4. Haase SC. Systematic reviews and meta-analysis. Plast Reconstr Surg. 2011;127(2):955–966.
5. Moher D, Liberati A, Tetzlaff J, Altman DG The PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6(7):e1000097. http://www.plosmedicine.org/article/info%3Adoi%2F10.1371%2Fjournal.pmed.1000097.
6. Booth A. Clear and present questions: formulating questions for evidence based practice. Library Hi Tech. 2006;24(3):355–368. http://testbed.cis.drexel.edu/I515/Booth%20Problem%20Formulation.pdf.
7. Bettany-Saltikov J. Learning how to undertake a systematic review: part 2. Nurs Stand. 2010;24(51):47–56.
8. Walters WH. Comparative recall and precision of simple and expert searches in Google Scholar and eight other databases. portal: Libraries and the Academy. 2011;11(4):971–1006. http://www.litguide.press.jhu.edu/journals/portal_libraries_and_the_academy/portal_pre_print/archive/articles/11.4walters.pdf.
9. Folb B, Hartman L, Klem M, Ratajeski M, Saleh A, Wessel C. Systematic Review Workshop: The Nuts and Bolts for Librarians, November 7-9, 2011. Health Sciences Library System, University of Pittsburgh, Pittsburgh, PA. (PowerPoint Presentation.)
10. Haig A, Dozier M. BEME Guide no 3: systematic searching for evidence in medical education—Part 1: sources of information. Med Teach. 2003;25(4):352–363.
11. Maggio LA, Tannery NH, Kanter SL. Reproducibility of literature search reporting in medical education reviews. Acad Med. 2011;86(8):1049–1054.
12. Savoie I, Helmer D, Green CJ, Kazanjian A. Beyond Medline: reducing bias through extended systematic review search. Int J Technol Assess Health Care. 2003;19(1):168–178.
© 2014 by Wolters Kluwer Health | Lippincott Williams & Wilkins.