• Tidak ada hasil yang ditemukan

CHAPT ER 8

Dalam dokumen Evidence-Based Public Health (Halaman 196-200)

w

Searching the Scientific Literature and Using Systematic Reviews

Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information?

T.S. Eliot

A

s you develop an issue statement and begin to understand the epidemio- logic nature of a particular public health issue along with the interven- tion options, the scientific literature is a crucial source of information on what works. Because of the considerable growth in the amount of information available to public health practitioners, it is essential to follow a systematic approach to literature searching. The underpinnings of an evidence- based pro- cess rest largely on one’s ability to find credible, high- quality evidence as effi- ciently and exhaustively as possible. A systematic searching process also helps ensure that others can replicate the same results. With modern information technologies, virtually all public health workers have an excellent opportunity to find valuable information quickly. Published information resources are now increasingly available for anyone with an Internet connection, enabling pro- fessionals outside major institutions to perform professional and thorough searches for needed resources.

This chapter provides guidance on how to identify existing evidence of effective interventions using systematic reviews and online resources, and if necessary how to conduct a primary search of the scientific literature. It focuses on the importance of a literature search, where to search, how to find evidence, and how to organize the results of a search. Evaluation of the qual- ity of the evidence is covered in other chapters (primarily chapters 3 and 11).

( 178 ) Evidence-Based Public Health BACKGROUND

As noted in chapter  1, there are many types and sources of evidence on public health programs and policies. Scientific information (the “sci- entific literature”) on theory and practice can be found in textbooks, government reports, scientific journals, and policy statements and at sci- entific meetings. Three levels of reading the scientific literature have been described:  (1)  browsing— skimming through actual books and articles, looking for anything of interest, and browsing topic- related sites on the Internet; (2) reading for information— approaching the literature in search of an answer to a specific question; and (3) reading for research— reading to obtain a comprehensive view of the existing state of knowledge on a specific topic.1 In practice, most of us obtain most of our information through browsing.2,3 However, to conduct a literature review for building evidence- based programs efficiently, it is important to take a more struc- tured approach. We focus primarily on journal publications here because they have gone through a process of peer review to enhance the quality of the information and are the closest thing to a gold standard that is available (see chapter 3).

When conducting a search of the scientific literature, there are four broad categories of publications to consider for evidence- based decision making (Figure 8.1):

1. Original research articles: These are the papers written by the authors who conducted the original research studies. These articles provide details on the methods used, results, and implications of results. A  thorough and comprehensive summary of a body of literature will consist of careful read- ing of original research articles, particularly when a topic area is changing rapidly or there are too few original articles to conduct a review.

Practice guidelines

Systematic reviews Meta analyses

Narrative reviews

Individual Studies

Figure 8.1: Hierarchy of scientific evidence for evidence- based decision making.

se arching t he s cien t ific Li t er atur e ( 179 )

2. Narrative review articles: These provide a descriptive summary of what is known on a particular topic. A narrative review article presents a summary of original research articles. The Annual Review of Public Health is an excel- lent source of review articles on a variety of topics (http:// arjournals.annu- alreviews.org/ loi/ publhealth). A  limitation of narrative review articles is that they do not always follow systematic approaches, a practice that some- times leads to selection bias and inconsistent results.4

3. Systematic review articles: These review articles involve a detailed, struc- tured, and exhaustive search of original research articles with the goal of reducing selection bias and synthesizing all relevant articles on a particu- lar topic.4– 6 Systematic reviews often feature a quantitative synthesis of results, or meta- analysis, to produce a summary statistical estimate of the measure of association or effect. For example, the Cochrane Collaboration, an international organization of clinicians, epidemiologists, and others, has produced quantitative reviews on the effectiveness of various health care interventions and practices covering a wide range of subjects (www.

cochrane.org). A more detailed discussion of finding evidence of effective interventions using systematic reviews is provided later in this chapter.

4. Guidelines: Practice guidelines are formal statements that offer advice to clinicians, public health practitioners, managed- care organizations, and the public on how to improve the effectiveness and impact of clinical and public health interventions. Guidelines translate the findings of research and demonstration projects into accessible and usable information for pub- lic health practice. There are several examples of useful guidelines. The ter- minology used within them differs across the globe. Thus, in the European Community, directives are stronger than recommendations, which are stronger than guidelines.7 No such hierarchy exists in North America.

Review articles and guidelines often present a useful shortcut for many busy practitioners who do not have the time to master the literature on multiple public health topics.

In addition to the type of publication, timeliness of scientific information is an important consideration. To find the best- quality evidence for medical decision making, Sackett and colleagues recommended that practitioners burn their (traditional) textbooks.8– 10 Although this approach may seem radi- cal, it brings to light the limitations of textbooks for providing information on the cause, diagnosis, prognosis, or treatment of a disorder. To stay up to date in clinical practice, a textbook may need to be revised on a yearly basis.

Though considered to provide more timely scientific findings, research and publication of results in a journal are a deliberative process that often takes years from the germination of an idea, to obtaining funding, carrying out the study, analyzing data, writing up results, submitting to a journal, and waiting out the peer- review process and publication lag for a journal.

( 180 ) Evidence-Based Public Health

The number of scientific publications has increased dramatically since the 1940s.11 In Medline alone in 2015, there were 806,326 new citations, 5,618 journal titles indexed, and 2.8 billion searches conducted.12 There are an estimated 25,000 to 40,000 scientific journals in the world, publishing approximately 1.4 million new research papers each year. About 8% to 10%

of these are published in open access journals, and only about 20% of scien- tific articles are available free of charge.13 To assimilate even a fraction of this large body of evidence, the practitioner needs to find ways to take advantage of the vast amount of scientific information available and to find informa- tion quickly. Of increasing interest to health professionals is the ease with which this literature may be accessed by those not directly supported by major library resources. Consequently, there is interest in open access availability of scientific publications. A recent study reported that in 2011, 12% of articles became immediately available, and an additional 5% become available within 12 months of publication.14 Therefore nearly 20% of articles may be readily accessible, and professionals may also use the PubMed author information to obtain the author’s email address for direct requests of articles. With easy access to abstracts and an increasing ability to obtain research articles, pub- lic health professionals— regardless of their institutional resources— may be able to actively work with the scientific literature in their areas of concern.

Methods for searching the literature have changed dramatically. Thirty years ago, a practitioner wishing to find information on a particular topic would speak with a librarian and inform him or her of the type of information being sought, perhaps provide a sample article, and help in selecting some key words. The librarian would run the search, consult with the practitioner as to whether it captured the desired types of articles, modify the search as needed, rerun it, consult with the practitioner again, and so forth. This whole iterative process could take weeks. Current practitioners with an Internet connection can now search for relevant information from the world’s scientific literature and, with training and experience, can discern relevance and quality so as to improve the practice of public health. There also are numerous online training modules on how to search the literature, such as those at www.ebbp.org or www.nlm.nih.gov.

FINDING EVIDENCE FOR EFFECTIVE INTERVENTIONS USING SYSTEMATIC REVIEWS AND ONLINE RESOURCES

Systematic Reviews

As noted earlier, systematic reviews are syntheses of comprehensive collec- tions of information on a particular topic. Given the huge number of new scientific articles published each year and the fact that no single study could provide a conclusive answer with regard to an intervention’s effectiveness,

Dalam dokumen Evidence-Based Public Health (Halaman 196-200)

Dokumen terkait