Health Research Evidence Sources

Sources of Evidence

• Various sources provide evidence in health research, including peer-reviewed journal articles, books, conference proceedings, government documents, websites, clinical practice guidelines, and expert opinions.
• Peer-reviewed articles are published in scientific journals after expert review. Academic databases make finding peer-reviewed literature across multiple journals efficient.
• Benefits of peer-reviewed articles: High quality research and results, detailed topics, and easy access to millions of articles through online databases.
• Limitations of peer-reviewed articles: Users may not need detailed analyses (only descriptive statistics/basic measures). Access to articles may require a journal or database subscription.

Books

• Books can be valuable sources of information, but their currency (up-to-date nature) and author expertise are key considerations.

Conference Proceedings

• Conference proceedings document presentations at scientific conferences.
• They are sources of new research, but presentations aren't peer-reviewed.

Government Documents

• Government documents are informative on public health policy and guidelines.

Websites

• Websites provide various health research information (e.g., Wikipedia).
• Evaluating website information critically and checking for reputation ensure reliability.

Clinical Practice Guidelines

• Guidelines recommend best practices for diagnosis and treatment of specific conditions.
• They are developed by experts using the best available evidence.

Gray Literature

• This includes information not published in traditional academic journals or books.
• Often not peer-reviewed.
• Examples include: government reports, technical reports, conference abstracts, newsletters, and reports, and websites. Requires careful evaluation.

Hierarchy of Evidence

• Evidence hierarchy structures the levels of evidence from RCTs and systematic reviews, considered most robust, to expert opinions as least robust.
• Different levels of evidence are relevant for different types of research questions, e.g., therapeutic vs. prognostic studies.

Systematic Reviews & Meta-Analyses

• Critical appraisal and synthesis of all relevant studies on a specific research question.
• Identify patterns and inconsistencies, providing a quantitative summary of an intervention's or association's overall effect.

Randomized Controlled Trials (RCTs)

• Gold standard for assessing intervention effectiveness.
• Participants randomly assigned to intervention or control groups to compare outcomes.
• Minimizes bias through randomization, blinding participants/researchers from treatment assignment, ensuring any observed differences are due to the intervention.

Cohort Studies

• Track individuals over time to investigate associations between exposures and outcomes (e.g., disease development).
• Prospective cohort studies follow individuals without the outcome, collecting data and monitoring for outcomes.
• Retrospective cohort studies use existing data to identify exposures/outcomes. Assessing multiple outcomes, examining rare exposures, calculating incidence rates are possible but require long-term follow-up and are prone to loss to follow-up.

Case-Control Studies

• Compare individuals with a specific outcome (cases) to those without (controls) to determine exposure associations.
• Cases and controls are selected based on outcome presence/absence; exposure history is retrospectively assessed.
• Useful for rare diseases/outcomes with long latency but prone to recall bias and inaccurate past exposure recall.

Cross-Sectional Studies

• Collect data on exposures and outcomes at a single point in time from a representative sample.
• Provide estimates of disease prevalence, behaviours, or exposures, associations between variables.
• Cannot establish causality and temporal relationships as data is collected at a single point in time.

Case Reports and Expert Opinion

• Detailed descriptions of individual patients' unusual presentations/outcomes.
• Can generate hypotheses or identify rare adverse events, but don't provide strong evidence for causation due to their anecdotal nature.
• Expert opinions are perspectives/recommendations of recognized authorities, valuable for decision-making but should be considered with higher levels of evidence.

Types of Journal Articles

• Original Research: Most common, reports research data (Introduction, Methods, Results, Discussion).
• Short Reports/Letters: Concise reporting of original research; often for time-sensitive fields to inspire further investigation.
• Narrative Review: Summarises research on a specific topic and its current status, widely read and cited, citing over about 100 primary research publications.
• Case Studies: Specific instances of interesting phenomena; alert other researchers to potential phenomenon.
• Methodologies/Methods: Describe new experimental methods, tests, or procedures often improving over current knowledge.

Databases

• Bibliographic Databases: Contain references to published medical literature (e.g., PubMed, Scopus, Web of Science). Identifying relevant research articles, books or conference proceedings.
• Clinical Trial Databases: Provide information on clinical trials, currently recruiting/completed, useful for specific conditions/treatments.
• Patient Databases: Contain patient data (medical history, medications, test results); used to track outcomes and identify disease patterns.
• Genetic Databases: Contain genetic information (sequence, location), used in disease analyses and new treatment development.
• Imaging Databases: Contain medical images (X-rays, MRIs, CT scans); for diagnosing and monitoring diseases.
• Drug Databases: Contain drug information (side effects, interactions), crucial to safe medication prescription and management.

What Does "Maintained by" Mean?

• Organizations responsible for acquiring, organizing, storing, maintaining, and making data accessible. Illustrated by example: Web of Science, a bibliographic database, is maintained by Clarivate Analytics.

Indexing

• A journal is considered indexed when included in a database that catalogs/indexes scholarly publications. This allows researchers and others easy access to articles.

Journal Types: Open Access vs. Subscription

• Open Access: Accessible to all regardless of financial resources. Can increase publication visibility/impact and reduce "publish or perish" pressures. However, costs of publishing in open-access journals may be prohibitive.
• Subscription Journals: Typically more rigorous, longer history, more established, and indexed in major databases.
  • Can be expensive, limiting access for readers, especially in developing countries.

Impact Factor & Quartiles

• Impact Factor (IF): Measures how frequently a journal's articles are cited in other research. Calculated by dividing the number of citations to articles in the past two years by the number of articles published in the journal during the same period.
• Quartiles: Journals are ranked into quartiles (Q1-Q4) based on IF or other bibliometric measures. This helps determine a journal's rank within its field. Q1 is the top 25% and indicates high impact, with Q4 as the bottom 25%.