Principles of evidence based medicine
March 18, 2024Healthcare professionals are increasingly required to make clinical decisions based on the most reliable evidence. Evidence-based medicine (EBM) offers a systematic approach to clinical problem-solving, enabling the integration of the best available research evidence with clinical expertise and patient values. This paper aims to elucidate the concept of EBM and introduce the five-step EBM model: formulating answerable clinical questions, searching for evidence, critically appraising the evidence, assessing its applicability, and evaluating performance. Subsequent articles will delve into the principles and critical appraisal of randomized controlled trials, systematic reviews, and meta-analyses, while also providing a practical demonstration of the five-step EBM model using a real-life clinical scenario.
Table of Contents
WHAT IS EVIDENCE-BASED MEDICINE?
In recent years, the concept of evidence-based medicine (EBM), defined as the “integration of best research evidence with clinical expertise and patient values,” has gained increasing traction. EBM is a practice that entails a lifelong process of self-directed learning, driven by the need for critical information about clinical and healthcare matters arising from patient care. It acknowledges the dynamic nature of research literature, recognizing that what is considered the best practice today may evolve in the future. Staying abreast of these changes is facilitated by EBM tools, which enable the tracking and critical appraisal of evidence for integration into daily clinical practice.
In the field of pediatrics and child health, practitioners focus on the well-being of children, their families, and caregivers. This often involves encountering questions about diagnosis, prognosis, and treatment, necessitating the search for answers. EBM facilitates the incorporation of high-quality published evidence into clinical decision-making, alongside clinical expertise and the input of patients, families, or caregivers.
While evidence forms a crucial component of treatment decisions, EBM emphasizes that other factors, such as personal experience, judgment, skills, and, most importantly, patient values and preferences, must also be taken into account. Therefore, the practice of EBM aims to provide optimal patient care by integrating the best available evidence with patient preferences, while also incorporating expertise in clinical history-taking and physical examination. Figure 1 illustrates a typical flowchart of EBM, illustrating how knowledge and experience can be integrated with patient preferences and available evidence in clinical decision-making.
WHY PRACTICE EVIDENCE-BASED MEDICINE?
The primary goal of practicing EBM is to enhance the quality of care by identifying and promoting effective practices while eliminating those that are ineffective or harmful. EBM fosters critical thinking by requiring scrutiny of the effectiveness of clinical interventions, the accuracy of diagnostic tests, and the predictive power of prognostic markers, ensuring that their utility is established. It encourages clinicians to remain open-minded, willing to explore and adopt new methods that have been scientifically proven to be effective, while discarding those that have been shown to be ineffective or harmful. It is crucial for healthcare professionals to develop essential EBM skills, including the ability to locate, critically appraise, and integrate sound scientific evidence into their practice.
THE FIVE-STEP EBM MODEL
Practicing EBM entails following five fundamental steps: firstly, translating information needs into clear, answerable questions; secondly, locating the most relevant evidence to address these questions; thirdly, critically evaluating the evidence for its validity and relevance; fourthly, applying the findings of this appraisal to clinical practice; and finally, evaluating the outcomes of these practices.
Step 1: Formulating answerable clinical questions
One of the challenges in practicing EBM is transforming a clinical issue into a question that can be answered. When encountering a patient with a specific problem, numerous questions may arise for which we seek answers. These questions are often complex and unstructured, making them unclear in our minds. The practice of EBM should start with a well-crafted clinical question, requiring the ability to convert information needs into clear, answerable questions. Effective clinical questions should be focused directly on the issue at hand and should be answerable through medical literature research.
Sackett et al. proposed a framework for refining clinical questions, suggesting that a good clinical question should include four essential components: the patient or problem, the intervention, comparison interventions (if applicable), and the outcome(s) of interest. Therefore, an answerable clinical question should follow the PICO (Patient/Problem, Intervention, Comparison, Outcome/s) or PIO (Patient/Problem, Intervention, Outcome/s) format.
To illustrate the PICO/PIO concept, consider a scenario where a four-month-old baby is admitted to the hospital with viral bronchiolitis. The child’s symptoms worsen over time, prompting a question about whether administering corticosteroids could improve the child’s condition and shorten the hospital stay. The measure of improvement is decided to be the “clinical score.” The components of the clinical question would be:
Patient/Problem: 4-month-old baby with viral bronchiolitis. Intervention: Corticosteroids. Comparison: No corticosteroids. Outcomes: Clinical score, length of hospital stay.
Thus, a structured four-part clinical question could be formulated as follows: In a 4-month-old baby with viral bronchiolitis, does the administration of corticosteroids compared to no corticosteroids improve the clinical score and reduce the length of hospital stay?
Step 2: Locating the Evidence
Once you have formulated your clinical question, the next step is to find relevant evidence to answer it. While traditional sources like textbooks and journals can be useful, they are often disorganized or outdated. Seeking advice from colleagues or experts can also be helpful but the quality of information may vary. For quick, evidence-based answers, consider using reliable, summarized sources like Archimedes, Clinical Evidence, and BestBets. Additionally, electronic bibliographic databases offer a vast array of articles and can be searched efficiently with the right skills. Studies indicate that many healthcare professionals lack proficiency in conducting effective searches, leading to either too few or too many results. It is essential for healthcare professionals to undergo training in search skills to maximize the potential of retrieving relevant articles efficiently. This training can be provided through local library services or formal courses.
Basic Search Principles
- Formulate an Answerable Question: Transform your clinical problem into a clear, answerable question using the PICO/PIO format.
- Generate Keywords: Create a list of keywords based on your clinical question. For example, for the question regarding viral bronchiolitis and corticosteroids, keywords could include “viral bronchiolitis,” “corticosteroids,” “glucocorticoids,” “steroids,” “prednisolone,” “dexamethasone,” “clinical score,” and “hospital stay.”
- Choose a Bibliographic Database: Select a suitable online database. For most clinical practices, familiarity with one or two databases is sufficient. Recommended options include the Cochrane Library databases (including the Cochrane Database of Systematic Reviews), MEDLINE, EMBASE, and CINAHL. The Cochrane Library, updated quarterly, is a valuable resource for systematic reviews of healthcare interventions. MEDLINE, maintained by the National Library of Medicine, USA, is widely used and accessible through PUBMED, a free and regularly updated internet version.
- Utilize Secondary Sources: For articles on the effectiveness of interventions, consider using secondary sources such as Archimedes, Clinical Evidence, and BestBets in addition to the Cochrane Library databases. These sources provide summarized evidence.
- Search for Randomized Controlled Trials: The Cochrane Controlled Trials Register and MEDLINE are useful for finding randomized controlled trials and systematic reviews. SUMsearch is a search engine that allows direct searches of external databases focusing on clinical topics.
Conducting the Search
After identifying the keywords and databases, the next step is to execute the search. A straightforward method at the basic level is to combine individual words or terms using the Boolean operators “AND” and “OR.” Using “AND” retrieves articles containing both terms, while “OR” retrieves articles containing either term. This principle can be illustrated with a simple Venn diagram, where two overlapping circles represent the combination of terms using “AND” (fig 2A) and “OR” (fig 2B).
When the initial search yields too many articles (which is common), PUBMED offers a feature to limit the results. You can restrict your search by publication type (e.g., randomized controlled trials or review articles), date of publication, language, study population, and more. PUBMED also includes “Clinical queries,” a feature that facilitates evidence-based searching within the Medline database. “Clinical queries” is a preprogrammed filter that assists busy practitioners in accessing the best available evidence by providing quick access to reliable clinical studies related to therapy, diagnosis, etiology, or prognosis.
Example of a Basic Search Strategy
To find evidence for the earlier clinical question, we can search the Cochrane Database of Systematic Reviews and PUBMED using the keywords generated:
- Viral bronchiolitis
- Corticosteroids OR steroids OR glucocorticoids OR prednisolone OR dexamethasone
- Clinical score OR hospital stay
- Combine (1), (2), and (3) using AND. When this strategy was used to search the Cochrane Database of Systematic Reviews on December 10, 2004, four articles were retrieved, but only one was relevant. Other strategies to enhance the sensitivity and specificity of literature searches have been described by Sackett et al.
Step 3: Evaluating the Evidence
After gathering relevant articles, the next step is to assess the evidence for its validity and clinical relevance. The quality of available research articles varies, and using unreliable evidence could potentially harm patients or waste resources. Research evidence should be evaluated in terms of its validity, importance, and applicability to the patient or patients in question. Critical appraisal offers a structured approach to assessing research evidence in these areas. Developing critical appraisal skills involves learning to ask key questions about evidence validity and its relevance to specific patients. These skills can be acquired through tutorials, workshops, interactive lectures, and bedside learning. Various tools are available for appraising research articles. One recommended set of tools is developed by the Critical Appraisal Skills Programme (CASP) in Oxford, UK. CASP provides tools for appraising randomized controlled trials, systematic reviews, case-control studies, and cohort studies. These tools are user-friendly, freely accessible online, and offer a simple yet effective way to assess research articles. The critical appraisal of randomized controlled trials and systematic reviews will be discussed in detail in the next two articles of this series.
Step 4: Implementing the Evidence
After determining through critical appraisal that a piece of evidence is valid and significant, the next step is to assess its applicability to the individual patient or population. This involves considering the patient’s personal values and circumstances. It is essential to discuss both the efficacy and risks of the evidence thoroughly with the patient, parents, or both, to enable them to make an informed decision. This approach fosters a “therapeutic alliance” with the patient and their family, aligning with the core principle of EBM: integrating sound evidence with clinical expertise and patient values. When deciding to apply evidence, factors such as costs and the availability of the treatment in your hospital or practice should also be considered. The fourth article in this series will provide a practical demonstration of the considerations to take into account before applying research evidence.
Step 5: Assessing Performance
As we integrate EBM into our regular clinical practice, it’s important to regularly evaluate our approach and determine if there are areas for improvement in the previous four steps. Following the suggestions of Strauss and Sackett, we should assess whether we are formulating clear, answerable questions, efficiently finding quality evidence, accurately appraising the evidence, and effectively integrating clinical expertise and patient values with the evidence to develop a rational and acceptable management plan. Conducting formal audits of our performance may be necessary to demonstrate whether the EBM approach is enhancing patient care.
CONCLUSIONS
EBM strives to enhance the quality of care by integrating the best available research evidence with clinical expertise and the preferences of patients and parents. This article has outlined the five key steps for practicing EBM: formulating answerable clinical questions, searching for evidence, critically appraising the evidence, assessing its applicability, and evaluating performance.
