Evidence-based Medicine

What is evidence based medicine (EBM)? • The process of finding, appraising and using the best available external clinical evidence, in combination with individual clinical expertise and professional judgement, to make informed clinical decisions about the health and care of individual patients • You are essentially applying scientific research to everyday practice • It is a systematic (organised and non-random) approach that seeks to improve the effectiveness of health care decision-making • It involves a cultural shift away from opinion and past traditional practice towards greater use of scientific research and reproducible evidence • However, even with having a large number of evidence, professional judgement is still required to decide if the evidence is applicable to the individual patient. • There is a daily need for valid and up to date information by healthcare professionals, to enable pharmacists to choose the best therapeutic interventions for patients and make the right choices on management and treatment, to offer patient centred care • The type and complexity of medicine is Why is evidence based medicine impor- growing so understanding the evidence tant? behind these can help us to assess the value of new treatments and contribute to designing future health developments • It highlights areas where research is inadequate. Doing a literature search may not bring you the info you are looking for since it doesn't exist, so encourages research in this area • It overcomes some limitations associat1 / 16 ed with more traditional sources of information eg textbooks which are harder to update, information that has come about because of expert opinion, didactic education (intention to teach rather than entertain, in a way that is fixed and unwilling to change) • It reduces the gap between research and clinical practice • You can't always find the right evidence • You can't always access the evidence you need due to restrictions on publication eg copyright • The process of evidence based medicine is time consuming What are limitations of evidence based • There is a danger of information overmedicines? load which can lead to some health risks • Evidence based medicine doesn't always provide the perfect answer to a specific question, and you may still need to exercise professional judgement to come to a decision • Information overload and time constraints are major barriers to you being able to effectively apply evidence based medicine • Healthcare professionals must accept and feel comfortable about not knowing everything. What are some health risks of evidence • There is a need to keep up-to-date based medicine? with new information without being overwhelmed by the time this takes to do so. There is so much information that no one would be able to keep up to date with it all, but we need it to be able to make reliable decisions, so we have to effectively decide what the trusted sources of info are 2 / 16 • There is a need to develop effective ways to identify, select, evaluate and apply information at point of care. EBM isn't finding the original research and evaluating the trial papers, it is using reputable sources that summarise the relevant and valid evidence and making it accessible at the point of patient care How do we practice EBM? It's a cyclical process: 1. Formulate a clinical question to ask that is answerable 2. Find the evidence 3. Appraise the evidence - the evidence we find may not be able to be used for the patient so we need to critically appraise/assess/judge it to see how relevant it is, and how valid, accurate and precise the results are 4. Apply evidence to clinical practice - When we are satisfied that we have found the right evidence, we need to apply the evidence to clinical practice and the individual receiving treatment. 5. Evaluate steps 1-4 and decide whether it answered the clinical question, redefine the question if necessary and start the process again (Question, find, appraise, apply, evaluate/redefine) • Must be a well-structured question to make the process of searching for evidence efficient • There are two types of questions: backHow do we formulate a clinical question ground and foreground questions. As ex(step 1)? perience with a certain subject area increases, we will ask less and less background questions and more foreground questions, over time. 3 / 16 What are background questions? • They are questions that aim to find general knowledge of a condition or disease, pathophysiology, investigations, diagnosis and/or treatment • There are two components to background questions: a question root (who, what, when, where, why, how) followed by a condition or disease, investigation or treatment • Examples include - How does diabetes mellitus increase the risk of cardiovascular disease? - What group of microorganisms normally cause urinary tract infections? What are foreground questions? • They are questions that aim to gain specific knowledge to inform individual clinical decisions or actions using the evidence available • There are four components to foreground questions: - the Patient, Population or Problem - the Intervention being considered - the Comparison intervention (if relevant) - the clinical Outcomes of interest We can use the abbreviation PICO to remember (patient/population/problem, intervention, comparison, outcome) • Examples include - in adults aged <65 years with condition A, would adding drug B, compared to increasing dose of current therapy, lead to improved control of symptom C? 4 / 16 • Where you look for information depends on the type of question you are asking - resources for answering background questions are academic or medical textbooks, and narrative reviews in journals - resources for answering foreground questions are primary sources eg original research articles, and secondary sources eg syntheses or summaries of pre-appraised evidence (already been assessed for validity) • Information is available in different forms • Some general resources include How do we find the evidence (step 2)? - NICE - CKS - the Cochrane Library - Leicestershire & Rutland APC - SPS - CRD - SIGN guidelines (this is made in Scotland so treatment options may differ but research can be used since this would usually be on a national/regional level) - peer-reviewed journals for primary literature eg the British Medical Journal (BMJ), the New England Journal of Medicine, The Lancet, the Pharmaceutical Journal • Studies, which are related to a particular focused question • Synopses of single studies, which provide brief summaries of results and implications of single high-quality studies • Syntheses, which are systematic reviews that provide rigorous summary of all primary research evidence that could be found relevant to a particular focused question 5 / 16 What are the different forms of evidence? • Synopses of syntheses • Summaries, which provide an outline of management options for a given health issue, using high quality research evidence • Systems, which are electronic and can be sophisticated enough to link to patient records and prompt practitioners about guidelines for care As you go down this list, the time taken to practice evidence based medicine decreases What is NICE? • National Institute for Health and Care Excellence • They review the best available evidence to develop recommendations to guide health and social care decisions on a wide range of therapeutic topics • They serve the department of health What is CKS? • Clinical Knowledge Summaries • They provide access to summaries of the current evidence base, and offer practical guidance on best practice in primary care, in the form of case studies on how to manage particular conditions with non-specialist medications What is the Cochrane library? • A regularly updated collection of highquality evidence based research • It contains The Cochrane Database of Systematic Reviews, which uses summaries of results of all clinical trials regarding a specific question. • Area Prescribing Committee • A local source rather than national publications, so the committee looks at evidence of medicines being used on a local or regional level • They provide evidence to manage the 6 / 16 What is the Leicestershire & Rutland APC? introduction of new drugs, and advise on the prescribing of medications within the local health community including in hospitals • They rely on the evidence base themselves and present some of this as guidance notes on their website What is the SPS? • Specialist Pharmacy Service • It is a comprehensive online resource • It gathers expert advice and evidence-based information to create a source of advice for pharmacists on the use, monitoring and prescribing of medicines within the NHS What is the CRD? • Centre for Reviews and Dissemination • It provides access to databases holding systematic reviews and economic evaluations of health and social care interventions • If you have used a reputable source of information then this is done for you by experts, so there is little need to assess the validity. • You need to appraise the evidence if you are using original articles or websites of unverified merit. This is done through literature review skills and critical & statistical analysis skills How do we appraise the evidence (step • Usefulness of information = (relevance 3)? x validity)/work involved Useful info is highly relevant (applies to the question, readily applicable to frequent problems in everyday practice, and matters to patients) and highly valid (the extent to which the info is correct and factual) and takes little work to locate, evaluate and understand. • One good way of filtering medical litera7 / 16 ture is to concentrate on published studies that measure outcomes that matter to the patient themselves. • Disease-orientated outcomes (DOOs) are outcomes of an intervention that offer surrogate, intermediate or objective physiological results eg evidence looking at systolic blood pressure, plasma glucose concentration, bone mineral density etc. This evidence isn't useful for a patient to see how useful a drug might be • Patient-orientated outcomes (POOs) / Patient-Orientated Evidence that Matters (POEMs) are outcomes of an intervention that are important to patients and assist in clinical decision-making eg eviWhat are the two types of outcomes that dence looking at morbidity and mortality, studies can have? quality of life, clinical events like number of hospitalisations • Published studies use DOOs to predict the clinical outcome of an intervention, mostly because they are easier to measure than POOs, but there are some problems with evidence that depends on surrogate outcomes: - only measures a single outcome - may not measure/correlate to patient harm or adverse events - may not accurately predict the clinical outcome 8 / 16 • A study on post-myocardial infarction (events occurring after having a heart attack) used the antiarrhytmic drug Flecainide and the DOO showed a reduction in ECG abnormalities, however the actual clinical outcome was that there was an increased mortality rate with taking the drug. The DOO showed a positive result however the POO was a negative thing. • A study on hypertension used the antihypertensive drug Doxazosin and the DOO showed a reduction in blood presWhat are examples of clinical trials sure of patients, however the POO was where consideration of DOOs can pro- that the drug ended up doubling the rate vide different information to POOs? of heart failure in these patients, increasing the rate of cardiovascular disease events • A study on osteoporosis used sodium fluoride to show an increase in bone mineral density, however it also showed an increased rate in fractures • A study on prevention of cardiovascular disease events used HRT (oestrogens) to show a reduction in LDL cholesterol levels however there was no reduction in mortality rate (when you would expect there to be) 9 / 16 • There is a danger of over relying on research, recommendations, or advice from people we deem to be experts, rather than taking the time to properly find evidence base ourselves, hence we need to beware of the quality of reviews or articles written by 'experts' as this can directly influence how we care for our patient. • When a study has been designed to get a positive outcome, we will often obtain POOs which are valid but often rejected, especially when they don't "make sense" What is the importance of 'experts' when or conflict with DOOs. This can be known it comes to appraisal of evidence? as the tomato effect. • Experts often cannot leave their past experience and inherent, preconceived views behind, because of the expertise they have developed in their area, and so fail to keep up to date with latest evidence that proves something different to what they are used to • Research indicates the quality of a review article varies inversely with the expertise of the author (the greater the expertise of the author, the poorer the quality of the article) 4000 patients with type 2 diabetes were assigned to receive either conventional treatment or more intensive treatment, and the study lasted over 11 years. The study aimed to demonstrate that tight glycemic control had positive outcomes (DOO). POOs were also found. 35 experts then write articles reviewing the findings from this study. However some discrepancies were revealed on the reporting and results of this study. These expert had a preconceived view 10 / 16 that tight glycemic control is the way forward to manage type 2 diabetes. Therefore more of the experts included POOs that supported their view in their review article eg the finding that tight conform What UKPDS study shows the influence decreased likelihood of 21 different comof expert advice? plications, and less experts reported the POOs that opposed their view eg tight control did not prevent premature mortality. This shows the influence that experts can have. • Medical research is published • Results are summarised and interpreted by those interested in the field • Those who summarise the results publish review articles, help to develop consensus guidelines at a national or local level, and continue education of medical How does publishing of research affect healthcare professionals who then use the patients themselves? this information for their own practice. • The information now becomes available to healthcare professionals to keep up to date and use in their practice • This information dictates the healthcare professionals's choices of treatments to offer to patients, influencing patient care • Medical research is published • Results are selectively summarised and interpreted by 'experts' based on their own experience • These experts publish review articles, help to develop consensus guidelines at a national or local level, and continue education of medical healthcare professionals. If the work published by experts is selective, then the review articles are then flawed and the integrity of medical 11 / 16 education is compromised since the information doesn't include the correct interpretation of the results. • The information now becomes available to healthcare professionals to keep up to date and use in their practice How does publishing of research by ex- • This information dictates the healthcare perts negatively affect the patients them- professionals's choices of treatments to selves? offer to patients, influencing patient care as it is now based on incomplete or inaccurate information This shows the influence of experts on patients and how they can directly affect how healthcare professionals choose to deliver patient care. • Once medical research is published, results should be summarised by independent 'information masters' rather than experts in the field, to reduce bias of reporting How can we avoid publishing of research • This would lead to publication of negatively affecting patients? high-quality, systematic reviews, evidence-based guideline development and continuing medical education forums • Ultimately, patient care is of the highest quality • Consider - author(s) - date of publication - edition or revision - publisher How do we evaluate the merit and judge - title of journal or article how trustworthy a source of information - intended audience is online? - writing style - objective reasoning - breadth of coverage - reference to evidence base • Tools can be used to aid the process of 12 / 16 evaluating a website eg the ABCD tool: Authority, Bias, Content, Date • To assess how useful a website is, look at the URL domains and these can help us decide how to interpret the information What comes under Authority in the ABCD tool? • Who is the author, publisher, organisation or sponsor? Is there an 'About Us' section on the company website? • What are the authors credentials or level of expertise? • Is the author or organisation qualified to write on the subject? • Can the author or organisation be contacted, and if they have their contact details readily available eg email, telephone number, post address, in case we have a query with anything they have presented? What comes under Bias in the ABCD tool? • Why does the website exist and what is the primary purpose of the information on the website? Was the website created to inform, sell, entertain, persuade etc? • Does the author make their intentions clear? Or is the information quite vague? • Does the authors point of view appear objective and impartial? Does it appear as if it is based on opinion or facts? • Is a particular point of view only being presented? Is there a balanced argument for the information presented? • Was there a title to the article, and was this worded in a certain way to direct the reader to one particular view? 13 / 16 • Is the style of presentation and writing appropriate for academic or professional research? • Is there sufficient breadth and depth of information provided on the subject What comes under Content in the ABCD area? Does it seem superficial or broad? tool? • Are there lots of grammatical or spelling errors that could raise doubts on the credentials of the author or website? • Is the information appropriately linked to reference sources? Is it all backed by evidence? What comes under Date in the ABCD tool? • When was the website produced? • How frequently is the information updated? When was this last done? This is important if the research is active and ongoing in the specific area. • Is it important for the information to be recently written? Would the information still be applicable and relevant now, if it was written many years ago, or is it outdated now? • Is the information consistent with your knowledge on the topic, or does it have any surprising or unexpected details? • How well is the website looked after and maintained? Do all links to additional sources of information still work? If the URL domain ends in • .com / .net These are commercial websites with the main objective to gain financial income from the information provided, through sponsorship or directly from the public • .org These are websites from an organisation, and these could be non-profit or profit-making 14 / 16 • .gov.uk These are websites from government institutions, and they often provide up-to-date facts, statistics and information on current issues How do URL domains influence the use• .ac.uk / .edu fulness of a website? These are websites from educational institutions, and they provide scholarly works and research from academic departments, which are usually useful and reliable to use How do we apply evidence to clinical practice (step 4)? • We are satisfied that the evidence is of high-quality, valid and relevant. We now need to use professional judgement and other skills to decide whether the evidence is applicable to the patient(s) with a particular disease. • You need to consider certain factors before applying the evidence to patient care - quality of the evidence available - characteristics of the patient - characteristics of the disease (we need to make sure these characteristics we see in the research and studies are the same as the ones we would see in real life in clinical practice. any discrepancies may then mean the evidence is not applicable) - baseline risk of the patient - risks vs benefits to the patient - preferences and values of the patient (some studies may not take these patient factors into consideration, and these would influence their decision to take the medication or not) - statistical vs clinical significance (most studies show results from a mathematical point of view) 15 / 16 We need to be sure that the process of EBM has been successful: the question has been answered and it has been applied to the patient. We can ask specific questions to ourselves to see if this is the case: • Are my questions specific, relevant and focused? • Have I asked the right number of background and foreground questions for my level of experience? • If I do not have an answer, am I able to formulate more refined questions? • If I do not have an answer from the How do we evaluate steps 1-4 (step 5)? information I have used so far, am I able to identify further sources of information that are of high quality? • Do I know the most appropriate sources of evidence to find answers to medical/ therapeutic questions? • Can I find and interpret the evidence quickly? • Can I judge the relevance and authority of any information I find, to make sure that the information I then provide to the patient is correct and relevant? • Can I apply the information I find to support individual patient care (ultimate goal of EBM)? 16 / 16