Written By Guest Blogger, Robert Marshall, MD, MPH, MISM, FAAFP, CMO, HIT GROUP

The internet has become an excellent resource for information on virtually any medical condition. Type in a search on a particular topic, and hundreds if not thousands of results will emerge, all offering information on the clinical condition you wish to know more about. While reading through such information (at least the more reliable information), you may find the terms such as ‘systematic review’, ‘meta-analysis’, ‘prospective studies’, ‘long-term outcomes’ and ‘mortality and morbidity’ being mentioned. These are particularly common in papers that are more scientific in nature and written in a healthcare professional rather than lay style.

Those terminologies are utilized in research papers which form a part of what is known as ‘evidence-based medicine’. The term and concept were introduced into medical literature by Guyatt in 1992. Evidence-based medicine refers to the practice of medicine based on high quality research studies assessing whether or not treatment options or diagnostic methods actually have benefit in a wider population group rather than in an individual or based on anecdotal/expert opinion. The ‘evidence’ refers to the research studies supporting specific diagnostic methods or therapies. However, evidence-based medicine is a lot more complex than that. In this article, we shall take a look at what evidence-based medicine is, discuss a few terminologies that may help you understand the different aspects of research papers and show/discuss how medicine can be practised when backed by reliable, repeatable science.

Defining evidence-based medicine

The most common definition of Evidence-Based Practice (EBP) is from Dr. David Sackett. EBP is “the conscientious, explicit and judicious use of current best evidence in making decisions about

the care of the individual patient. It means integrating individual clinical expertise with the best available external clinical evidence from systematic research.”

EBP is the integration of clinical expertise, patient values, and the best research evidence into the decision making process for patient care. Clinical expertise refers to the clinician’s cumulative experience, education and clinical skills. The patient brings to the encounter his or her own personal preferences and unique concerns, expectations, and values. The best research evidence is usually found in clinically relevant research that has been conducted using sound methodology.

The evidence, by itself, does not make the decision, but it can help support the patient care process. The full integration of these three components into clinical decisions enhances the opportunity for optimal clinical outcomes and quality of life. The practice of EBP is usually triggered by patient encounters which generate questions about the effects of therapy, the utility of diagnostic tests, the prognosis of diseases, and/or the etiology of disorders.

Put in simple terms, evidence-based medicine looks at a number of different aspects of a clinical condition or a treatment. It looks at the benefits and harm that a diagnostic test can do, assesses these risks in population groups carefully selected to represent the average patient and helps the clinician to use this information when treating a patient. Evidence-based medicine will inform the treating physician about whether or not a treatment that they wish to offer the patient is more likely to cause them harm or be beneficial.

The process of evidence-based medicine

In a study published by Martin Dawes, the process of evidence-based medicine was broadly classified into five steps:

  1. Translation of an uncertain aspect of medicine into an answerable question (and includes critical questioning, study design and levels of evidence [were the results due to high or low quality study design])
  2. Systematically retrieving the available evidence by looking at different research studies
  3. Critical appraisal of evidence for validity that can be broken down into aspects regarding:
  4. Systematic errors as a result of bias and confounding (not accounting for external or internal factors that might invalidate the results)
  5. Quantitative aspects of diagnosis and treatment (the numbers)
  6. How large was the effect size and was it precise
  7. Clinical importance of results
  8. External validity or generalizability (can it be applied to patients other than those actually in the study)
  9. Application of the available results in clinical practice
  10. Evaluation of performance of these methods (was the study or review conducted well)

While the above steps may sound a little complicated, it basically means that evidence-based medicine starts at the utter basics with a question regarding a diagnostic test, particular treatment or clinical condition, follows a stepwise approach looking at the different studies that have been conducted in the area, critically analyses the available data from the studies and puts together a final review on the application of the data in clinical practice.

How a research study is conducted

While a detailed description on how exactly a research study is conducted is out of the scope of this article, it is worthwhile for you to understand what steps are undergone when a research study is conducted.

Primarily, clinical studies can be either prospective or retrospective. A prospective study is one that is currently ongoing and assesses data that is collected over a period of time into the future. On the other hand, a retrospective study is conducted based on information that is already available from a previous period of time. In the scientific world, prospective studies are considered of higher quality than retrospective studies, although valuable information can still be obtained from retrospective studies.

Another aspect of clinical studies is called ‘blinding’. Blinding refers to the availability of or absence of information to the researcher regarding the study. For example, in a blinded treatment study, one or more of the researchers is/are unaware of who is assigned what treatment. This makes the chances of this person intentionally or unintentionally biasing the results highly unlikely, making the measured/stated outcomes much more reliable. Most studies conducted these days are double blinded studies, meaning that both the researcher and the patient are unaware of the treatment that they are receiving. This information is only available after the study is concluded and the results have been analysed.

Another term and you may come across when reading about studies is the word ‘placebo’. Placebo-controlled trials refer to clinical trials that involve one group of patients receiving a medical treatment while the other group receives one that should have no effect on the disease being studied. The latter is called a placebo. It is a useful way to compare the effect of a medicine on clinical conditions as compared to no active treatment.

‘Outcomes’ of studies is yet another terminology commonly encountered in clinical studies. An outcome refers to a particular aspect/result of the treatment that is being studied. For example, if a clinical study is looking at whether or not a new drug can help reduce the number of people dying from heart attacks, the outcome of the study would be to see how many people actually died from a heart attack despite the medicine and how many were saved due to the medicine. The outcomes can be either short-term or long-term; short-term outcomes are those seen over a few weeks to months, while long-term outcomes are those seen over a number of years.

Once a decision has been made to conduct a clinical study, a research plan is put forth in front of an institutional review board (IRB). The IRB consists of both experts and patients who will assess the research study to ensure that all the aspects of this study are both scientifically reasonable and within the ethical boundaries defined by the institution and the Federal government. The dictum ‘do no harm’ is essential to be followed and every aspect of this will be covered during the ethics review and the IRB meeting.

Once the study has be approved and subsequently commenced, different pieces of data are collected to assess various aspects of the diagnostic test or treatment that is being studied. Care is taken to ensure that no harm is done to patients during the study. In the event that initial results demonstrate that a treatment or a test is doing more harm than good, the study is immediately stopped and reviewed. If no harm has occurred/occurs, the study is continued till its conclusion or at a predesignated interim evaluation point (for long term studies). If harm is proven upon review, the study is terminated. When the study has concluded, the data is put together and analysed using pre-approved statistical analysis.  The results are then presented or published in the form of a research paper.

Types of evidence-based medicine

Broadly classified, there are a number of different types of evidence-based medicine. Below is a brief description on each:

  1. Systematic review

The term ‘systematic review’ refers to a type of scientific study that is conducted in a strategic manner, assessing the available research studies that have been conducted in a particular field. For example, systematic reviews may be conducted looking at the benefit of particular tablets in managing patients with heart failure. Systematic reviews may also be conducted assessing the risks or benefits of a particular treatment for managing fractures. In essence, a systematic review takes information from a number of different studies, large and small, and creates a more powerful set of conclusions for the physician. It often does so in an easy to understand format.

But why exactly are systematic reviews performed? Well, given the amount of scientific research that is being conducted these days, it is not uncommon for studies to show varied results depending on how they were performed, how large they were and/or what their ultimate goal was. Systematic reviews help pick out particular aspects of the studies, assess the quality/validity of the studies (and throw out the bad ones), collate the higher quality information and put it together in a format that makes the results a lot more clear and easier to apply to patients.

  1. Meta-analysis

A meta-analysis is similar to a systematic review, and takes into account results from different scientific studies, put them together and brings to light either new evidence or points towards a conflict in those results. It looks at a large number of similar studies and put the results together in a single paper, highlighting the important aspects of the study.

  1. Randomized controlled trials

A randomized controlled trial refers to a clinical study that is conducted in a random population. This means that the individuals in the study are selected at random for each arm of the study, although they may have to meet certain criteria to be included. For example, if one were looking at the effect of a new medicine in treating stroke, the included group might not contain individuals who are already taking anticoagulants like Coumadin®, Xarelto® or Plavix®. A ‘controlled’ trial implies that the study is conducted in a strict and controlled manner, with constant observation of the patients in the study.

Often randomised controlled trials are conducted in a double-blind manner as previously discussed.

How is the quality of the evidence assessed?

In order to address this issue, there are specific groups of experts who analyse research studies and decide at what ‘level’ the evidence is for a particular study. In the United States, the US Preventive Services Task Force (USPSTF) classifies evidence into five different categories. Level I refers to the evidence that is obtained from a well designed randomised controlled trial, while Level III refers to evidence available through clinical experience and not backed by scientific research. Similarly, in the United Kingdom, the National Health Service also has their own level of evidence assessment criteria. Level A includes evidence obtained from randomized controlled trials, while level D refers to  that obtained through study of different cases and anecdotal evidence.

In clinical practice, healthcare professionals will ideally use information obtained from the highest level of evidence-based studies to drive diagnosis or therapy decisions.

In some cases, guidelines may be put forth by expert groups to manage different clinical conditions. The American Heart Association has clear-cut guidelines on the management of heart attacks. The recommendations that are made in these guidelines are categorised into different levels as well. Level A refers to strong scientific evidence of more benefits than risks. Subsequent levels involve lesser availability of scientific evidence and an increasing risk to benefit ratio.

The bottom line is that there are different levels of evidence, and the highest level of evidence should be utilized when providers decide what diagnostic test to order or what medical treatment to offer the patient.

Evidence-based medicine terminologies

There are other terminologies within evidence-based medicine. Here are a few of them that you may come across when reading a research paper review.

  1. Number needed to treat (NNT) and number needed to harm (NNH)

A number needed to treat (or harm) is a measure of how safe or harmful a treatment is. It refers to the number of patients that need to be treated in order to prevent one bad outcome (such as heart attacks, stroke or death). For example, an NNT of 10 means that 10 patients need to be treated to prevent one additional bad outcome. Similarly, NNH looks at the number of people that may be harmed from the treatment. The higher the NNH, the lower the risk.

  1. Endpoints

Endpoints are long-term outcomes in clinical studies. If looking at a new drug that treats stroke, an endpoint would be the development of stroke or the prevention of it. Statistical calculations are conducted to see whether or not the endpoints are significant. One caveat: statistical significance does not always infer clinical significance. It is important for both providers and patients to be aware of/ask about clinical significance of an endpoint/outcome.

  1. Mortality and morbidity

These are two commonly used terminologies in evidence-based practice. Mortality refers to the deaths due to a treatment or clinical condition. On the other, morbidity refers to a state where the individual is suffering from a disease or has an untoward side effect from a therapy.


Evidence-based medicine is essential to ensure that patients are receiving the best possible treatment. These days, it is the norm in clinical practice. There are different steps involved in evidence-based medicine, and a good knowledge of this is important for both the patient and the healthcare practitioner.


David L Sackett, William M C Rosenberg, J A Muir Gray, R Brian Haynes, W Scott Richardson. “Evidence based medicine: what it is and what it isn’t.” British Medical Journal (1996). 312: 71.

Guyatt, Gordon, John Cairns, David Churchill, Deborah Cook, Brian Haynes, Jack Hirsh, Jan Irvine. “Evidence-based medicine.” Journal of the American medical Association (1992). 268, no. 17: 2420-2425.

Guyatt, G. Rennie, D. Meade, MO, Cook, DJ.  Users’ Guide to Medical Literature: A Manual for Evidence-Based Clinical Practice, 2nd Edition 2008

Greenhalgh T. Papers that summarise other papers (systematic reviews and meta-analyses). BMJ (1997 Sep 13). 315(7109):672-5. PubMed PMID: 9310574.

Ried K. Interpreting and understanding meta-analysis graphs: a practical guide. Aust Fam Physician (2006 Aug). 35(8):635-8. PubMed PMID: 16894442

Dawes M, Summerskill W, Glasziou P, et al. “Sicily statement on evidence-based practice”. BMC Med Educ (January 2005).  5 (1): 1. doi:10.1186/1472-6920-5-1. PMC 544887. PMID 15634359.