Skip to main content

Evidence-Based Medicine: The Process in Detail

A basic introduction to the history and steps of evidence-based medicine, with a focus on the EBM resources held by the Whittemore Library.
What is Evidence-Based Medicine?

Evidence-Based Medicine (EBM) is the integration of  clinical experience and patient values with the best available research information.

This methodology is now widely used in the healthcare field.

Variations of this technique are also in use in other fields, such as education, human services, and business, there most commonly known as 'Evidence-Based Practice' or 'Best Practices'.

Here at the Henry Whittemore Library, we can assist you with the steps of the EBM process in which you develop a foreground question, then search for research literature that will help you answer clinical questions.

ebm venn diagram

The 5 Steps in Detail & The Levels of Evidence

To form a focused clinical question, also known as a foreground question, health care professionals use a PICO or PICOT form. These are sometimes embedded in the clinical information system / electronic medical records system in your workplace.

The PICO(T) form contains the following sections:

 

P - This stands for Patient or Problem. When P = Patient, it may include the chief complaint, any co-morbidities, sex and/or gender, age - anything relevant.

I - This stands for Intervention.  This will be the first possible intervention that is being considered.  Sometimes there will be only one, as you may be considering one possible treatment versus none.

C - This stands for Comparison Intervention, when more than one is being considered.  A PICO form may have an Intervention and as many additional Comparison Interventions as necessary.

O - This stands for preferred Outcome - the outcome that the patient and the clinician both want to achieve. Sometimes when the preferred outcome for a health complaint is well understood (e.g. a disease with high morbidity, such as pain), it is not always necessary to fill out this section.

T - Sometimes a Time element (time of outcome or time of observation) is added.

 

What the PICO(T) form does is force you to be very specific about your clinical question - which may be focused on possible treatments, but which can also be used to create questions about diagnosis, prognosis, etc.

 

Here's an example of a filled-out PICO form that is focusing on trreatment/management:

 

P = COPD

I =  mist inhaler

C = dry powder inhaler

O = management

 

In short, the PICO form helps you find your topic's keywords - which will allow you to efficiently search health science databases to find the research articles that will help you answer your clinical question.

 

pointing handPubMed/Medline offers a PICO search form at: go.usa.gov/xFn  (great for hand-held devices)!

PubMed's PICO Search

Let's take the PICO form and use the keywords to create a search query to use in Medline:

This PICO...

P = COPD

I =  mist inhaler

C = dry powder inhaler

O = management

...becomes:

COPD AND (mist AND inhaler) AND (dry AND powder AND inhaler) AND management

 

Let's see how many results we get:

Screenshot of a Medline search results list - 6 results.

 

As you can see, that was a very focused set of results - only 6!  In EBM, one goal is to achieve precise searches, with fewer but vastly more focused results. This saves time for busy clinicians.

Not every clinician is an expert in study design.  Fortunately, you don't have to be - the EBM process was designed to allow people who aren't to still be able to fairly quickly appraise the research articles they have found - to get a feel for whether the research being reported on has been done well, and whether the conclusions the researchers have come to are reasonable.  This will help you come to a decision about which evidence to use.

This is done by using Critical Appraisal Checklists. These are checklists that ask you a number of questions about the research study:

  • Some  about the internal validity of the research  (Was the study well-designed? Was it carried out properly?)
  • Some about the external validity of the research (Was the patient population from that study similar enough to your patient for the research to be relevant?).

  You generally answer either Yes, No, or Unclear, based on what the researchers actually tell you in the article. As a general rule, the more Yeses, the better. If you're coming up with a lot of Nos or Unclears...the study might not be that helpful to use to answer your clinical question.

These questions do include some asking about the results of the study's statistical analysis and results; again, one need not be that well-versed in statistics to answer the checklist questions (though it certainly helps).

There obviously isn't always going to be one clear answer to every clinical question - in the end, it's a judgment call, and again, any evidence that you find is just one of several elements that go into a clinical decision.  Your experience and wisdom, the patient's preferences, and what evidence you have found are all involved, as you implement the knowledge (evidence) you found into your clinical decision..

One example: Even if you found what you felt was good evidence that favored a particular intervention / treatment, your patient might still reject it due to their preferences. Rather than simply attempting to over-ride their wishes, you would need to communicate / negotiate with them further.

At other times, there is simply a dearth of 'evidence' (or a lack of good quality evidence), whereupon you must do the best you can with what there is.

The following open-access article is a good reminder of the need to integrate ethics with evidence-based practice.

There is definitely a 'learning curve' for EBM. Again, not every clinician is an expert at quickly searching databases, or knows the details of what makes good study design. The EBM protocol contains tools that help.  As with any skill, the more you do it, the better you get.

Some amount of reflection is involved in EBM - did the evidence I found help me in my decision-making and contribute to a good outcome? What was it like, carrying out the 5 steps... is it getting easier?

It is also important to acknowledge EBM's limitations. The medical community continues to ask itself how effective EBM is, including questions such as: How much does it truly help in providing care for each patient as an individual?

Many health care professionals find the time needed to carry out the EBM process (despite being deliberately-designed to make clinical research quicker and less frustrating) prohibitive. There is a reason that there are now a number of evidence-based clinical point of care tools available - these time-saving products are very attractive and, with hindsight, were inevitable.  They are, of course, expensive, but health care practices over a certain size tend to subscribe to at least one and sometimes multiple products.

Some examples of these evidence-based point-of-care products include:

The Levels of Evidence in EBM

In EBM, not all evidence is created equal. There is a definite hierarchy, as seen in the figure* below:

(Below this level would be evidence from research such as animal studies and 'bench science')

The 'gold standard' for EBM is the randomized controlled trial (RCT), and further, systematic reviews (often involving meta-analysis)  carried out on groups of RCTs.

Statistically-speaking, those types of studies offer better 'evidence', because when done correctly, they offer the best chance to avoid the various types of bias amd confounders that occur when carrying out research. - the type of evidence generated by RCTs and SRs make it easier to determine whether that research's data has sufficiently disproven the researchers'  'null-hypothesis'.

* Image used with the kind permission of Dr. Sam Keim.

Loading ...