June 27, 2022

Evidence-based medicine – PICO and How to develop answerable clinical questions?

Put yourself in this scenario: You are a physician, coming to the end of a long week of surgical procedures, Friday evening, one last patient who is a 35-year-old man, with a serious type-III, open tibial fracture. The patient understandably wants to save his leg. You want to discuss the case with a couple of trusted, more experienced colleagues, but they have left for the weekend, so it is down to you and science.  You decide on a flap reconstruction of the limb because you are led by emotion and you cannot handle the added stress and strain of telling him that amputation is probably the best option, based on scientific evidence.

The problem is you have a gap in knowledge, you have looked at a couple of expert opinion articles, but time is short, you do not know how to ask the right question to find the right answer, and anywhere you ‘take a punt’ to look does not have your precise problem on record.

This is where PICO (Patient/Problem, Intervention, Comparison Intervention, and Outcomes) comes into prominence.

PICO is a tool to aggregate the most relevant, focused data from Evidence based Medicine (EBM), allowing the physician to find the best outcomes to his specific clinical question. In this scenario, utilizing PICO, you are able to reduce your patient population to Type IIIB/Type IIIC fractures (based on the lack of soft-tissue coverage of the fracture requiring flap reconstruction). You will now have data concerning limb salvage with reconstructive surgery, amputation, with several example cases. You simply have to input the question: What procedure results in a better quality of life and limb outcomes in Type IIIB/C fractures?

So how does PICO work?

Simply, there are 2 types of clinical questions: Background and Foreground.

In Background questions, you ask for general information about a condition, test, or treatment. Within this, you have 2 components:

  1. Question root and verb. For example: What causes cellulitis?
  2. Disorder, test, and treatment. For example: How do I treat cellulitis?

Background questions can be answered using traditional textbooks.

In Foreground questions, you have specific knowledge about managing specific patients with a specific disorder. Within this, you incorporate PICO. For example: In a diabetic patient with cellulitis (P), do IV antibiotics (I) compared to oral antibiotics (C) lead to higher cure rates (O)?

You can see that this is a much more detailed, higher-order question than those that represent background questions.

So how do well-formulated questions help us?

First, it is important we focus our scarce learning time on evidence that is directly relevant to our patient’s clinical needs. Second, it is important to focus scarce learning time on evidence that directly addresses our particular knowledge needs (or those of our learners). Third, they can suggest high-yield search strategies (and with that you can see if that patient population is fully covered in the data you look at?). Fourth, they can enhance hand-offs or referrals. And finally, as a tool for teaching: learners can better understand content and model adaptive processes for lifelong learning.

So which study/methodology types offer the best answers to different sets of questions/domains?

  1. In Therapy/Treatment type questions (as in the selection of treatments or interventions that do more good than harm and that are worth the effort and cost), double-blind, randomized controlled trials or systematic review/meta-analyses of random controlled trials offer the best data.
  2. In Diagnosis (as in the selection and interpretation of diagnostic tests, in order to confirm or exclude a diagnosis, based on considering their precision, accuracy, acceptability, expense, safety, etc.,), a controlled trial or systematic review/meta-analyses of controlled trials offer the best data.
  3. In Prognosis (as in the estimation of a patient’s likely clinical course over time and anticipation of likely complications of disease), cohort studies, case controls, and case series offer the best data.
  4. In Harm/Etiology (as in the identification of causes or risk factors for diseases), cohort studies offer the best data.
  5. In Prevention, random controlled trials and cohort studies are best, and
  6. In Quality Improvement, random controlled trials offer the best data to determine management.

So how do you find the best elements for the perfect search?

Over the last 10 years, a new set of standards for reporting literature searches were developed under the mnemonic STARLITE (see image below). These include standards in the reporting of Random Controlled Trials (CONSORT – Consolidated Standards of Reporting Trials), meta-analysis of RCTs (QUOROM – Quality of Reports of Meta-analysis19), and meta-analysis of observational studies (MOOSE – Meta-analysis of Observational Studies in Epidemiology).

If researching a paper, it is therefore important to report the elements of the literature search so that your results can be reproduced.

Now back to PICO, let us try to convert the following scenario to a PICO question:

On morning rounds in the hematology/oncology unit, a first-year resident turns to you for consultation. She wants to discuss options for managing moderate nausea and vomiting that result following chemotherapy. She shares an experience of one of her relatives who was taking ginger, when prochlorperazine did not provide effective relief, and asks for your input.

In this case, chemotherapy is the Problem, ginger is the Intervention, prochlorperazine is the Comparison Intervention, and moderate nausea and vomiting are the Outcome.

Utilizing this method, you can determine the following PICO question: In patients undergoing chemotherapy, is ginger as good as or better than prochlorperazine in reducing moderate nausea and vomiting?

This is much better than asking “Does ginger work? Work in who? For what? Better than what? Everything is relative to something else, so forming detailed PICO question will help you focus on getting a good answer to your question and will help you find the resources that you need in multiple study designs, to help you focus your scarce time and effort in providing high-quality care and improving your own knowledge base.

Leave a Reply

Your email address will not be published.

%d bloggers like this: