Tip for data extraction in meta-analysis – 19

October 28, 2019

What if there’s no intervention?

Kathy Taylor

In my last post (video) I introduced the PICO as the abbreviation for the key elements of a clinical question. In this post I’ll highlight other abbreviations that may be used, and also cover the varied use of the PICO abbreviation.

Describing a clinical question

The key elements of a clinical question may be defined by the PICO:

P Population, Patient or Problem Simple illustration: a group of people, each a different bright colour.
I Intervention Illustration: a pill bottle surrounded by yellow and red pills.
C Comparator or Control  

Pill bottle surrounded by red and yellow pills, the word 'or', then a pill bottle surrounded by green and red pills

 

O

 

Outcome(s)

 

Illustration: a figure looking up at a big red question mark

 

Regarding clinical questions in intervention studies, including randomised control trials (RCTs), the intervention will often be a drug or combination of drugs, but could also be another intervention such as a surgical procedure. For the drug-based interventions, the comparator might be a placebo drug (which looks the same as the intervention drug and will be in the same packaging) or a different drug.

The example I gave in my video was a drug intervention study. It was a trial of glucose lowering drugs and had the following PICO:

P – Patients with type 2 diabetes
I – Metformin combined with insulin
C – Placebo combined with insulin
O – Changes in the carotid intima media thickness (primary outcome)

Rather than an intervention, the study might involve an exposure, as in a cohort study. For this, the key elements of the clinical question may be abbreviated by PECO:

P Population, Patient or Problem Simple illustration: a group of people, each one a different bright colour.
E Exposure Simple illustration: a group of people, each one a different bright colour; surrounded by grey and white stripes creating the illusion of perspective.
C Comparator or Control Simple illustration: a group of people, each one a different bright colour.
O Outcome(s) Illustration: a figure looking up at a big red question mark.

 

For example, a retrospective cohort study using data from a database of general practice records considered the relative risks associated with type 2 diabetes of all-cause mortality and cardiovascular mortality in middle aged people. The PECO is

P – Middle-aged patients
E – Type II diabetes
C – Without diabetes
O – All cause mortality, cardiovascular mortality

For diagnostic accuracy studies, we can use another abbreviation, PIRT:

P Population, Patient or Problem Simple illustration: a group of people, each one a different bright colour.
I Index Test Close up photograph of a doctor pricking someone's fingertip
R Reference Test A medical scan
T Target Condition Illustration: a red target with an arrow in the bullseye.Illustration: a figure looking up at a big red question mark.

 

An example of this is a study of point-of-care testing for heart failure:

P – Primary care patients
I – Point-of-care testing of N terminal fragment pro B-type natriuretic peptide (NT-ProBNP)
R – Clinical examination
T – Heart failure

The PICO, PECO or PIRT for a systematic review will be broader as it will cover multiple studies. For example, the following PIRT refers to the diagnostic accuracy systematic review that included the study described above:

P – Patients in ambulatory care settings
I – Point-of-care testing of B-type natriuretic peptide (BNP) or N terminal fragment pro B-type natriuretic peptide (NT-ProBNP)
R – Echocardiography, clinical examination or a combination of both
T – Heart failure

As I said in the video, the key elements of a clinical question will typically be found in the title and abstract of the article. Note that study questions won’t always contain all the elements of a PICO, PECO or PIRT. For example, you may only find a population, exposure and outcome. This is illustrated by a study of the association between sugar-sweetened drinks and colorectal cancer:

P – Female teachers and administrators
E – Consumption of caloric soft drinks, sweetened bottled waters and teas, and fruit drinks
O – Colorectal cancer

Varied uses and extensions of PICO

The PICO and PECO are often combined and referred to as PICO, where I for intervention is broadly defined to encompass a treatment, exposure or prognostic factor.

PICOTT is an extension to PICO, which combines all three abbreviations, adding
T – Type of question (therapy, diagnosis, harm, prognosis or prevention)
T – Type of study (systematic review, cohort study, RCT or case-control)

The PIRT example that I gave above would have the following PICOTT:

P – Patients in ambulatory care settings
I – Point-of-care testing of B-type natriuretic peptide (BNP) or N terminal fragment pro B-type natriuretic peptide (NT-ProBNP)
C – Echocardiography, clinical examination or a combination of both
O – Heart failure
T – Diagnosis
T – Systematic review

Alternatively, for systematic reviews, there’s the PICOTS abbreviation:

P – Population/Problem
I – Intervention (broadly defined)
C – Comparison
O – Outcome
T – Timeframe
S – Setting

In the context of systematic reviews of the prediction model performance, the PICOTS is

P – Population in which the prediction model will be used
I – Prediction model
C – Competing models
O – Outcome for which the model is validated
T – Timeframe, for prognostic models
S – Setting

Others remain using PICO but alter its elements according to the type of question.

All the above abbreviations are intended to help define a clinical question. A well-defined question will provide a clear focus for the literature search and will be more likely to provide a useful answer and ensure that research resources are well spent.

Illustration: a group of figures. The figures in the background are shades of grey, there is one red figure at the front holding its arms above its head.

Starting with a poorly defined question will more likely produce confusing and unhelpful answers.

Here’s a tip…

Abbreviations such as PICO, PECO and PIRT are intended to help clearly define the clinical question for a study or a systematic review.

Dr Kathy Taylor teaches data extraction in Meta-analysis. This is a short course that is also available as part of our MSc in Evidence-Based Health Care, MSc in EBHC Medical Statistics, and MSc in EBHC Systematic Reviews.

Follow me on Twitter @dataextips for updates on my blog, related news, and to find out about other examples of statistics being made more broadly accessible.

Leave a Reply

Your email address will not be published. Required fields are marked *

* Checkbox GDPR is required

*

I agree