The backbone of your systematic review or guideline are the questions that you seek to answer. The questions should reflect a central, clinical issue and these should be formulated prior to going through the literature.
The questions constitute the foundation on which studies to select (inclusion – and exclusion criteria), the focus when reading and extracting data and finally basis for formulating the recommendation.
Thus, it is essential to ask the right questions, and to include all the relevant components to make it as precise as possible. The question should specify the patient population, intervention, the alternative to the intervention and the outcome. This is abbreviated into PICO (Population, Intervention, Comparison and Outcome)
Population (patient group)
Definition of the disease and patient target group, including relevant demographic factors (e.g age, gender).
Definition of the intervention (s), including variations (e.g dosage, way of administration)
Which alternatives do you want to compare the intervention with? What would be clinically relevant to compare with? In situations where the effect of an intervention is uncertain, it would be relevant to compare with placebo/no intervention. In case the effect is well-known, the intervention may rather be compared to standard treatment or another type of intervention (head-to-head comparison).
Which outcomes are necessary in order to evaluate the effect and side-effects of the intervention?
What to measure: Outcomes may include clinical events (e.g fractures, stroke) as well as other patient relevant outcome (e.g funtionality, quality of life). You should always remember to include adverse events, her amongst others serious adverse events (e.g hospitalisation, death)
How to measure: In addition to chosing the most relevant outcome, you also should have to choose how the individual outcome should be measured. Ask yourself, what scale is used to measure this outcome? In case there are several scales (e.g measuring quality of life can be done using a variaty of scales), which one is prefered?
When to measure: You also have to consider when to measure the effect of the specific outcome. Are you interested in both short – and longterm effects? This is important, since the effect may vary across time, and thus chosing when to measure, may have a large impact on the overall assesment of the intervention.
Nice to know versus need to know: Finally, you need to specify if information on the individual outcome is critical or important in the decision making. Thus, there is something that you simply need to know (critical outcome) and then there is something that would be nice to know (important outcome) when seeking the full picture of an intervention.
With this said. you should avoid including less important outcome, since this may be overwhelmning to the reader and be disturbing when wanting to assess the overall effect of the intervention. Less important outcomes include the usage of indirect or surrogate outcome, that are not considered a precise measurement of a given clinical effect. It may however be neseccary to include surrogate outcomes, in case there are no studies that provide information on the relevant outcomes. This will however, affect the overall quality of the evidence and subsequently the final recommendation.
Need help? It may be difficult to define the most critical and important outcome within your field. Check out comet-initiative.org that includes a collection of core outcomes sets. Remember to consider outcomes relevant to both the clinician as well as the patient.