1. We aim to support informed decisions, NOT to give recommendations
The main goal of all of our work is to support informed decision making at all levels of the health care service: decisions concerning the systems (e.g. how to organise, govern, finance and implement health care), decisions about preventive services, and decisions about collective and individual clinical care.
Being informed means having access to reliable evidence, for instance about “what works” (what are the important benefits and harms of the interventions under consideration, and how reliable is that evidence?). It may also mean having access to what different options might cost, or how feasible they are to implement, or how acceptable they are to the people delivering or receiving them. Being informed also means knowing when reliable evidence is lacking.
But decisions do not necessarily flow directly from evidence alone. Value judgments are always involved, and information from other sources may also play an important part.
For instance, one person may choose to not undergo a particular surgical intervention, because she considers the risk for harm too high compared with the potential benefit the surgery represents for her. In addition, she may consider the weeks of follow up therapy as a heavy burden, due to her life situation. Another patient might make different choices based on the exact same information. For this reason, evidence presentation must remain neutral, so that decision makers can add their own value judgments to the available evidence and information.
Our goal is to support decision makers at all levels (policy makers, health professionals, and consumers) in making decisions - for themselves or for others - that are well informed by the available research evidence. This means that we strive to make the evidence understandable, transparent and accessible for decision making contexts.
2. Summary of Findings tables are a critical starting point for our products
When producing summaries for different audiences from Cochrane Reviews, we have often run into the following problems:
Reviews report too many outcomes, and it’s not apparent which are most important for the decision maker
The quality (or certainty) of the evidence is not always explicitly presented
Effects are presented in many different statistical formats (making them sometimes difficult to understand and often difficult to compare)
Sometimes the way results are reported in the Abstract are inconsistent with the presentation in the Plain Language Summary or Authors’ Conclusion.
For these reasons, we always start with a GRADE Summary of Findings table of the evidence when we are creating any other kind of summary. The Summary of Findings table presents, in a very condensed format, the effect of the intervention for the most important outcomes, including benefits and harms, as well as an evaluation of the certainty of this evidence. It also indicates where evidence for important outcomes is lacking. This provides us with a precise starting point for creating any kind of summary of the evidence – paper, web-based, video, quantitative, qualitative or visual.
3. We use a layered presentation approach: the “1:3:25 principle”
The 1:3:25 principle is a concept for structuring information in layers. The “1” stands for one-page of key messages, “3” for three pages of executive summary, and “25” means the entire unabridged document. One of the most robust findings from our studies – across all target groups and products - is that people appreciate this layered structure of evidence and find it very useful. The concept comes from the Canadian Health Services Research Foundation, and can be read about in more depth here.
4. When it comes to understanding research, different target groups may have similar problems
Evidence regarding health care interventions is often labeled as “For consumers” or “For health professionals”. Likewise, many of our projects are organized around developing approaches for a specific target group, such as policy makers or health professionals, or lay people. However, taking a step back and looking at the body of work as a whole, we have found that user responses don’t arrange themselves clearly into these kinds of professional/non-professional categories. In fact, when it comes to understanding scientific evidence, there seems to be much more similarity than differences across users in all these groups.
As a consequence of this, when developing an approach for presenting evidence to non-scientists, we aim to find one basic strategy that can work for educated people across all groups.
5. No product without feedback and user testing!
Most of our work involves designing things: templates, web sites, explanations and instructions. We employ a user-centered approach, which means that we try to create solutions that cater explicitly to the needs of the user. In order to achieve this, we involve both end users and other stakeholders in our development processes, and organize our work in cycles of feedback gathering and making revisions.
Potential users of a product can be drawn into all development phases of design and testing, from listening closely to people’s needs and ideas when setting goals in the early stages, drawing on their ideas and suggestions during prototype development, and collecting feedback on sketches or prototypes to help identify and fix problem issues. One method we utilize in all projects for gathering feedback is user testing. We've put together a package of instructions and user test materials.