Risk of Bias
Could limitations in how the studies were designed, conducted, analyzed, or reported have distorted the results?
Autism Foundation of America
Loading pageAutism Foundation of America
Loading pageEvidence Atlas Methodology
The AFA Support Evidence Atlas is designed to show not only what research found, but how confident we can be in that finding, who was actually studied, what important limitations remain, and where the evidence does not yet provide an answer.
Our Core Principle
The same support can have stronger evidence for one outcome and weaker evidence for another. Evidence may also apply more directly to one age group or support profile than another.
Atlas ratings are therefore designed around specific outcomes and populations rather than reducing an entire therapy, strategy, or service to a single star rating.
Evidence Certainty
The Atlas uses a GRADE informed certainty framework. Certainty describes confidence in the evidence for a specific outcome. It is separate from whether the finding itself is positive, negative, mixed, or uncertain.
We have strong confidence that the research finding is close to the true effect for this specific outcome and population.
We have reasonable confidence in the finding, but important new research could change the estimate or our confidence in it.
Confidence is limited because the evidence has important weaknesses, gaps, or uncertainty.
The available research is highly uncertain and does not support a confident conclusion.
There is not enough reliable evidence to determine the effect for this outcome.
Why Certainty Changes
Could limitations in how the studies were designed, conducted, analyzed, or reported have distorted the results?
Do studies generally point in the same direction, or are results substantially different across studies without a clear explanation?
Does the research directly apply to the population, support, comparison, setting, and outcome we are trying to understand?
Are there enough participants and sufficiently precise estimates to support confidence in the finding?
Is there reason to believe that available published research may provide an incomplete or selectively positive picture?
Certainty Is Not the Result
The evidence suggests a meaningful beneficial effect for the specific outcome being assessed.
Some evidence suggests benefit, but uncertainty remains important.
Research findings differ enough that a consistent effect cannot be confidently described.
Research has not demonstrated an important benefit for the specific outcome studied.
The available evidence does not allow a reliable conclusion.
Evidence, authoritative guidance, or credible safety information identifies meaningful concerns that visitors should understand.
Who Was Actually Studied?
Applicability is reported separately from evidence certainty. The Atlas will make important population gaps visible instead of assuming results apply equally to everyone.
Population Representation
Evidence Sources
Source type alone does not guarantee quality. Every source still needs to be evaluated for relevance, methods, limitations, and applicability.
These are generally the preferred starting point when they are current, methodologically sound, and directly relevant.
Guidelines can provide structured evaluations of benefits, harms, certainty, and recommendations.
High quality individual trials may be used when systematic reviews are unavailable, outdated, or do not address an important question.
These studies can be important for long term outcomes, harms, rare outcomes, real world implementation, and populations poorly represented in trials.
Atlas Review Process
Specify the support, population, comparison where relevant, outcomes, setting, and decision the review is intended to inform.
Prioritize systematic reviews, meta analyses, high quality guidelines, controlled trials, and other evidence appropriate to the question.
Assess relevance, study design, methodological limitations, population, outcomes, follow up, and other factors that affect reliability.
The Atlas does not give one universal grade to an intervention. Certainty is assessed separately for the important outcomes being reported.
Document who was actually studied and whether the evidence applies directly to populations visitors may be asking about.
Benefits and harms are considered separately, including whether harms were adequately measured or poorly reported.
Translate the findings without exaggerating certainty, hiding important limitations, or using cure language.
Every completed Atlas review should identify its evidence sources and when the evidence was last reviewed.
Harms and Concerns
The Atlas will report potential harms separately from effectiveness and identify when harms were not adequately measured or reported in the underlying research.
What We Do Not Know
Each review can identify missing populations, short follow up, small studies, unmeasured quality of life outcomes, weak harms reporting, or other limitations that affect how the evidence should be interpreted.
Transparency Is the Standard
Completed reviews will show the evidence used, important limitations, certainty reasoning, population representation, evidence gaps, and the date the evidence was reviewed.
Important: The Support Evidence Atlas is a decision support and educational resource. It is not individualized medical advice and does not replace evaluation or guidance from an appropriate qualified professional.