For each research question addressed by the systematic review, the body of evidence was consolidated into evidence statements and rated according to the matrix shown in Table 2.2, which considers five domains: evidence base, consistency, clinical impact, generalisability and applicability. For included studies, evidence base and consistency were derived directly from the literature identified for each research question, whereas clinical impact, generalisability and applicability were assessed with guidance from the CRG. To ensure that the best available evidence was used, studies of higher levels of evidence (i.e. Levels I or II) were included in preference to those presenting lower levels (i.e. Levels III or IV) of evidence. This was done to minimise the potential for bias in the evidence base for each systematically reviewed question. However, lower level studies were reviewed where evidence was not available in higher level studies for any of the primary outcomes.
| Component | A - Excellent | B - Good | C - Satisfactory | D - Poor |
|---|---|---|---|---|
| Evidence base | Several Level I or II studies with low risk of bias | One or two Level II studies with low risk of bias or a systematic review, or multiple Level III studies with low risk of bias | Level III studies with low risk of bias, or Level I or II studies with moderate risk of bias | Level IV studies, or Level I–III studies with high risk of bias |
| Consistency | All studies consistent | Most studies consistent and inconsistency can be explained | Some inconsistency reflecting genuine uncertainty around clinical question | Evidence is inconsistent |
| Clinical impact | Very large | Substantial | Moderate | Slight or restricted |
| Generalisability | Population/s studied in the body of evidence are the same as the target population for the guideline | Population/s studied in the body of evidence are similar to the target population for the guideline | Population/s studied in the body of evidence are different to the target population, but it is clinically sensible to apply this evidence to the target population for the guideline | Population/s studied in the body of evidence are different to the target population and it is hard to judge whether it is sensible to generalise to the target population for the guideline |
| Applicability | Directly applicable to the Australian health-care context | Applicable to the Australian healthcare context, with a few caveats | Probably applicable to the Australian health-care context, with some caveats | Not applicable to the Australian healthcare context |
Evidence statements were only transformed into ‘action-oriented’ recommendations where:
- the body of evidence was sufficient – that is, wherever the evidence yielded support for recommendations of at least NHMRC grade C (see Table 2.3)
- the question type was interventional – that is, it evaluated the effectiveness of an intervention.
The recommendations were carefully worded to reflect the strength of the body of evidence.
| Grade | Definition |
|---|
| A | Body of evidence can be trusted to guide practice |
| B | Body of evidence can be trusted to guide practice in most situations |
| C | Body of evidence provides some support for recommendation(s) but care should be taken in its application |
| D | Body of evidence is weak and recommendations must be applied with caution |
Where there was insufficient quality or quantity of evidence, it was not possible to develop evidencebased recommendations. In this situation, the CRG developed practice points through a consensus-based process, to guide clinical practice.
For prognostic and aetiological questions, the evidence base provided only an indication of the risk associated with a particular factor; thus, it was not possible to make an evidence-based recommendation for a change in practice. Instead, the CRG’s consensus-based process, used to develop practice points to guide practice, was informed by the prognostic and aetiologic review, and by clinical experience.