If a consensus approach is not possible non-consensus practice statements can be issued, but there should be clear reference to all schools of thought and consumers should be made aware of the lack of consensus (see Appendix C).
2.2.2 The relevance of the evidence
‘Relevance of the evidence’ refers to the study question’s similarity to the clinical question and the extent to which the findings from the study can be applied in other clinical settings to different patients. The types of outcomes that are measured and reported in a clinical trial determine whether all of the possible benefits and harms associated with a treatment can be assessed properly—among the factors that are important are the nature of the outcomes that have been measured in the trial and the period over which the measurements have been made.
It may be useful to consider outcomes as being of one type or a combination of two or three types: ‘surrogate’ outcomes, ‘clinical’ outcomes, and ‘patient-relevant’ outcomes. In general, the types of outcomes that clinical trials measure and report are either surrogate or clinical outcomes; these may or may not be of importance to the patient.
A surrogate outcome is commonly a physiological variable; for example, serum cholesterol concentration or blood pressure. There is a statistical association between the surrogate outcome and the clinical outcome of interest; for example, bone mineral density and fracture or measures of HIV viral load and progression of AIDS. There is a biological and pathophysiological basis for believing that the surrogate outcome is a primary determinant of the clinical outcome in the disease being studied; for example, glycosylated haemoglobin measurements and diabetic complications.
If the trials that form the basis of the recommendations are restricted to surrogate outcomes it is important to determine that there is a relationship between these surrogate outcomes and the clinical outcome of interest.
Clinical outcomes tend to be defined on the basis of the disease being studied. In many cases the outcome that appears to the clinician to be of primary concern, and that can be measured, is chosen; for example, survival in cancer, vertebral fracture in osteoporosis, peptic ulcer healing and relapse, walking distance in angina or claudication, or microbiological ‘cure’ of infection. But such an approach does not necessarily capture all of the relevant outcomes from the patient’s perspective: other factors that are often more difficult to measure may be more relevant, in particular factors relating to improved quality of life.