START provides the following diagnostic and treatment options. The “standard” and the “individualised” options are coupled with ranked types of basis.
(“standard”, “recommended” or “not recommended”)
This can be considered a conventional choice for the average patient.
(“suitable for individual clinical use”)
This is not a standard option, but it can be a reasonable choice for the individual patient.
The patient should be informed that the option is not standard and the decision must be shared with the patient.
This is something which, in principle, can be offered to the patient only within a clinical study.
Type of Basis
START provides an appropriate basis for each clinical option.
Types of basis are ranked in five levels.
“TYPE C basis”
There is a widespread consolidated consensus. Randomised trials have not been carried out or have been inadequate, but the issue is settled without major controversy:
currently, no (further) experimental evidence is felt to be needed.
“TYPE 1 evidence”
(Randomised trial(s) available, strong evidence)
Consistent results have been provided by more than one randomised trials, and/or a reliable meta-analysis was performed. In some instances, one randomised trial can be considered sufficient to support this type of evidence.
Further confirmatory trials do not seem necessary.
“TYPE 2 evidence”
(Randomised trial(s) available, weak evidence)
One or more randomised trials have been completed, but the evidence they provide is not considered definitive (their results are not consistent, and/or they are methodologically unsatisfactory, etc.).Some controlled evidence has therefore been provided, but confirmatory trials would be desirable.
“TYPE 3 evidence”
(External controlled comparisons available)
Evidence is available from non-randomised studies, with external controls allowing comparisons. Some uncontrolled evidence has therefore been provided, but trials would be desirable.
“TYPE R basis”
Little or no direct evidence from clinical studies is available. Yet clinical conclusions can be rationally inferred from available data and knowledge (e.g. by rationally combining pieces of information from published studies and observations; for a rare neoplasm, or presentation, through analogy with a related, more common tumour, or presentation; etc.).
The inference can be more or less strong, and trials may, or may not, be desirable
(although sometimes unfeasible).