Table 1 Article quality rating
Quality rating questions | Quality categories |
|---|---|
• Were the test(s) clearly described (number of loci tested, MMR genes, etc?) AND did the Index test(s) meet NIH standards? | • Good: Studies with a low risk of bias and minimal concerns of applicability |
• Was the spectrum of patients/tumors representative of the patients/tumors who will receive the test in practice? | • Fair+: Studies with some risk of bias or concerns regarding applicability; testing does meet NIH standards |
• Was the patient (sample) selection process from the source population (retrospective studies) clearly described? If prospective, were patient selection criteria clearly described? | • Fair -: Studies with some risk of bias or concerns regarding applicability; testing does not meet NIH standards |
• In a retrospective study, were selected samples representative (50% of original sample number; not statistically different on key characteristics e.g. stage distribution) of the original complete sample set? | • Poor: Studies with a significant risk of bias or greater concerns regarding applicability |
• Were patient withdrawals (prospective) or sample losses (retrospective) from the source population explained? | |
• Were un-interpretable, indeterminate, or intermediate test results reported? (Includes samples with insufficient DNA) | |
• Was follow-up sufficiently long? (minimum 3 years) | |
• If prospective, was treatment assignment blinded to MSI status? |