Overview of proposed confidence levels for grading mutations associated with phenotypic resistance
| Symbol | LR+ and OR | ||
| p-value | value | ||
| High (Hi) confidence for association with resistance Strong association of the mutation with phenotypic drug resistance; sufficient evidence that the mutation confers or is strongly associated with drug resistance | #• | <0.05 | >10 |
| Moderate (Mo) confidence for association with resistance Moderate association of the mutation with phenotypic drug resistance; additional data desirable for improved evidence that the mutation confers or is strongly associated with drug resistance | #• | <0.05 | 5< … ≤10 |
| Minimal (Mi) confidence for association with resistance Weak association of the mutation with phenotypic drug resistance; inconclusive evidence that the mutation confers or is strongly associated with drug resistance. Substantial additional data required | #• | <0.05 | 1< … ≤5 |
| No association with resistance No evidence of association between the mutation and drug resistance | #• | <0.05 | <1 |
| Indeterminate No statistically significant threshold reached; additional data required | Indeter | ≥0.05 | |
The table shows the thresholds applied to likelihood ratios (LR) and odds ratios (OR) to grade the association of mutations with phenotypic drug resistance. LR+: positive likelihood ratio. “Additional data” is defined as a requirement for 1) more phenotypically drug resistant and susceptible isolates tested with the mutation in question; and/or 2) better understanding of the mechanism of drug resistance (e.g. to investigate epistasis, or the interactions between drug-resistance conferring mutations, lineage-specific genetic factors and compensatory mutations [23, 24] or synergistic factors when more than one mutation is required to confer resistance [25]).