Handling BLLQ data
- Below Lower Limit of Quantification
- not always possible to improve the LLOQ of the bioanalytical assays [📖]
- ignoring data below LLOQ [📖]
- biased and slightly overpredicted near or below LLOQ
- BLQ observations should be considered if the fraction BLQ is larger than 10% [📖]
- censoring BLQ in data with a low fraction BLQ (10%) [📖]
- 1-CMT model: minor impact
- 2-CMT model: more impact
- 4 types of methods
- the improvement depends on model complexity and fraction of observations below LLOQ [📖]
M1
- Censoring observations below LLOQ
M2
M3
- the likelihood for the observations below LLOQ is assessed
- may not be possible for more complex models
- generally provides more accurate parameter estimates over M1 [📖] [📖]
M4