Accuracy required and achievable in radiotherapy dosimetry: have modern technology and techniques changed our views?

In this review of the accuracy required and achievable in radiotherapy dosimetry, older approaches and evidence-based estimates for 3DCRT have been reprised, summarising and drawing together the author's earlier evaluations where still relevant. Available evidence for IMRT uncertainties has been reviewed, selecting information from tolerances, QA, verification measurements, in vivo dosimetry and dose delivery audits, to consider whether achievable uncertainties increase or decrease for current advanced treatments and practice. Overall there is some evidence that they tend to increase, but that similar levels should be achievable. Thus it is concluded that those earlier estimates of achievable dosimetric accuracy are still applicable, despite the changes and advances in technology and techniques. The one exception is where there is significant lung involvement, where it is likely that uncertainties have now improved due to widespread use of more accurate heterogeneity models. Geometric uncertainties have improved with the wide availability of IGRT.