Michael Tallhamer, MSC, DABR
I use a statistical evaluation to determine thresholds for our patients. You obviously have to start somewhere so we chose +/- 5mm since our old SSD tolerances were 1.0cm. You can then dial them over time from there as your therapists gain more familiarity with the system.
To do that you can have the vendor setup the system to dump out the real-time deltas to a .csv file for all your patients and then use MS Excel (or your own software) to plot the distribution of the deltas for the various body sites you are concerned about (i.e. all your left side DIBH patients) after some period of time.
Once you have that information you can determine if the thresholds need to be tighter or looser based on the standard deviation of the distribution and set it so that you are trying to limit both Type I and Type II errors. You will find that in most cases as everyone gets more familiar with using the system the distribution will get narrower so you may need to adjust once and then re-evaluate 3-6 months later and see if it needs to be tightened up again. This will help you set tolerance both that are appropriate and representative of your patient population and the skill level of your therapists. It also is a great Process Quality Improvement project that TJC and ACR love to see on accreditation and audits so it is a win win all around for very little effort up front.
CAUTION: If you tighten them up too far then you will know as it will become too difficult to get the patient setup to within the proper range or keep them in it over the course of a fraction. You will begin to pay a time penalty which may or may not be worth the added accuracy for some body sites (e.g. don’t need 1mm accuracy on a breast as it is dosimetrically inconsequential)