[111455091]
A clinical near-miss was reported from a clinic. When checking the dose distribution for an extreme off-axis field, it was noticed that the dose in the corners did not match that of the previously used planning system. When investigating the beam model, it was found that the beam profile corrections had inappropriate values outside the 20cm radius. The issue was detected before treatment and no patient was affected. This is abnormal use, using the system contrary to instructions. The responsibility of the user to validate the beam model for all clinical situations for which it is to be used is clearly stated in the ifu, giving field size and off-axis fields as explicit examples of what needs to be included. The clinic's responsibility to validate the beam model is also "common knowledge" and has not been disputed by the customers. In these cases, the validation done by the clinics has been insufficient. We are still investigating why this is and the extent of validation done, but it is clear that they have not validated dose for field size used or for extreme off-axis fields. Other beam models exhibiting the same unreasonable beam profile correction have been identified, so it may be common to skip some of the required validations. As is also stated in the ifu, all beam model parameters must be reviewed before model commissioning. Although the rayphysics manual clearly states that the auto-modeling may cause unreasonable values for radii points larger than the measured curves and that manual tuning often is necessary, this has not been done. This could be because users have put too much faith in the auto-modeling, neglecting the required parameter review. In general, failure to adhere to instructions could be caused by time constraints or information fatigue. There is no risk when using the system according to instructions. However, as the reports show, this is not always done. Based on the severity and probability of occurrence only, the risk is high. The high detectability reduces the risk, but there is still a significant risk of serious harm resulting from this issue.
Patient Sequence No: 1, Text Type: D, B5