Review of the patient positioning reproducibility in head-and-neck radiotherapy using Statistical Process Control

Radiother Oncol. 2018 May;127(2):183-189. doi: 10.1016/j.radonc.2018.01.006. Epub 2018 Jan 31.

Abstract

Background and purpose: A remarkable improvement in patient positioning was observed after the implementation of various process changes aiming to increase the consistency of patient positioning throughout the radiotherapy treatment chain. However, no tool was available to describe these changes over time in a standardised way. This study reports on the feasibility of Statistical Process Control (SPC) to highlight changes in patient positioning accuracy and facilitate correlation of these changes with the underlying process changes.

Materials and methods: Metrics were designed to quantify the systematic and random patient deformation as input for the SPC charts. These metrics were based on data obtained from multiple local ROI matches for 191 patients who were treated for head-and-neck cancer during the period 2011-2016.

Results: SPC highlighted a significant improvement in patient positioning that coincided with multiple intentional process changes. The observed improvements could be described as a combination of a reduction in outliers and a systematic improvement in the patient positioning accuracy of all patients.

Conclusion: SPC is able to track changes in the reproducibility of patient positioning in head-and-neck radiation oncology, and distinguish between systematic and random process changes. Identification of process changes underlying these trends requires additional statistical analysis and seems only possible when the changes do not overlap in time.

Keywords: Head and neck; Patient positioning; Quality management; Radiotherapy; Statistical Process Control.

MeSH terms

  • Head / radiation effects
  • Head and Neck Neoplasms / radiotherapy*
  • Humans
  • Neck / radiation effects
  • Patient Positioning / methods*
  • Radiotherapy Planning, Computer-Assisted
  • Radiotherapy, Intensity-Modulated
  • Reproducibility of Results
  • Statistics as Topic