ESTRO 2021 Abstract Book

S1469

ESTRO 2021

Conclusion The pre-trial quality assurance demonstrated a substantial variation in the delineated CTV volumes in SRT. A consensus meeting will contribute to a standardized delineation practice within this trial. Furthermore, the dose constraints of the PERYTON-trial are considered feasible.

PO-1746 Redash based radiotherapy surveillance system R. Ayala 1 , A. García 1 , J.M. Becerro 1 , Á. Soza 1 , R. García 1 , G. Ruiz 1 , M.J. García 1 1 H.G.U. Gregorio Marañón, Servicio de Dosimetría y Radioprotección, Madrid, Spain

Purpose or Objective R&V systems are currently an essential tool in the RT treatment process. These systems house a large amount of frequently underused data that can be further exploited to identify a wide range of potentially correctable adverse errors and “near miss” events. Moreover, it is essential that errors are addressed at a fast pace in order to reduce their frequency and their potential negative consequences. We have developed a continuous surveillance system for Mosaiq (Elekta AB) based on the open source software Redash ( https://redash.io ). Although Redash can be used as a powerful data visualization tool, it can also perform scheduled queries and trigger alerts when they do not return the expected results. Several error prone processes are monitored with SQL queries that run against the Mosaiq DB. Email alerts are sent to the physics department when a preconfigured condition is met. Materials and Methods Mosaiq data is stored in a MSSQL database (DB) that is accessible from other computers from the same network. Redash software is installed on a hospital server and connects to Mosaiq database with read-only permissions to avoid potential DB corruption. At the moment, Redash monitors the items in table 1. Mosaiq item Query Alert condition Stage of the RT process Severity level* Course/Si te Check for documents in plan Incomplete documentation Pre-treatment C L (Low) Beams Check for beam overrides Beam overrides found Treatment C M (Med)

QA delivered beams by non-physics staff Beam tolerance does not match type of beam

QA delivered beams

Treatment

C H

(High)

Check beam tolerances

Pre-treatment

C M

(Med)

Check

planned

vs.

Changes machines

between

non-twin

Site Setup

Treatment

C H

(High)

delivered linac

Check site setup fields

Missing fields

Pre-treatment

C L

(Low)

Check privileges within groups

Users

Privilege escalation

Whole RT process C H

(High)

Table 1 Mosaiq items monitored by Redash queries. *Severity levels based on Risk Matrix Methodology (IAEA- FORO, 2013a;2013b; Vilaragut et al., 2013) All Redash queries are run during the night to mitigate a possible performance reduction during clinical hours. Results The surveillance system has been running for one month. During this period, 76.4% (3986) of the non-QA irradiated beams were overridden, all of them were couch overrides except for 5 related to wrong electron applicator block codes (fig1).