Accidental exposures in modern radiotherapy

Draft document: Accidental exposures in modern radiotherapy
Submitted by Maria do Carmo Lopes, IPOCFG, E.P.E.
Commenting as an individual

Preventing Accidental Exposures from New External Beam Radiation Therapy Technologies General comment This is a very important document. All those that have been involved in the implementation of new techniques in modern radiotherapy can see here many of their concerns. A safety culture starts from each of us but needs the contribution of this kind of recommendations and views. 1. Small corrections: Page Paragraph Line Where it is written Suggestion 20 3.1. 13 A number or reports A number of reports 21 (39) 1 (39) (39) 26 (49) 1 (49) (49) (not bold) 26 (49) 2 between, devices between devices 28 (57) 1 (57) (57) (not bold) 31 Table last line The Comment for the last Task The sentence is not completed 35 (70) 9 photons protons 41 1, 7 (ref.) References are missing 68 (156) 8 Thomadsen (2003) No reference at the end of the chapter 71 (165) 1 diagram below No diagram is included. Does it refer to Fig.3a? 71 (165) 5 calibration output 71 (166) 7 (see Fig. 3) (see Fig. 3a) 71 (166) 9 fig xx Fig. 3b 72 (168) diagram It is not clear what is the diagram 75 Table 5.1 last line 1,4% 91,3% 1.4% 91.3% the same decimal sign as above 77 5. 1 . Independent Independent 90 (207) 20 FORUMs FORA 2. Small suggestions: Page Paragraph Line Where it is written Suggestion 8 (1) 9 (IGRT), gated radiation therapy, radiosurgery Include robotic systems in this enumeration 8 (3) 6 (for example the Accuray machine) add: or the Cyberknife 17 (30) 3 and may not even be feasible in the case of IMRT Delete this last sentence. EQUAL-Dose is an example of an independent MU calculation software for IMRT but there are many other certified ones 21 (37) 12 irregular fields elementary beam segments 24 (46) 9 different outcomes different dose distributions 25 Footnote 1 3 the beams are defined on the computer For IMRT the beams definition as part of the virtual simulation (VS) does not make sense. In the context of new techniques it would be appropriate not to include this step of VS 28 (56) 3 post-treatment during-treatment 29 (59) 1: geometrical or dosimetric deviation from the “correct” plan without expected significant clinical consequences (low severity) low probability, low severity. The scoring for 1: should be equivalent of that for 2: and 3: 33 (61) 2 from the TPS to the treatment machine from the TPS to the R & V system 35 (69) 7 the patient shift aims at the patient shift is calculated from 39 (77) The R&V systems are perhaps nowadays still the most used ones, so the past phrasing seems not too appropriate Concerning the session Accidental exposures involving single and multiple patients: From the 443 failure modes, we can see, from Table 5.1., that 8 event sequences are responsible for 91.3 % of accidental exposures involving multiple patients. But then in (180) it is said that TPS is the most vulnerable step with a contribution of 93% to the total probability of accidental exposures involving multiple patients. If we see the table, the TPS is involved in 45.5+32%=77.5% over 91.3%. The percentage of 93% is not clear. Also in (181) it is said that Single patient accidental exposures represent 98% of all accidental exposures. This means that the preceding paragraphs (multiple patients) refer just to 2% of the accidental exposures. Perhaps it would be good to say it at the beginning. In (182) it seems that the sum of the contributions of the various steps should amount in 100% and not 112%. 3. Some contributions (suggestions): Paragraph (28), page 17 It is not clear what is said in this paragraph. If what is meant is referring an independent MU calculation as a safety layer to verify the MU calculation from TPSs, it is not clear how this kind of verification can lead to accidental exposures… If this is the case, a remark should be done that any independent software for MU calculation should be properly commissioned before put in clinical use. On the other hand if the intention is to refer an independent MU calculation apart from the calculation of dose distribution in the TPS, I think that nowadays, the most common is that the MU are directly calculated in the TPS, so the phrasing is not appropriate. Another comment to this paragraph is the last sentence “Mistakes were also made at treatment simulation due to incorrect labeling of the images resulting in treating the wrong side of the patient”. This sentence is not related to the anterior issue. It should be included in an independent paragraph. The particular error in “treatment simulation” should be included in a more general type of errors concerning images (correct labeling, correct orientation, etc.) and not just concerning simulation. This kind of error can also involve a clinical responsibility namely during delineation by the physician, using different image sets (CT, MR, etc.). The target volume delineation should not be independent of a conscientious knowledge of the clinical case. Paragraph (32), page 18 Many other causes can be mentioned namely related with the new automatisms like: automatic treatment sequences that are repeated or missed for a given patient, automatic couch movements wrongly aloud during treatments, etc. Paragraph (39), page 22 Should it be useful to include here a reference to the “pre-paid” commissioning services that are offered by the linac companies? The linac can be purchased with the commissioned done by the company. This is sometimes quite attractive to some hospital administrations that think they can save time to start treating patients. In my opinion the local physicists will miss then an important phase of knowledge of that particular machine. Paragraph (40), page 22 Some reference to a complete quality control program should be included. The complete quality control program, involving periodic tests (daily, weekly, monthly, etc.), is mandatory. Also the training of emergency procedures is an important safety layer to prevent major accidents. Paragraph (51), page 26 A mention to stereotactic fixation devices should be done – starting with special thermoplastic masks, with dental support strips and nose bridges to head rings and body frames. Paragraph (54), page 27 Mention that concerning this issue there are basically two types of TPS - capable of using custom CT number to electron density tables or vendor supplied tables. Paragraph (55), page 28 May be it would be good to mention that this kind of errors does not occur in IMRT planning. Once more the reference to virtual simulation should be updated with the reality of the new techniques where the definition of the beams at the time of simulation does not make sense. Virtual simulation for the new techniques is focused in image acquisition and reference set-up. Paragraph (56), page 28 Also the doses for IGRT with kV-Cone Beam CT or MV-Cone Beam CT should be taken into account and referred here Table 3.1. , page 30 The Comment of the first Task (Preparation of beam data library (parameterization) should include “also geometric parameters like angle orientation for couch, collimator, gantry, etc”. In fact the “consistency of collimator and table rotation scales” mentioned in the Comment of the eighth Task (Beam set-up definition) comes from this first task. Another comment to this table is that in modern RT departments, sometimes, the patients structures (target volumes and critical organs) may be delineated in an external system and then imported as a RT Dicom object into the TPS. Caution should be paid in this data transfer and I think that a mention to this possibility should be included in this table. The Comment and Risk Index to the Task of Beam set-up definition are typical for conventional RT, not modern techniques. Paragraph (60), page 32 It would be convenient a remind to: Independent MU calculation is an efficient safety layer to prevent such accidental exposures. Session 3.4, page 33 The last bold paragraph referring dynamic wedges should include something like: Also treatment interruption with dynamic wedges should follow proper procedures (like “treatment resumption”) that should be trained to prevent wrong irradiations (repeating the whole wedges beam, or loosing the already treated part of the beam) Paragraph (65), page 33 May be it should be added something like: A daily verification of the table absolute scales is mandatory. Paragraph (69), page 35 After the reference (line 2) to “under the overall responsibility of the radiation oncologist” I would add (for instance a tolerance level like – shifts below 3 mm are not corrected). Paragraph (75), page 38 I would add: Also proper periodic tests should be developed in order to assure the maintenance of the linac calibration Page 39, after (78) I would include a (79) paragraph here with a mention to the importance of the step “Approval” in R&V systems. In fact, most of the referred errors involving R&V systems would be prevented by the implementation of the approval of any significant changes (in beams, in doses, etc.) by a qualified physicist. This is a common feature to all R&V systems that is often skipped or misused. This is linked to security privileges that have to be defined for the different users (physicians, technologists, physicists). Paragraph (183), page 76 I think that it would be good to comment on the delineation phase, more profoundly. Radiation oncologists sometimes have no training in different image modalities. If they delineate on the base of a single image set (CT) they can make mistakes. Fusion functionalities of TPS (or other systems) have to be approached in a multi-disciplinary way involving radiation oncologists, radiologists, surgeons, etc. Page 77, point 5. Shouldn’t it be important to add an ending sentence like “Independent MU calculation should be implemented”? Figure 3a. , Page 80 Here it is the suggestion for the labelling: (Unfortunately I could not paste the figure with the suggested labelling...) Probably it would be a good idea to let the attachment of a comment file, instead of printing it here. Paragraph (194), page 84 I would add: The “customer information bulletins” (CIB) are a way to disseminate the reported failures or malfunctions of the TPS, for instance. These CIBs emanating from the manufacturers require, at the hospital level, some kind of operational procedure in order to assure that all users are informed, that they have understood the problem and the proposed work-around. Coimbra (Portugal), April 13, 2009 Maria do Carmo Lopes, PhD. Serviço de Física Médica IPOCFG, E.P.E. Av. Bissaya Barreto 3000-075 Coimbra Portugal Email: Telephone: +351-230 400346 Mobil Phone: +351-962471324