Advice
Clinical and technical evidence
Clinical and technical evidence
A literature search was carried out for this briefing in accordance with the published process and methods. This briefing includes the most relevant/best publicly available evidence relating to the clinical and cost effectiveness of the technologies. The literature search strategy and evidence selection methods are available on request by contacting mibs@nice.org.uk.
Published evidence
This briefing summarises 10 studies involving CT scans, fluoroscopically guided procedures (FGP), X‑rays, positron emission tomography (PET) scans and PET/CT.
One study reported results using DoseTrack (Kim et al. 2016), 1 using OpenREM (Platten and Thomas 2015) and 8 using DoseWatch. No published evidence was found for the other technologies included in this briefing.
Table 3 summarises the clinical evidence as well as its strengths and limitations.
Strengths and limitations of the evidence
Published evidence is only available for 3 of the included technologies (DoseTrack, DoseWatch and OpenREM), most of which comes from retrospective observational studies. Although prospective data gathering ensures that important values such as patient weight are correctly documented, the same bias effect is not present for these software systems as for other medical technologies, because standard collection of data can be done retrospectively with equivalent accuracy. Five of the studies were only available as conference abstracts and so there was limited information on their methodology.
The 2011 review of the Public Health England report CRCE-013: Doses from CT examinations in the UK includes data from 47,000 patients across 127 centres (an average of 370 examinations per centre). The total number of CT examinations analysed in 3 single-centre studies included in this briefing equates to more than 1,000 examinations per centre, demonstrating that radiation dose monitoring technologies may be used to collect and analyse sufficiently large datasets (Boos et al. 2015, Heilmaier et al. 2016, Manco et al. 2016).
None of the studies provided any information on the amount of time and staffing needed to use the software.
Comparative studies assessing the usability or functionality of 2 or more of the technologies would improve the evidence base. Radiation dose monitoring technologies do not have any measurable directly related clinical outcomes, so studies should be designed to support their main functionality claims (particularly their ability to collect and analyse large datasets fast and efficiently).
Table 3 Summary of selected studies
|
Study size, design and location |
296 CT scans, retrospective observational study, multicentre, Belgium. |
|
Intervention and comparator(s) |
DoseWatch, no comparator. |
|
Key outcomes |
The radiation dose monitoring software identified that there was a consistent dose level between scanners in the same hospital. Large dose variations were observed between hospitals. The radiation dose monitoring software revealed that erroneous selection of adult protocols for children occurred. |
|
Strengths and limitations |
There was a relatively small sample size (<300). Results were pooled over 5 different scanners with technological differences, which can affect the lowest achievable radiation dose. |
|
Study size, design and location |
357 FGP, prospective observational study, single-centre, Switzerland. |
|
Intervention and comparator(s) |
DoseWatch, no comparator. |
|
Key outcomes |
The software successfully transferred dose data from all the procedures. Highest dose values were seen during transarterial chemoembolisation procedures. Complex cases were associated with higher doses, whereas there was no direct correlation of dose and total fluoroscopy time. |
|
Strengths and limitations |
Although operators based their grading of case complexity of the FGP on predefined criteria, grading remains subjective to a certain degree. C‑arm angulation was not recorded. It is well known that this is an important parameter in the calculation of the dose delivered. |
|
Study size, design and location |
8,883 CT scans, retrospective observational study, single-centre, Switzerland. |
|
Intervention and comparator(s) |
DoseWatch, no comparator. |
|
Key outcomes |
A total of 316 alerts were generated by the system. The overall alerts percentage ranged from 2% to 5% and the main reasons were high BMI and patient being positioned off‑centre. |
|
Strengths and limitations |
Scanners from only 1 company were used. In some patients, alerts may have been caused by more than 1 source, but the authors tried to estimate which of the causes might have been the principal reason. |
|
Study size, design and location |
30,045 conventional digital X‑ray images, retrospective observational study, single-centre, Switzerland. |
|
Intervention and comparator(s) |
DoseWatch, no comparator. |
|
Key outcomes |
Radiation dose values decreased significantly after implementation of the dose monitoring software. The software successfully transferred dose data from the procedures. |
|
Strengths and limitations |
Data were obtained from only 1 department, so radiographs of certain areas of the anatomy were limited. The department was fully equipped with digital detectors, so the data may not be representative of institutions still using film-screen systems. |
|
Study size, design and location |
5,359 X‑rays, 413 CT scans, 98 PET scans, 82 PET/CT scans, retrospective observational study, single-centre, Korea. |
|
Intervention and comparator(s) |
DoseTrack, no comparator. |
|
Key outcomes |
The CT component of PET/CT scans contributed nearly half of the total cumulative dose in children with neuroblastoma. The radiation dose received from X‑ray was higher than expected because of the large number of images. The software was used to analyse the cumulative radiation dose attributed to different imaging modalities in children having multiple imaging investigations because of cancer. |
|
Strengths and limitations |
The authors did not consider changes in study protocols or equipment models for the 12‑year length of the study. |
|
Study size, design and location |
30,000 CT scans, retrospective observational study, single-centre, Italy. |
|
Intervention and comparator(s) |
DoseWatch, no comparator. |
|
Key outcomes |
Using an iterative reconstruction, average effect doses were reduced by 25% for thorax-abdomen scans and 30% for head scanning protocols. In addition, dose reductions of 52% on body and 30% on neurology CT scans were achieved. |
|
Strengths and limitations |
This study was published as an abstract so there is limited information to assess its methodological quality. |
|
Study size, design and location |
45 CT scans, retrospective observational study, single-centre, France. |
|
Intervention and comparator(s) |
DoseWatch, no comparator. |
|
Key outcomes |
5 examinations, representing 12% of the total, were above the dose alert threshold. |
|
Strengths and limitations |
This study was published as an abstract so there is limited information to assess its methodological quality. Additionally, the study is based on a limited number of scans. |
|
Study size, design and location |
1,393 interventional procedures (368 IR and 1,025 NIR), retrospective observational study, single-centre, US. |
|
Intervention and comparator(s) |
DoseWatch, no comparator. |
|
Key outcomes |
10 of 368 (2.7%) IR and 52 of 1,025 (5.1%) NIR procedures exceeded estimated doses of 5 Gy with reported reference point Air Kerma (kinetic energy released per unit mass of air). All IR cases were abdominal/pelvic trauma angiograms with/without embolisation; there were no reported tissue reactions. 5 of the 49 (10.2%) NIR patients reported skin/hair injuries, all of which were temporary. |
|
Strengths and limitations |
This study was published as an abstract so there is limited information to assess its methodological quality. |
|
Study size, design and location |
70 CT aortic angiograms, retrospective observational study, single-centre, UK. |
|
Intervention and comparator(s) |
OpenREM, no comparator. |
|
Key outcomes |
Mean DLP was 1,180 mGy cm for CT aortic angiograms with a median of 1,030 mGy cm, which is comparable to the national reference dose for this type of exam (1,040 mGy cm). |
|
Strengths and limitations |
This study was published as an abstract so there is limited information to assess its methodological quality. The sample size was relatively small. |
|
Study size, design and location |
2,165 FGPs, prospective observational study, single-centre, Italy. |
|
Intervention and comparator(s) |
DoseWatch, no comparator. |
|
Key outcomes |
From the 32 patients enrolled in the follow‑up programme, 1 case of transient alopecia was observed at 1 month. Patients at risk of skin injuries because of high doses from FGPs were followed up as an integral part of the radiation dose management protocol. |
|
Strengths and limitations |
This study was published as an abstract so there is limited information to assess its methodological quality. |
|
Abbreviations: DLP, dose length product; FGP, fluoroscopy-guided procedures; IR, interventional radiology; NIR, neuro-interventional radiology; PET, positron emission tomography. |
|