Cumulative radiation dose due to diagnostic investigations in seriously injured trauma patients admitted to critical care

Abstract

A retrospective review of case notes and radiology records was performed in order to estimate the amount and sources of ionising radiation multiply injured trauma patients are exposed to, during their initial investigations and subsequent critical care admission. Data were available for 431 radiological investigations from 36 patients. Results showed initial emergency department imaging (combined computed tomography (CT) and plain radiographs) contributed 70% of the total radiation dose. Overall, CT scans were responsible for 80% and plain radiographs 15% of the total radiation dose. Plain radiographs performed after the initial resuscitation period contributed the greatest number of investigations but accounted for only 8% of total radiation exposure. Median cumulative effective dose was estimated to be between 16 and 29 millisieverts, resulting in an estimated increased life time risk of carcinogenesis of between 1 in 614 and 1 in 1075 above baseline.

Introduction

In 2007, the National Confidential Enquiry into Patient Outcome and Perioperative Death published a report examining the management of trauma in England.¹ As a result of this document, a network of major trauma centres was established, and this led to an increased interest in research examining how to optimally manage the trauma patient.

As the management of polytrauma evolved computed tomography (CT) scanning rapidly became a central component in the assessment of the multiply injured patient.² The use of whole-body CT scans (from occiput to sacrum) in the early resuscitation period of trauma patients, has been shown to increase the probability of survival in some retrospective studies^3,4 but evidence of definitive improvements in clinical outcomes is lacking at the present time.⁵

Causal links between ionising radiation and carcinogenesis became apparent from the lifespan study of Japanese atomic bombs survivors, with the most recent Biological Effects of Ionising Radiation report attributing a 1% risk of carcinogenesis from exposure to 100 mSv of ionising radiation.^6,7 A study by Berrington de González and Darby⁸ estimated the risk of cancer from diagnostic radiological studies. Based on the UK population as a whole, they hypothesised that 0.6% of the cumulative risk of cancer in patients up to the age of 75 years could be attributed to ionising radiation based investigations. This risk showed geographical variation, ranging from 0.6% to 3%. For example, in North America the cumulative risk was 0.9%, with the greatest risk in Japan (>3%) due to differences in radiological practice.

As a result of this concerns have been raised regarding the radiation dose associated with the liberal use of whole-body CT,^9,10 especially as the trauma population tend to be relatively young, and critically injured patients often undergo repeated CT scans during their admission.¹¹ This has led to suggestions that selective CT scanning may be more appropriate and a large, multicentre randomised trial is in progress that compares whole-body CT scanning with conventional radiographic imaging in conjunction with selective CT scanning.¹²

Most of the current data on radiation exposure in trauma patients come from North American and Canadian populations. Obvious differences exist in funding, clinical practice and mechanism of trauma (blunt vs. penetrating) with the rate of radiological investigations in North America double that of the UK.⁸ As a result, it is not clear how applicable these data are to UK practice.

The aim of this study was to estimate the amount and sources of ionising radiation that multiply injured patients are exposed to during their initial resuscitation and subsequent stay in a UK intensive care unit (ICU). On the basis of these data, we aimed to provide an estimate of the increase in the risk of carcinogenesis in trauma patients as a result of radiological imaging.

Methods

Following approval from our institutional clinical effectiveness board, data were retrospectively collected for all adult polytrauma patients (age ≥18 years) who were admitted to the Sheffield Teaching Hospital Major Trauma Centre, via the emergency department (ED) over a six-month period (January to July 2012). Patients were included if they had been admitted following trauma that had necessitated a whole-body CT scan according to our trauma protocols and had required subsequent admission to the ICU. Demographic data and injury severity scores (ISS) were obtained from electronic patient records (Metavision, iMDsoft, Needham, MA, USA).

The type and number of radiological imaging investigations performed from admission to the ED, until discharge from the ICU, were obtained from our Picture Archiving and Communication Software. The dose area product (DAP) was measured using the on-board DAP meter and was obtained from the investigation dosimeter report stored as part of the radiology records (HSS CRIS, Healthcare Software Solutions, Mansfield, UK).

As there are multiple radiological imaging devices from a variety of different manufacturers within our institution, it was not possible to precisely measure the cumulative effective dose (CED) for each individual investigation. Instead, an estimate was calculated, using standard reference values, based on typical doses of radiation for each imaging investigation undertaken.¹³ It was not clear from the retrospective data whether patients received an uninterrupted scan or if the separate body parts were scanned individually. The radiation doses for individual scans compared to uninterrupted scans differ. For example, CT scans of chest (6.6 mSv), abdomen (5.6 mSv) and pelvis (6.0 mSv) scanned individually would have a combined dose of 18.2 mSv, whereas the value stated for combined chest, abdomen and pelvis is 10 mSv. For this reason, both values were calculated, thereby providing a conservative and liberal estimate of cumulative effective radiation dose.

Data were analysed using standard spreadsheet software (Excel 2010, Microsoft Corporation, Redmond, Washington, USA). All data are presented as median (interquartile range, [IQR]) or number (%).

Results

During the study period, 60 major trauma patients were admitted to our centre, 36 of which met inclusion criteria (78% male with median (IQR) age 38 (27–68) years) with a median (IQR) ISS of 25 (17–34) and ICU stay of 5 (2–10) days. Injuries were predominantly due to blunt trauma (Figure 1). A summary of the injuries sustained by body region is given in Figure 2. In total, 480 radiological images were analysed of which the DAP could be calculated for 431 (90%).

Figure 1.

Mechanism of injury.

Figure 2.

Distribution of injuries according to body region.

C-spine: cervical spine; T/L spine: thoraco-lumbar spine.

The median (IQR) radiation DAP was 4289 (358–6051) mGycm⁻². Initial ED imaging (combined CT and plain radiographs) contributed 70% of the total radiation dose (Figure 3). Overall, CT scans were responsible for 80% and plain radiographs 15% of the total radiation dose. Although plain radiographs performed after the initial assessment period contributed to the greatest number of radiological investigations, they were only responsible for 8% of total radiation exposure (Figure 4).

Figure 3.

Source of radiation exposure in trauma patients; comparison of number of tests performed versus total effective dose of radiation.

ED: emergency department; CT: computed tomography.

Figure 4.

Contribution of different modalities to total radiation exposure.

ED: emergency department; CT: computed tomography.

Conservative and liberal estimates of CED produced median (IQR) values of 16.32 mSv (15.03–23.08) and 28.56 mSv (25.03–35.47), respectively. The population risk of developing cancer or the heritable effects of ionising radiation exposure provided by the International Commission for Radiological Protection is 5.7 × 10⁻⁵ cases per millisievert (mSv) of exposed radiation.¹⁴ Therefore, the conservative and liberal risks in our cohort are estimated to be 1 in 1075 and 1 in 614, respectively. This is in addition to the baseline risk of fatal cancer development that is present in the population.

Discussion

The purpose of this study was to estimate the doses and sources of ionising radiation that critically injured polytrauma patients are exposed to at a UK major trauma centre. As would be expected from the epidemiology of UK trauma, our study population was mainly male, had a median age less than 40 years and had predominately suffered blunt injury. All subjects met the classification of major trauma, namely having an ISS ≥16.¹⁵ We have demonstrated that polytrauma patients are exposed to significant levels of ionising radiation, which may equate to lifetime risk of fatal and non-fatal cancer development and heritable risks of between 1 in 614 and 1 in 1075. This additional risk is difficult to quantify at an individual level, however, when it is considered that everyone has a natural lifetime risk of cancer of 1 in 3–4.¹⁶

Our conservative and liberal estimates for median radiation exposure dose were 16.32 and 28.56 mSv, respectively, which are similar to other published results, although it is difficult to draw direct comparisons between studies due to a variety of factors. The severity of injury varies between the cohorts studied as does the length of time that radiation doses were measured. Due to improvements in radiological equipment (in particular CT scanners) historical comparisons are difficult to make, as similar scanning practices will now result in much lower doses of radiation. In a study based in a US trauma centre, Sharma et al.¹⁷ estimated a mean exposure of 14.56 mSv for all trauma patients within the first 24 h of admission. A US study with similar methodology to ours described a median DAP of 1700 mGycm⁻² for trauma patients who stayed at least one day in critical care.¹⁸ This population, however, was less severely injured than ours (13% had an ISS <16), and whole-body CT scanning was not employed. It is of note that a subset of more severely injured patients in this cohort received radiation doses from CT in excess of 100 mSv. Winslow et al.¹⁹ looked at the radiation dose in the first 24 h received by less severely injured US trauma patients. In all, 92% of their population had a whole-body CT (‘pan-scan’), and the median dose was 40.2 mSv.

The length of hospital stay is obviously a key factor in cumulative radiation dose, as increased length of stay (LOS) allows more opportunity for performing imaging and suggests a more severe or complex injury that may require serial investigations. The median LOS in our study was five days. A study in which only patients with a LOS >30 days were included unsurprisingly showed higher radiation doses (mean cumulative dose 106 mSv) than that seen in our population.¹¹ Another study that measured radiation dose using dosimeters directly attached to patients for the entirety of their LOS (median 15 days) showed a median dose of 22.7 mSv.¹⁰

Despite initial trauma scanning contributing to the largest proportion of radiation exposure, the consensus amongst clinicians involved in the management of trauma is that the risks of exposure to ionising radiation at this stage are probably outweighed by the potential risk of missing occult injuries. Data from retrospective studies suggest that full body trauma CT scanning is associated with a reduction in missed injuries and overall mortality.^3,4,20 The REACT-2 study is currently examining this premise by determining whether immediate total-body CT scanning during the primary survey in trauma patients compared with conventional imaging supplemented with CT scanning reduces mortality.¹² Until the results of this study are published, whole-body CT scanning is likely to continue which means that radiation dose reduction strategies in critical care patients will have to focus on imaging undertaken after discharge from the ED.

This would mean the use of alternative imaging modalities that do not necessitate ionising radiation, such as ultrasound and magnetic resonance imaging (MRI). Although this may seem like an attractive option, the benefit of reduced radiation exposure has to be offset against the inherent disadvantages of these modalities. MRI scanning can be time consuming and the scanners are often located in relative isolation compared to CT, and as such, may not be appropriate for an unstable patient. This of course is in addition to the generic risks of the MRI environment. Ultrasound, although rapid, easily accessible, and without the need for transfer from the critical care unit, lacks the sensitivity of CT and is not suitable for all body tissues. In our cohort of patients, CT scanning undertaken after ED discharge was only responsible for 18% of total radiation exposure and as such, efforts to reduce exposure in this area may have limited effects. Plain radiographs performed after ED discharge (which will predominately be requested by critical care physicians to assess for respiratory disease or after interventional procedures such as central venous catheterisation) contribute to the greatest number of investigations, but only equate to 8% of the total radiation dose.

The main limitation of our study is that we have estimated, rather than precisely measured, the cumulative effective radiation dose. Obtaining the absorbed dose for a patient is a relatively simple procedure, as this is recorded on the radiology equipment and can be used to calculate the DAP, which takes account of the surface area being exposed to radiation. However, in order for comparisons to be made regarding radiation exposure, this unit of measurement should be converted to CED, measured in mSv. This conversion takes into account the type of radiation used and applies a weighting factor according to the type of tissues exposed, as some tissues are more sensitive to the ionising effects of radiation than others. This conversion would be a time-consuming and specialist process requiring a medical physicist and was, therefore, not possible due to logistical and financial constraints. It is for this reason that estimates were made with standard reference tools. Using this approach does have the potential to introduce some inaccuracies. Use of fluoroscopy in trauma patients is common during surgical fixation and the amount it is used is variable, so a standard value for this investigation is difficult to formulate. There were also some non-standard investigations, for which there was no value to incorporate and so were excluded from the CED. Having calculated a conservative and liberal estimate for radiation exposure, we feel it is likely given these inaccuracies that the true estimate lies somewhere between the two.

We have also been unable to contextualise the risk:benefit ratio of ionising radiation in trauma patients. At the present time, the only meaningful dose reduction strategy available to clinicians would be to reduce CT scans and plain radiographs undertaken whilst on intensive care which equates to 24% of total exposure. The potential risks associated with missing potential reversible pathologies, however, are likely to be greater than that of the radiation exposure. Indeed, it may prove impossible to modify the risks of radiation exposure in critical ill trauma patients in a meaningful way, until more advanced CT scanners are introduced. New-generation CT scanners are now able to generate images of a similar quality to older devices, but with a reduction of over 60% in dose-length product.²¹ In the interim, efforts should focus on ensuring radiation dosages in imaging investigations are undertaken in line with as low as reasonably achievable principles and strategies.²²

Conclusion

Seriously injured UK trauma patients who are admitted to critical care are exposed to significant levels of ionising radiation. This is associated with an estimated lifetime risk of increased cancer development of between 1 in 614 and 1 in 1075. CT imaging in the initial assessment contributes to the largest dose, though a significant contribution comes from subsequent imaging, particularly further CT scanning. Although plain films are responsible for the greatest number of tests, they contribute only 15% of the total radiation dose, with plain films taken after ED discharge contributing 8%. Clinicians should be aware of the cumulative dose of radiation trauma patients are exposed to and consider alternative imaging modalities where possible.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.