European Radiology

, Volume 21, Issue 10, pp 2046–2055

Patient acceptability of CT colonography compared with double contrast barium enema: results from a multicentre randomised controlled trial of symptomatic patients

Authors

  • Christian von Wagner
    • Department of Epidemiology and Public HealthUniversity College London
  • Samuel Smith
    • Department of Epidemiology and Public HealthUniversity College London
    • Centre for Medical ImagingUniversity College London
    • Centre for Medical ImagingUniversity College Hospital
  • Alex Ghanouni
    • Department of Epidemiology and Public HealthUniversity College London
  • Emily Power
    • Department of Epidemiology and Public HealthUniversity College London
  • Richard J. Lilford
    • Department of EpidemiologyBirmingham University
  • Dion Morton
    • Department of SurgeryBirmingham University
  • Edward Dadswell
    • Department of Surgery and CancerImperial College London
  • Wendy Atkin
    • Department of Surgery and CancerImperial College London
  • Jane Wardle
    • Department of Epidemiology and Public HealthUniversity College London
  • For the SIGGAR Investigators
Gastrointestinal

DOI: 10.1007/s00330-011-2154-y

Cite this article as:
von Wagner, C., Smith, S., Halligan, S. et al. Eur Radiol (2011) 21: 2046. doi:10.1007/s00330-011-2154-y

Abstract

Objectives

To determine patient acceptability of barium enema (BE) or CT colonography (CTC).

Methods

After ethical approval, 921 consenting patients with symptoms suggestive of colorectal cancer who had been randomly assigned and completed either BE (N = 606) or CTC (N = 315) received a questionnaire to assess experience of the clinical episode including bowel preparation, procedure and complications. Satisfaction, worry and physical discomfort were assessed using an adapted version of a validated acceptability scale. Non-parametric methods assessed differences between the randomised tests and the effect of patient characteristics.

Results

Patients undergoing BE were significantly less satisfied (median 61, interquartile range [IQR] 54–67 vs. median 64, IQR 56–69; p = 0.003) and experienced more physical discomfort (median 40, IQR 29–52 vs. median 35.5, IQR 25–47; p < 0.001) than those undergoing CTC. Post-test, BE patients were significantly more likely to experience ‘abdominal pain/cramps’ (68% vs. 57%; p = 0.007), ‘soreness’ (57% vs. 37%; p < 0.001), ‘nausea/vomiting’ (16% vs. 8%; p = 0.009), ‘soiling’ (31% vs. 23%; p = 0.034) and ‘wind’ (92% vs. 84%; p = 0.001) and in the case of ‘wind’ to also rate it as severe (27% vs. 15%; p < 0.001).

Conclusion

CTC is associated with significant improvements in patient experience. These data support the case for CTC to replace BE.

Keywords

CTColonographyVirtual colonoscopyBarium enemaColonCancerPatient preference

Introduction

Colorectal cancer is the second most common cause of cancer-related death in the western world; approximately 1 in 20 people develop the disease [1, 2]. Timely and effective diagnosis is therefore a priority, and colonoscopy is currently the most frequently used test for symptomatic patients. However, colonoscopy can be associated with potentially serious adverse events, especially in older people [3]. Double contrast barium enema (BE) is the standard alternative whole-colon investigation for symptomatic patients and has been widely used for the past 40 years. However, while safer than colonoscopy, BE is less sensitive [4] and the mobility requirements can make BE difficult for some patients [5, 6].

Computed tomographic colonography (CTC) is increasingly used in frail elderly patients who might find BE difficult to tolerate and/or who are at risk of colonoscopy-associated adverse events [6, 7]. CTC is also used as an alternative to BE to complete colonic imaging after incomplete colonoscopy [8, 9]. Dissemination of CTC has been accelerated by studies suggesting that it is superior to BE for detection of cancers and large (≥10 mm) colonic polyps [1013]. Opinion leaders have therefore proposed that BE should be phased out and replaced by CTC [14]. Research also suggests that CTC is more acceptable than BE in both symptomatic [15, 16] and asymptomatic patients [17]. However, most evidence is based on intra-individual comparisons whereby patients experience both tests. Acceptability judgments are therefore relative rather than absolute, which could exaggerate differences in experience. To date, studies comparing acceptability of CTC or BE alone, have been small, non-randomised, and have compared very different samples [15].

To address these limitations we aimed to determine the acceptability of BE or CTC via a randomised controlled trial for patients being investigated for symptoms suggestive of colorectal cancer [18]. We administered a post-examination survey of patient-reported experiences so that differences in test acceptability and post-test complications could be assessed in a sizeable population for whom the impact of clinician preference on outcomes had been reduced via randomisation. Moreover, a randomised design whereby patients follow normal clinical pathways after the primary randomised procedure (i.e. receive usual care) means that results are likely to better reflect experience in normal practice than designs where patients receive both tests. We hypothesised that patients allocated to CTC would report higher acceptability (including less discomfort, less worry and higher overall satisfaction with the experience) and lower levels of post-test complications than patients allocated to BE.

Materials and methods

Participants

The SIGGAR trial is a multicentre randomised controlled trial (International Standard Randomised Controlled Trials Number 95152621) comparing CTC and standard investigation of patients with symptoms suggestive of colorectal cancer [18]. Ethical approval was granted and the trial was conducted according to guidelines for good clinical practice [19]. Recruitment ran from April 2004 to December 2007 at 21 National Health Service hospitals. Eligible patients had to (1) be aged 55 years or older, (2) be able to give informed consent, and (3) have symptoms suggestive of colorectal cancer (e.g. change in bowel habit, rectal bleeding, abdominal pain, anaemia) needing investigation by a whole-colon examination according to the clinician in charge of care. Exclusion criteria were (1) inability to undergo full bowel preparation, (2) known genetic predisposition to cancer, (3) having previously received a diagnosis of inflammatory bowel disease or colorectal cancer, and (4) having had a whole-colon examination within the preceding 6 months. The clinician in charge assessed patients’ fitness to undergo bowel preparation and indicated whether they would normally investigate the patient using colonoscopy or BE—the “standard” examination. Consenting patients were then randomised to either the indicated standard examination or CTC, creating two parallel prospective randomised multicentre studies (BE vs. CTC and colonoscopy vs. CTC), each with a 2:1 randomisation in favour of the standard test. The present study took place during the last 12 months of recruitment after a series of qualitative interviews had been conducted and analysed [20], and only considers patients randomised between BE and CTC.

Diagnostic tests

For both BE and CTC, patients underwent dietary restriction and full bowel purgation. All centres used sodium picosulphate and magnesium citrate (Picolax®, Ferring Pharmaceuticals, Slough, UK) as the primary laxative for both procedures except for two centres, which used macrogol (Klean-prep®, Norgine Pharmaceuticals Ltd., Uxbridge, UK) and sodium/meglumine diatrizoate (Gastrografin®, Bayer Healthcare, Newbury, UK) respectively for CTC. BE was performed by radiologists or experienced radiographic technicians. An intravenous spasmolytic was administered routinely (20 mg of hyoscine-N-butyl bromide) unless contraindicated. Typically, 600 ml of barium suspension was instilled via a rectal catheter and the colon distended via manual compression of an enema bag filled with air (9 centres) or carbon dioxide (11 centres). Multiple digital fluoroscopic spot images of the colon were supplemented by lateral decubitus overcouch radiographs. CTC was performed following intravenous spasmolytic (20–40 mg of hyoscine-N-butyl bromide unless contraindicated) and insufflation of either room air (6 centres) or carbon dioxide administered via an automated insufflator (14 centres); one centre used air and carbon dioxide in approximately equal amounts. Investigators followed local practice where no consensus existed (e.g. with respect to administration of intravenous contrast and faecal tagging agents). Prone and supine acquisitions were obtained using multi-detector row CT with maximum 2.5-mm collimation.

Questionnaires

Consenting patients recruited on or after December 1 2006 received a post-test questionnaire at 18 centres; the other 3 centres commenced questionnaire administration on February 12 2007. One centre performed no barium enemas. The questionnaire was either given to patients directly by a research nurse (n = 508) or mailed (n = 510) according to centre preference. Patients were asked to complete it at home the day following their test, and return in a pre-paid envelope. A reminder was mailed if they had not responded within 14 days of their randomised procedure.

The questionnaire asked patients to rate the acceptability of bowel preparation (“How acceptable did you find the bowel preparation”) using a 4-item scale ranging from “not at all acceptable” (score 1) to “very acceptable” (score 4). It also asked patients to identify the least acceptable part of the test: “bowel preparation”, “bowel test” and “other-(please specify)”. Acceptability of the experience of the randomised procedure was assessed using an adapted version of a previously validated scale to assess satisfaction with colonoscopy [21]. There were 29 items each rated on a 7-point scale anchored at the end point (e.g. “felt out of control” to “felt in control”). Several items were added to make the scale more relevant for participants having BE or CTC (see highlighted items in Appendix Table 4). In accordance with previous research [15], we divided the scale into three sub-components: satisfaction, worry and physical discomfort, each of which met the standard threshold for adequate internal reliability (Cronbach’s α = 0.84, 0.72, 0.87 respectively). Higher scores reflected greater degrees of the construct being measured by each subscale. The questionnaire also assessed patients’ experience in the 24 h following the test with respect to the following eight complaints: “abdominal pain/cramps”, “nausea/vomiting”, “faint feeling or dizziness”, “wind”, “soreness”, “soiling”, “anxiety” and “sleep deprivation”. Four response options ranged from “none” to “severe”.

Basic patient information (randomised procedure, age, gender, postcode) was obtained from data registered in the trial dataset. Each patient’s postcode was used to derive a marker of area-based socio-economic deprivation [22], the Index of Multiple Deprivation (IMD). The IMD uses census-derived indicators of income, education, employment, environment, health and housing at a small-area level to generate a scale from 0 (least deprived) to 80 (most deprived). We grouped IMD scores into tertiles of lower, medium and higher levels of deprivation.

Statistical analysis

Full data were available for gender, age and postcode. Missing values in the patient acceptability scales were imputed for individuals who had completed at least 50% of the scale using the Expectation Maximisation (EM) method [23]. This uses regression analyses to estimate missing values based on individuals with complete data and other significant variables (chosen by the researcher).

Initial analysis revealed that responses were skewed towards the upper end of the distribution (indicating high levels of acceptability), so non-parametric methods were used to test for differences between groups and any influence of patient characteristics, including age (<70 years vs. ≥70 years), sex and level of patient-reported acceptability of bowel preparation (acceptable vs. unacceptable). The Mann–Whitney test statistic was used to compare two groups and the Kruskal-Wallis test was used to examine differences between three or more groups (i.e. deprivation tertiles). Post-test complications were reported using frequency statistics and analysed using Pearson’s Chi-squared statistic.

Results

Over the study period, a total of 1,018 patients were randomised to either BE (n = 675) or CTC (n = 343) (see Fig. 1). Ninety-seven patients were excluded because they did not undergo a randomised procedure (n = 84), or because they had withdrawn consent (n = 13). The final sample of eligible patients therefore comprised 921 patients (606 BE; 315 CTC) of which 674 (60.8% female, 39.2% male; median age = 68 years) responded to the questionnaire, giving a response rate of 73.2% (450 randomised to BE and 224 to CTC).
https://static-content.springer.com/image/art%3A10.1007%2Fs00330-011-2154-y/MediaObjects/330_2011_2154_Fig1_HTML.gif
Fig. 1

CONSORT diagram of participant flow through the trial

Patients were more likely to return the post-test questionnaire if they were from a less socioeconomically deprived area (IMD median score = 13.9, interquartile range [IQR] 7.4 to 22.6 for responders vs. 16.6, IQR 8.9 to 26.6 for non-responders; p = 0.004). Response status was unaffected by gender, age, or randomised procedure. Sixteen responders failed to complete at least 50% of the satisfaction, worry and discomfort scales and were excluded from imputation. The number of patients completing questions relating to test after-effects differed by question.

Acceptability of bowel preparation

The median score for tolerability of bowel preparation was 3 (IQR 3 to 4) indicating a response of “fairly acceptable”. A substantial proportion of patients in both groups rated bowel preparation as either completely unacceptable or only slightly acceptable (23.0% of patients receiving BE and 17.3% of patients receiving CTC).

Patient satisfaction, worry and discomfort

Individuals undergoing BE were slightly (albeit significantly) less satisfied than those having CTC (median 61, IQR 54 to 67 vs. median 64, IQR 56 to 69 respectively; p = 0.003). Two individual items on this scale reached significance: “dignity” and “modesty” with patients undergoing BE responding less favourably than those having CTC (Table 1).
Table 1

Scores on the patient experience scale by randomised procedure. A 7-point score was used with higher scores reflecting greater degrees of the construct being tested

Item

BE (n = 436)

CTC (n = 222)

 

Median

IQR

Median

IQR

p

Satisfaction Scale

 Satisfied

7

5–7

7

5–7

0.131

 Staff were interested in me

7

6–7

7

7–7

0.279

 I was pleased with how it went

7

6–7

7

6–7

0.087

 Staff were warm

7

6–7

7

7–7

0.290

 Staff were informative

7

6–7

7

6–7

0.343

 Dignified

4

2–6

5

3–7

0.000

 I was interested

7

6–7

7

6–7

0.111

 Confident in staff

7

7–7

7

7–7

0.629

 No loss of modesty

5

3–7

6

4–7

0.001

 Enough privacy

7

6–7

7

6–7

0.692

Worry Scale

 Worried

3

1–5

4

1–5

0.984

 Agitated

2

1–4

2

1–4

0.127

 I was worried about what they would find

4

3–6

4

2–7

0.850

 Did not understand what was happening

1

1–2

1

1–2

0.749

 I felt puzzled

1

1–2

1

1–2

0.494

 I was confused

1

1–2

1

1–2

0.567

Physical Discomfort Scale

 Painful

3

1–4

2

1–4

0.005

 I would have preferred to be less awake

4

2–4

4

2–4

0.739

 Uncomfortable

4

3–6

4

2–5

0.003

 A bad experience

4

2–4

3

1–4

0.000

 Felt out of control

4

1–4

3

1–4

0.071

 Soreness

2

1–4

1

1–4

0.000

 Afraid of “making a fool of myself”

2

1–4

2

1–4

0.012

 Claustrophobic

1

1–2

1

1–2

0.206

 Bloated afterwards

5

3–7

4

2–7

0.031

 Intrusive

3

1–5

3

1–4

0.004

 Hard to cope with

2

1–4

2

1–3

0.000

 Difficult to do what was required

2

1–3

1

1–2

0.000

 Tired afterwards

3

1–5

2

1–5

0.135

BE barium enema; CTC computed tomographic colonography; IQR interquartile range

There were no significant differences between BE and CTC with regard to reported worry overall (median 15, IQR 10 to 20 vs. median 15, IQR 9.75 to 19 respectively; p = 0.617) nor were there any significant differences between BE and CTC for any of the individual items (Table 1).

Overall physical discomfort was rated as significantly worse for patients having BE than CTC (median 40, IQR 29 to 52 vs. median 35.5, IQR 25 to 47 respectively; p < 0.001). Significant differences were observed for 9 individual items on the discomfort subscale including: “pain”, “discomfort”, “bad experience”, “soreness”, “afraid of making a fool of myself”, “bloated afterwards”, “intrusiveness”, “hard to cope with”, and “hard to do what was required” (Table 1).

Regarding patient characteristics, only age was associated with patient experience. Specifically, among participants randomised to CTC, older patients reported less physical discomfort and worry than younger patients. No such differences were found for patients randomised to BE (Table 2).
Table 2

Patient demographics and experience of the randomised procedure

 

BE (n = 436)

CTC (n = 222)

 

Satisfaction

Worry

Discomfort

Satisfaction

Worry

Discomfort

 

(Median, IQR)

(Median, IQR)

Age

 Under 70

60, 54–66

15, 10.5–20

41, 31–53.5

63, 55–68

16, 11–20

37, 27–50

 70 Years +

62, 55–68

15, 10–21

38, 28–52

65, 57–70

13, 8–18

33, 23–45

 p

0.057

0.889

0.130

0.138

0.017

0.009

Gender

 Male

62, 54–66

14, 10–20

38.5, 29–49.3

63, 55.8–69

14, 9–19

32.5, 22–45

 Female

61, 54–67

15, 10–21

41, 29–56

64, 57.3–69

16, 10–19

37, 26–48

 p

0.859

0.491

0.104

0.343

0.544

0.066

IMD

 Low

61, 54–67

15, 10.5–20

41, 29.5–54

63, 56–68

14.5, 9–19

35.5, 25.3–47

 Mid

60.5, 54–66

15, 10–20.3

40, 30–52.3

64, 58.3–68

15, 10–19.8

37, 26–48.8

 High

63, 57–67.5

15, 10–21

39, 28–51.5

65, 55.8–70

17, 9–19

33.5, 22–44.5

 p

0.254

0.841

0.883

0.518

0.825

0.194

IMD Index of Multiple Deprivation

Post-test complications

Patients in both groups reported a number of post-test complaints. “Wind” (92% for BE vs. 84% for CTC respectively (χ2 [1, 647] = 11.15; p = 0.001), “abdominal pain/cramps” (68% vs. 57%; χ2 [1, 645] = 7.61; p = 0.007), “soreness” (57% vs. 37%; χ2 [1, 646] = 21.81; p < 0.001), “nausea/vomiting” (16% vs. 8%; X2 [1, 636] = 6.72; p = 0.009) and “soiling” (31% vs. 23%; X2 [1, 639] = 4.67, p = 0.034) (Table 3) were all more common in BE than CTC. The severity of post-procedural “wind” was greatest for BE (27% reporting ‘severe’ as opposed to 15% for CTC, χ2 [1, 647] = 11.47; p = 0.001). Similarly 2% of BE patients reported severe nausea/vomiting compared with no CTC patients (χ2 [1, 636] = 4.22; p = 0.056). No other symptoms differed significantly between the groups (Table 3). There were no serious adverse events for either procedure.
Table 3

Post-test complications by randomised procedure, i.e. BE vs. CTC

 

Test

None (%)

Mild (%)

Moderate (%)

Severe (%)

Abdominal pain/cramps

BE

31.8

26.6

29.3

13.2

CTC

42.7

29.5

18.6

9.1

Nausea/vomiting

BE

84.4

8.9

4.8

1.9

CTC

91.7

6.4

1.8

0.0

Faint feeling or dizziness

BE

76.3

16.1

7.1

0.5

CTC

73.9

18.8

6.9

0.5

Wind

BE

7.7

22.8

42.3

27.2

CTC

16.3

37.6

30.8

15.4

Bottom soreness

BE

43.5

28.2

21.9

6.4

CTC

62.9

24.0

9.5

3.6

Soiling

BE

68.6

15.2

12.1

4.0

CTC

76.7

16.4

4.6

2.3

Sleep difficulties

BE

72.4

14.0

11.7

1.9

CTC

78.2

12.6

6.0

3.2

Anxiety

BE

61.8

21.8

13.5

2.8

CTC

67.7

18.6

10.0

3.6

Least acceptable aspects of the patient experience

More than two-thirds of respondents stated that bowel preparation was the least acceptable aspect of their experience (68.9% for barium enema and 74.4% for CTC). The test itself was rated the least acceptable aspect of the experience by 24.3% of patients who had BE as opposed to 13.7% of those who had CTC. “Other aspects” (e.g. administrative issues) were cited as the least acceptable aspect by 6.8% of those who had barium enema and 11.8% of those who had CTC.

Discussion

This study reports the experience of symptomatic patients randomised to undergo either CTC or BE in order to diagnose or exclude significant colorectal neoplasia. This is the first time a randomised design has been used rather than cohort studies in which patients undergo all tests under investigation. A randomised design is advantageous because patients’ experience is not influenced by multiple colorectal tests (that would not normally have all been administered), thereby achieving a more valid representation of patient experience in daily clinical practice. Furthermore, test ordering is fixed in intra-individual studies, for example with CTC usually performed before colonoscopy, which may also bias experience.

We found that experience differed significantly according to the diagnostic test administered, with those undergoing CTC generally rating the test as more acceptable than BE. These data support previous findings from non-randomised cohort studies [1517]. In particular, we found significant differences between procedures for items measuring physical sensation (e.g. pain, soreness, intrusiveness, hard to cope with), as well as the physical requirements necessary to comply with the tests. It was reassuring that most patients having CTC did not report claustrophobia, which had been identified by previous research as a potential drawback [24]. Our findings also support and extend previous evidence suggesting that bowel preparation constitutes the worst aspect of patients’ experience overall.

We tested experience across 21 different centres. Because we wanted to enhance the generalisability of our results we did not stipulate that each centre use exactly the same bowel preparation (either for BE or CTC) as this does not reflect normal daily practice, although Picolax (Ferring, UK) was used overwhelmingly. Similarly, we did not stipulate that carbon dioxide or air be used as both gases are used for both procedures in daily practice, according to availability and preference. We did stipulate however that “full bowel preparation” was used for both procedures (i.e. full colonic purgation) so the experience would be comparable despite the use of different pharmacological agents to achieve this. While full purgation is necessary for BE, CTC is an evolving technology and it is possible that bowel preparation will be reduced or eliminated altogether in the future [25, 26], whereas BE will probably evolve no further. As a result, the acceptability of CTC will likely improve beyond the level documented in our study, strengthening its superiority to BE. The fact that CTC is more acceptable than BE while also offering superior diagnostic sensitivity indicates that it should replace BE as the standard radiological whole-colon investigation [27].

Our study had several limitations. We did not test patient experiences in the entire sample of the randomised controlled trial because we needed to make qualitative assessments of patient experience in order to design the questionnaire. While a randomised design offered the opportunity to subject patients to only one diagnostic test, it also has some drawbacks. Perhaps most importantly, patients were aware that two tests were being compared and a description of each was inevitably part of the consent process. Therefore it is possible that a pre-randomisation preference, for example for the more novel and appealing CTC (perhaps conveyed unconsciously by the investigator consenting the patient), may have biased patients who were then subsequently randomised to BE, and impacted negatively on their reported experience. Recent qualitative research indicates that for lay people, the novelty of CTC leads them to hold somewhat unrealistic expectations about its attributes compared with more standard tests [24]. The observed acceptability of the test may also have been affected by factors we were unable to control for, including additional demographic patient characteristics (e.g. ethnic background, participants’ general attitudes towards medical testing), or the clinical outcome of the test, but randomisation should balance these across conditions. It should be noted that the test was negative for cancer in most patients, reflecting the relatively low prevalence of colorectal cancer despite symptoms suggestive of the disease (e.g. rectal bleeding, change in bowel habit).

As noted already, we did not stipulate which gas was used for insufflation because both air and carbon dioxide are used for both procedures in day-to-day practice. Although it is long-established that carbon dioxide is more comfortable for patients having barium enema [28], we were surprised that 43% of centres were still using air (indeed 32% were using air for CTC). This may impact on physical discomfort (where we would expect the gas used to have most influence) but, at the same time, it is important that our data are generalizable to daily practice—universal adoption of carbon dioxide for barium enema now seems unlikely, especially since use of the test is declining. We did perform a sub-group analysis that found patients experienced significantly more physical discomfort when air was used for barium enema than when air was used for CTC (median 41 vs 35, p = 0.003). No significant difference was found when the same comparison was made for CO2. However, these analyses were not pre-specified and are underpowered.

A number of important questions arise from our findings that warrant future research. First, while it is tempting to conclude that eliminating bowel preparation will make CTC more attractive, future research is needed to directly compare patient experiences of the procedure with and without bowel preparation. Furthermore, research evaluating the extent to which patients prefer a more convenient/comfortable test (e.g. one that avoids bowel preparation) while trading off some degree of sensitivity (i.e. ability to identify colorectal neoplasia) would be of interest to both healthcare professionals and policy makers. Despite preliminary qualitative evidence suggesting unwillingness to trade any level of sensitivity [24], further research is required with patient populations. The finding that the physical requirements associated with CTC were perceived as less demanding also suggests that this test may be an attractive option for specific sub-groups, such as patients with mobility issues. Finally, from a health economics perspective, more work is required to frame the findings in a wider context: The modest, albeit consistent, finding of better patient experience among CTC patients needs to be considered alongside other aspects of the test, notably financial costs and the expense and anxiety that might be generated by the detection of incidental extracolonic abnormalities [29].

In conclusion, patients undergoing investigation of symptoms suggestive of colorectal cancer perceive CTC as more acceptable, and report fewer post-procedural side-effects, than patients undergoing BE. In conjunction with clinical efficacy data for CTC, our findings support the wider implementation of CTC for diagnosis of symptomatic colorectal disease.

Acknowledgements

This trial was funded by the UK Department of Health through the Health Technology Assessment programme of the UK National Institute for Health Research (NIHR grant HTA 02/02/01). This work was undertaken at UCL/UCLH and Imperial College London, each of whom received a proportion of funding from the NIHR Comprehensive Biomedical Research Centre funding scheme. The views expressed in this publication are those of the authors and not necessarily those of the UK Department of Health. We wish to thank the SIGGAR investigators and independent data monitoring and trial steering committees for their time and help (see web-Appendix 1).

SH is remunerated for R/D advice by Medicsight PLC, a company that develops CAD software for CTC.

Copyright information

© European Society of Radiology 2011