Cancer Risk and CT Scans: Understanding the Findings of a Data Linkage Study

[ZeGato]

Hello,

I don't have a background in medical research, but I was interested in understanding the conclusions of this paper: Cancer risk in 680 000 people exposed to computed tomography scans in childhood or adolescence: data linkage study of 11 million Australians

I'm not understanding a few things in the conclusion since I'm not familiar with some of the terms. In the subsection Discussion > Comparison with other studies, it says:

For lag periods of one, five, and 10 years, the incidence rate for all cancers combined increased by 24%, 21%, and 18%, respectively, in the CT exposed group compared with the unexposed group (table 3). The estimated numbers of excess cancers attributable to CT scans for these three lag periods were 608, 402, and 209, respectively. These excesses were, at most, 1% of the 60 674 cancers observed to date in this cohort of 10.9 million people.

So I have 2 questions regarding this quote:

1. For a lag period of one (this means that the cancer incidence rate started being measured one year after the CT scan, discarding patients who were diagnosed with any type of cancer before 1 year had passed after the CT scan), the incidence rate, when compared to the unexposed patients, for all cancer, equals 24%.
If I understood correctly, this means that there is a 24% higher chance of getting cancer if a CT scan was performed one year ago, when compared to an unexposed patient with the same characteristics (age, sex, CT scan location, etc...)? Is this the same as the odds ratio?

2. Then it says "The estimated numbers of excess cancers attributable to CT scans for these three lag periods were 608, 402, and 209, respectively. These excesses were, at most, 1% of the 60 674 cancers observed to date in this cohort of 10.9 million people".
If the incidence rate is 18% to 24% higher in the exposed group compared to unexposed group, shouldn't there be 18% to 24% excess cancers due to CT scans, instead of 1% (at most)?

Thank you for any help!

 

[Tom.G]

Sounds like the CT scan triggered an earlier Cancer diagnosis but the long-term increase in Cancer was approx 1%. My way of viewing the results is that the 24%, 1 year increase was in those people that were already heavily pre-disposed to contracting Cancer anyhow and the CT scan hurried the process along.

 

[ZeGato]

Sounds like the CT scan triggered an earlier Cancer diagnosis but the long-term increase in Cancer was approx 1%. My way of viewing the results is that the 24%, 1 year increase was in those people that were already heavily pre-disposed to contracting Cancer anyhow and the CT scan hurried the process along.

The incidence rate ratio was 24% for the entire follow-up period, which was an average of 9.5 years. Table 5 shows that this higher incidence rate ratio remains fairly constant during the follow-up, even for 15+ years. Of course, patients diagnosed with cancer were immediately excluded from further calculations, so that the higher incidence a few years after the CT scan doesn't influence subsequent incidence rates.

 

[russ_watters]

I think this is the part you missed integrating into the conclusions:

I don't have a background in medical research, but I was interested in understanding the conclusions of this paper: Cancer risk in 680 000 people exposed to computed tomography scans in childhood or adolescence: data linkage study of 11 million Australians

Only 6% of the people studied got CT scans. If the incidence rate for the people who got CT scans was 24% higher, then 6% of 24% is 1% for the entire group.

 

[ZeGato]

I think this is the part you missed integrating into the conclusions:

Only 6% of the people studied got CT scans. If the incidence rate for the people who got CT scans was 24% higher, then 6% of 24% is 1% for the entire group.

Oh that's right, thanks.

Considering that the increased incidence rate continues to show those same high values even after 15 years of follow-up, I'd say this is pretty concerning... If the IRR doesn't go down with time, and considering that the lifetime risk of cancer is about 41%, this could mean a 24%*41% = 9.8% additional probability of getting cancer over a lifetime, due to one CT scan. If this was proven to be true, CT scans usage would of course be greatly reduced, since it's an absurdly high risk for a diagnosis technique, that many times can be replaced by the safer MRI or ultrassound. This research didn't go that far though, since the oldest members were in their early 40s by the end of the study, but it'd be indeed interesting to see the results when extended to those older ages. Other research I've seen on the topic give much lower estimates, of around 1 in 1000 increased odds of getting cancer over a lifetime due to a pediatric CT scan. However, some extrapolate the results from the cancer incidence rates of Japan's nuclear bombings survivors, for the same amount of radiation as regular CT scans, while others use data from pediatric patients exposed to CT scans, but they also have a short follow-up period such as this one.

 

[FactChecker]

BEWARE of confusing cause and effect. There may be a much higher rate of people with cancer getting CT scans. And the worst cases would be most likely to get the scan, thereby increasing the early death rate. Pay special attention to how the study dealt with this. It is critical before a legitimate conclusion can be drawn.

 

[ZeGato]

BEWARE of confusing cause and effect. There may be a much higher rate of people with cancer getting CT scans. And the worst cases would be most likely to get the scan, thereby increasing the early death rate. Pay special attention to how the study dealt with this. It is critical before a legitimate conclusion can be drawn.

The study deals with it by considering different lag-periods, from 1 to 10 years. In a 10 year lag-period for example, they discard any patient that was diagnosed with any cancer during the 10 years following the CT scan, and measure the incidence rate ratio (IRR) solely for the ones who didn't get cancer during that initial period. Table 3 shows that the IRR went from 1.24 for 1 year lag-period, to 1.18 for the 10 year lag-period, and the null hypothesis that the IRR is the same at every lag-period cannot be rejected at a 5% significance level.

 

[FactChecker]

The study deals with it by considering different lag-periods, from 1 to 10 years. In a 10 year lag-period for example, they discard any patient that was diagnosed with any cancer during the 10 years following the CT scan, and measure the incidence rate ratio (IRR) solely for the ones who didn't get cancer during that initial period. Table 3 shows that the IRR went from 1.24 for 1 year lag-period, to 1.18 for the 10 year lag-period, and the null hypothesis that the IRR is the same at every lag-period cannot be rejected at a 5% significance level.

Yes, but some cancers are very slow to develop and be diagnosed. Do they account for why the CT scan was taken in the first place? There must have been some reason. I say this because those seem like very extreme results and it is hard to believe that there is not a great uproar if it really means what it sounds like.

 

[ZeGato]

Yes, but some cancers are very slow to develop and be diagnosed. Do they account for why the CT scan was taken in the first place? There must have been some reason. I say this because those are very extreme results and it is hard to believe that there is not a great uproar if it really means what it sounds like.

No, they did not - I guess that information wasn't available. It'd be indeed interesting to check the results corrected for the CT scan's reason, but I find it hard to believe that a cancer of an undetectable size could cause symptoms prompting a CT scan...

 

[jim mcnamara]

I read the paper.

The deliberately created lag rate is meant to control for those patients who were CT-scanned as part of cancer diagnosis. Somehow. Or so I think. Why those patients were not identified and then subsequently excluded completely from the study evades me. So does completely understanding the methodology.

Agree with @FactChecker

Caveat:
However, in some kinds of clinical studies there may be unstated assumptions that physicians have that biologists do not have.
I worked with physicians long ago, mostly helped with experimental design and statistical analysis. I encountered confusion in this crossover area. So it means I could be missing something important here as well.

 

[ZeGato]

I read the paper.

The deliberately created lag rate is meant to control for those patients who were CT-scanned as part of cancer diagnosis. Somehow. Or so I think. Why those patients were not identified and then subsequently excluded completely from the study evades me. So does completely understanding the methodology.

Agree with @FactChecker

Caveat:
However, in some kinds of clinical studies there may be unstated assumptions that physicians have that biologists do not have.
I worked with physicians long ago, mostly helped with experimental design and statistical analysis. I encountered confusion in this crossover area. So it means I could be missing something important here as well.

I'm not sure but I suspect that the Australian Medicare records don't store the reason for a CT scan.

Here's what the study's discussion says about the reverse causation issue:

Our study shows that CT scans during childhood and adolescence are followed by an increase in cancer incidence for all cancers combined and for many individual types of cancer. We cannot, however, necessarily assume that all the excess cancers seen during the current period of follow-up were caused by CT scans, because scanning decisions are based on medical indications and are not allocated at random. Thus, we cannot rule out the possibility of reverse causation, whereby symptoms of precancerous conditions (including genetic conditions20 35 36 37) or early symptoms of the cancer itself might themselves prompt a CT scan. Such reverse causation is most likely to be present for brain cancers following a brain CT, where low grade cancers could lead to symptoms prompting investigation several years before the eventual diagnosis of cancer.38 We have, therefore, repeated our main analyses excluding brain cancers occurring after a brain CT; this did not change the overall results substantially (tables 4, 5, and 7; web figure E). For most cancers other than brain cancer, the prediagnostic phase is unlikely to last more than one year in young people, and is very unlikely to last more than 10 years; therefore, reverse causation cannot explain all the cancer excess observed in this study.

Although the inference that CT scans cause most of the excess cancer in exposed people cannot be conclusively proven, it is supported by several observations:

• The increase in IRR with increasing number of CT scans (fig 2), even after excluding brain cancers following brain CTs (web figure E)
• The larger proportional increase in incidence rate (IRR) after exposures at younger ages (table 7), as seen in the Life Span Study of survivors of the Japanese atomic bombings and in other studies of cancer following exposures at larger average doses8 14 17 18 19 20
• The larger absolute excess incidence rate (EIR) for solid cancers in female patients than in male patients (web table D), as seen in other studies20
• The correlation between the site of the CT scan and the site of cancer, with a large proportional increase in the incidence rate (IRR) for leukaemias and myelodysplasias following exposure to red bone marrow from CTs of the abdomen or pelvis (fig 3)
• The increased risks per CT scan and per unit of radiation dose for all cancers (other than brain cancer after a brain scan), even when considering lag periods of five and 10 years (table 8).

 

[ZeGato]

A study on Nagasaki and Hiroshima's bombings survivors also reaches the same conclusion about a constant higher relative risk of cancer after radiation exposure, though no figures are given:

The excess in leukemia mortality has continued to decline with time, but remains slightly but significantly elevated in 1981-1985 in Hiroshima. For cancers other than leukemia, as a group, excess deaths continue to increase over time in direct proportion to the normal increase in natural cancer mortality with increasing age, and the relative risk seems unchanged over time within age ATB cohorts.

 

[Ygggdrasil]

Other research I've seen on the topic give much lower estimates, of around 1 in 1000 increased odds of getting cancer over a lifetime due to a pediatric CT scan.

The BMJ paper you cite in the OP found that "The absolute excess incidence rate for all cancers combined was 9.38 per 100 000 person years at risk."

No, they did not - I guess that information wasn't available. It'd be indeed interesting to check the results corrected for the CT scan's reason, but I find it hard to believe that a cancer of an undetectable size could cause symptoms prompting a CT scan...

CT scans are commonly used for cancer screening, and because those who get screened for cancer are often those who are at higher risk of getting that cancer (e.g. thorugh family history), some population of patients undergoing CT scans started at a higher cancer risk that those not undergoing CT scans.

 

[ZeGato]

The BMJ paper you cite in the OP found that "The absolute excess incidence rate for all cancers combined was 9.38 per 100 000 person years at risk."

CT scans are commonly used for cancer screening, and because those who get screened for cancer are often those who are at higher risk of getting that cancer (e.g. thorugh family history), some population of patients undergoing CT scans started at a higher cancer risk that those not undergoing CT scans.

Both figures (9.38 per 100 000 and 1 per 1000) refer to a low amount of years of follow-up after a pediatric scan, without reaching older ages when cancer is more likely to appear. I was comparing with the figure I gave above of an excess cancer frequency of 1 in 10 over a lifetime, that is valid if the increased incidence remains constant over a lifetime.

CT scans are commonly used for cancer screening, and because those who get screened for cancer are often those who are at higher risk of getting that cancer (e.g. thorugh family history), some population of patients undergoing CT scans started at a higher cancer risk that those not undergoing CT scans.

But then there is the correlation between the site of the CT scan and the site of cancer, with a large proportional increase in the incidence rate (IRR) for leukaemias and myelodysplasias following exposure to red bone marrow from CTs of the abdomen or pelvis (fig 3). And there's also a higher incidence in younger children, suggesting it is the scan's radiation (which has a higher effect in young children, as it was concluded from the nuclear blasts data) that is causing these excess cancers. I also doubt that a CT scan would be done in children just for scanning purposes if no serious symptoms were present.

 

[Ygggdrasil]

Both numbers (9.38 per 100 000 and 1 per 1000) refer to a low amount of years of follow-up after a pediatric scan, without reaching older ages when cancer is more likely to appear. I was comparing with the figure I gave above of an excess cancer frequency of 1 in 10 over a lifetime, that is valid if the increased incidence remains constant over a lifetime.

The figure from the BMJ source (9.38 per 100 000 person years) accounts for the low amount of follow up. The study looked at a population from 0-19, so assuming a scan at age 10, an average life expectancy of 70 years and a constant increased incidence over a lifetime, this would give closer to a one excess cancer case over the lifetime of the individual per 200 scans, not one per 10.

 

[ZeGato]

The figure from the BMJ source (9.38 per 100 000 person years) accounts for the low amount of follow up. The study looked at a population from 0-19, so assuming a scan at age 10, an average life expectancy of 70 years and a constant increased incidence over a lifetime, this would give closer to a one excess cancer case over the lifetime of the individual per 200 scans, not one per 10.

The issue is that the figure is estimated from the number of cancers in young people, to at most early 40s (the oldest members in the cohort). At those ages, the baseline cancer risk is pretty low, and so the excess number of cancers will also be very low. Just as an example: a 20% higher incidence when the natural cancer risk is 1% would give an excess rate of 0.2%, or 1 in 500. However, if the ratio of ~24% increased incidence is maintained for a lifetime, assuming a 41% lifetime cancer risk, that would give about 1 in 10 excess cancers. That's my extrapolation, but it does seem consistent with the study's conclusions that the IRR doesn't change over time, and also with the study I've cited earlier.

The study's discussion says this:

By the end of our current follow-up, the oldest study members were only in their early 40s, when many cancers are still uncommon. Yet the absolute excess incidence rate (EIR) for all solid cancers other than brain cancer after a brain CT increased with time since exposure (table 5; P=0.01 for trend), suggesting that the number of excess cancers among the exposed cohort is likely to rise in future years, increasing the eventual lifetime risk.

 

[Ygggdrasil]

But then there is the correlation between the site of the CT scan and the site of cancer, with a large proportional increase in the incidence rate (IRR) for leukaemias and myelodysplasias following exposure to red bone marrow from CTs of the abdomen or pelvis (fig 3). And there's also a higher incidence in younger children, suggesting it is the scan's radiation (which has a higher effect in young children, as it was concluded from the nuclear blasts data) that is causing these excess cancers. I also doubt that a CT scan would be done in children just for scanning purposes if no serious symptoms were present.

It is clear that CT scans do increase the risk of cancer, so I agree that not all of the excess cases are attributable to "reverse causation." However, because the issue at debate is the exact number of excess cases caused by pediatric CT scans, "reverse causation" still remains a relevant issue. I agree with the authors' assessment that many of the pediatric brain CT scans could be related to symptoms related to precancerous growths, so some of the increased risk they report is attributable to "reverse causation."

The issue is that the figure is estimated from the number of cancers in young people, to at most early 40s (the oldest members in the cohort). At those ages, the baseline cancer risk is pretty low, and so the excess number of cancers will also be very low. Just as an example: a 20% higher incidence when the natural cancer risk is 1% would give an excess rate of 0.2%, or 1 in 500. However, if the ratio of ~24% increased incidence is maintained for a lifetime, assuming a 41% lifetime cancer risk, that would give about 1 in 10 excess cancers. That's my extrapolation, but it does seem consistent with the study's conclusions that the IRR doesn't change over time, and also with the study I've cited earlier.

I would advise some caution in interpreting your estimates. First, as noted above, the overall ~24% figure is likely inflated given the issue with "reverse causation" that the authors note about the brain scan data. Second, you miss an important part of the Hiroshima/Nagasaki study: the conclusion that the relative risk remains unchanged applies to cancers other than leukemia. The 24% increased risk figure also includes many leukemias (~20% of the excess cases), so this fact may also elevate your estimate.

Finally, doctors are aware that CT scans do increase an individual's lifetime risk of cancer, especially in children, and most will act accordingly. As with many medical decisions, one must weigh the risks and benefits of a particular medical intervention. There are many cases where the benefits of a CT scan would outweigh the increased lifetime risk of cancer, and there will be other cases where the benefit is not so clear, so declining a CT scan or opting for a different imaging method would be warranted. Of course, alternative imaging methods like MRI can be more expensive, so there are additional costs and benefits to consider for that choice as well.

 

[ZeGato]

It is clear that CT scans do increase the risk of cancer, so I agree that not all of the excess cases are attributable to "reverse causation." However, because the issue at debate is the exact number of excess cases caused by pediatric CT scans, "reverse causation" still remains a relevant issue. I agree with the authors' assessment that many of the pediatric brain CT scans could be related to symptoms related to precancerous growths, so some of the increased risk they report is attributable to "reverse causation."
I would advise some caution in interpreting your estimates. First, as noted above, the overall ~24% figure is likely inflated given the issue with "reverse causation" that the authors note about the brain scan data. Second, you miss an important part of the Hiroshima/Nagasaki study: the conclusion that the relative risk remains unchanged applies to cancers other than leukemia. The 24% increased risk figure also includes many leukemias (~20% of the excess cases), so this fact may also elevate your estimate.

Finally, doctors are aware that CT scans do increase an individual's lifetime risk of cancer, especially in children, and most will act accordingly. As with many medical decisions, one must weigh the risks and benefits of a particular medical intervention. There are many cases where the benefits of a CT scan would outweigh the increased lifetime risk of cancer, and there will be other cases where the benefit is not so clear, so declining a CT scan or opting for a different imaging method would be warranted. Of course, alternative imaging methods like MRI can be more expensive, so there are additional costs and benefits to consider for that choice as well.

My concern is that, at most, they have the "1 in 1000" figure in their mind, but that seems to be a serious underestimation of the real risk by 1 or 2 orders of magnitude. The BEIR model derived from the nuclear blastings also seems to be severely flawed in measuring the risk in these type of scans, yet it's widely used in studies. If the real risk is between 1 in 100 to 1 in 10, a CT scan is effectively a cancer-causing procedure and there is no justification whatsoever to do it at the current doses, in children or adults.

 

[Laroxe]

I think this should come under the joys of statistics, a few thoughts;
The study population were all between 0 -19 years old, there were 10.9 million records.
In this group, over the study period 60,674 cases of cancer were identified during the study period, this is an incidence of around 0.56%
In the 10.9 million they identified 680221 individuals that had received a CT scan and had not received a diagnosis within 12 months of this scan, this was to reduce the chance that that the scan was part of the diagnosis of cancer.
Now, of the 60,674 cases of cancer they identified 3150 who had received a CT scan at least 1 year before the diagnosis now in matching these cases with the overall number they identified that the number was higher than expected and that's where you get the percentage figures from. In terms of risk the numbers are miniscule. The population level risk of 0.56% and the summary doesn't give the breakdown numbers but of the 3150 cases identified 76% of these would have developed cancer anyway, this leaves us with around 600 cases cases more than expected out of a population of 10.9 million.
We've known about the risks from X rays for decades and C.T. scans use X rays, but sometimes they are the best investigation to use, the different types of scans are not directly inter-changable. Talking about levels of risk in the 1-100 is light years out, they would have killed millions.

 

[ZeGato]

I think this should come under the joys of statistics, a few thoughts;
The study population were all between 0 -19 years old, there were 10.9 million records.
In this group, over the study period 60,674 cases of cancer were identified during the study period, this is an incidence of around 0.56%
In the 10.9 million they identified 680221 individuals that had received a CT scan and had not received a diagnosis within 12 months of this scan, this was to reduce the chance that that the scan was part of the diagnosis of cancer.
Now, of the 60,674 cases of cancer they identified 3150 who had received a CT scan at least 1 year before the diagnosis now in matching these cases with the overall number they identified that the number was higher than expected and that's where you get the percentage figures from. In terms of risk the numbers are miniscule. The population level risk of 0.56% and the summary doesn't give the breakdown numbers but of the 3150 cases identified 76% of these would have developed cancer anyway, this leaves us with around 600 cases cases more than expected out of a population of 10.9 million.
We've known about the risks from X rays for decades and C.T. scans use X rays, but sometimes they are the best investigation to use, the different types of scans are not directly inter-changable. Talking about levels of risk in the 1-100 is light years out, they would have killed millions.

The lifetime risk would be small only if assuming that the 24% higher incidence goes to 0 after the first 10 years (average follow-up period in this study, and lower in similar studies). The problem is that the studies from the nuclear bombings show that the relative risk stays higher and constant for life, even for the radiation doses that are used on CTs.

As you said, 24% higher incidence of cancer over a lifetime would indeed result in a very high number of cancer cases, but they could easily be overlooked for many reasons:

• Only a small portion of the population goes through CT or PET scans. Even if there was an increase in the general population's cancer cases, there are many other factors, such as older population and increasing urbanization. It wouldn't be noticeable enough increase to pinpoint the main cause as increased radiation exposure. That doesn't mean however, that such scan couldn't be bringing a very high risk to the person undergoing one.
• Many of the patients who go through multiple CT scans are cancer patients, and those can't be used to evaluate the effects of radiation, since it's a population with an already high, and hard-to-estimate risk of cancer reincidence.
• There's a long lag between radiation exposure and cancer. Radiation-induced cancers may only show up a decade or longer after the exposure, making it impossible for physicians and other medical staff to make the connection between the two.
• The model to estimate cancer risk that is currently in use, is used for decades. There are many cases in Science where wrong models are used for a long time just because it's more comfortable to accept it than to question it; the fact is that the empirical studies so far contradict it, suggesting that either the effective radiation dose is being underestimated, the gamma and neutrons radiation from the nuclear blasts has a milder effect on the human body than X-rays for the same radiation amount, or a concentrated dose could be more hazardous than a scattered one.
• There are no empirical studies with a follow-up that reaches 40-60 years of age

The only way to settle this would be to do an empirical study of CT scans with a follow-up that reaches the ages of 50-60 years, when the natural cancer rate begins to rise substantially, having the reason for the CT scan into account (the study could be limited to CT scans due to physical trauma only). If there is indeed a 10-20% higher cancer incidence after a CT scan, we'd see a high number of excess cancers in such study. I'm sure that there is this data in many countries, and I have no idea why such study wasn't done so far. The results from empirical studies on pediatric populations, the long-term studies on the nuclear bombings and the enormous potential risk from the widespread use of CT and PET scans are surely asking for a more serious approach to this issue.

 

[Laroxe]

Peoples beliefs about the effects of radiation exposure tend to over estimate the risks, in fact the health effects of radiation exposure following the attacks on Japan were significantly less than predicted and we see a thriving ecosystem around Chernobyl.
At the time of and immediately following the nuclear explosions (over a few months) there were of course a massive number of casualties but we are interested in the long term effects of a low dose single exposure to radiation. We know it acts as a mutagen, which increases the chances of mutations and the great majority of cancers are associated with multiple cellular mutations. There are problems in using data from the attacks as a basis for predicting risk from X-ray exposure, the nuclear attack generated significant amounts of biologically active radioactive isotopes, which if absorbed would lead to prolonged irradiation of some tissues.

The risk was far lower than predicted with significant variation based on the age and sex of people exposed, the risk is not constant. They found that in the period 2 – 6 years following the attack children experienced the greatest risk, particularly in relation to leukaemia. This had an attributable risk—the percent difference in the incidence rate of a condition between an exposed population and a comparable unexposed one of 46%. The risk was highest in the very young and was dose related. The risks to adults occurred later with an increased risk of around 10%. Its worth noting that these numbers convert to an average reduction of life expectancy of only a few months.

Its effectively impossible to get a clear causal relationship between x-ray exposure and cancer, at every cell reproduction there is a risk of mutation and mutations accumulate with age. Cancer requires a number of very specific mutation to occur together, its like a lottery. We are all exposed to a number of different mutagens that can effect the overall rate of mutations and radiation is always present though at varying levels. Most will come from that large nuclear explosion in the sky and many naturally occurring rocks emit radiation.


Diagnostic Procedure; Typical Effective Dose (mSv)1

Chest x-ray (PA film) 0.02
Lumbar spine 1.5
I.V. urogram 3
Upper G.I. exam 6
Barium enema 8
CT head 2
CT chest 7
CT abdomen 8
Coronary artery calcification CT 3
Coronary CT angiogram 16

A CT examination with an effective dose of 10 millisieverts may be associated with an increase in the possibility of fatal cancer of approximately 1 chance in 2000. This increase in the possibility of a fatal cancer from radiation can be compared to the natural incidence in the U.S. population, about 1 chance in 5 (equal to 400 chances in 2000). So the risk of radiation-induced cancer is much smaller than the natural risk of cancer. If you combine the natural risk and the estimated risk from a 10 mSv CT scan, the total risk may increase from 400 chances in 2000 to 401 chances in 2000.
source; https://www.fda.gov/radiation-emitt...es/medicalimaging/medicalx-rays/ucm115329.htm

This gives some numbers from the atomic bombs
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3875218/
and this discusses the risk
https://www.sciencedaily.com/releases/2016/08/160811120353.htm

 

[ZeGato]

The risk was far lower than predicted with significant variation based on the age and sex of people exposed, the risk is not constant. They found that in the period 2 – 6 years following the attack children experienced the greatest risk, particularly in relation to leukaemia. This had an attributable risk—the percent difference in the incidence rate of a condition between an exposed population and a comparable unexposed one of 46%. The risk was highest in the very young and was dose related. The risks to adults occurred later with an increased risk of around 10%. Its worth noting that these numbers convert to an average reduction of life expectancy of only a few months.

Yes, the risk changes based on age and sex, but a study on the mortality of A-bomb survivors show that the increased relative risk of cancer is constant throughout life.

A CT examination with an effective dose of 10 millisieverts may be associated with an increase in the possibility of fatal cancer of approximately 1 chance in 2000. This increase in the possibility of a fatal cancer from radiation can be compared to the natural incidence in the U.S. population, about 1 chance in 5 (equal to 400 chances in 2000). So the risk of radiation-induced cancer is much smaller than the natural risk of cancer. If you combine the natural risk and the estimated risk from a 10 mSv CT scan, the total risk may increase from 400 chances in 2000 to 401 chances in 2000.
source; https://www.fda.gov/radiation-emitt...es/medicalimaging/medicalx-rays/ucm115329.htm

This gives some numbers from the atomic bombs
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3875218/
and this discusses the risk
https://www.sciencedaily.com/releases/2016/08/160811120353.htm

This empirical study shows that the LNT model, that is used to estimate those odds, underestimated the all cancer risk by a factor of almost 10 (Table 9). The ERR of cancer per mSv of effective dose was 0.027, while the LNT model would estimate it to be 0.003. This could've happened because the effective radiation dose is being underestimated, the gamma and neutrons radiation from the nuclear blasts has a milder effect on the human body than X-rays for the same radiation amount, or a concentrated dose could be more hazardous than a scattered one. This is unfortunately the only empirical study that studies the incidence of all cancers after CT scans, so this result wasn't reproduced anywhere else so far. But given the widespread use of CT scans, and given that there is empirical data available, there is no reason to keep using rough estimates.

 

[ZeGato]

There are other empirical studies available after all: this one was done on patients from Quebec who underwent cardiac medical imaging, with a median age of 63 years and a mean follow-up of 5 years. The hazard ratio of all cancer incidence was estimated to be 1.03 per 10mSv (95% confidence interval 1.02–1.04), or a 3% increased cancer risk per 10mSv, at an average age of 63 years. This is significantly lower compared to the Australian study, probably because children are known to be more radiation-sensitive, and cancer incidence due to radiation seems to decrease with age, as it was observed in the Japanese Life-Span Study.

Comparison with Japanese Life-Span Study:

The Japanese Life-Span Study involving atomic bomb survivors showed a relative risk of 1.1–1.2 for solid cancers for an average dose of 200 mSv3 (HR 1.751 in extrapolation from our model)

 

[Nik_2213]

Having sundry family members who required CT scans, may I mention several points ?

First, IIRC, there have been historical issues with eg CT software, such that some people received higher than necessary dosages.

Second, the technology improves. IIRC, resolution is now much, much better for the same dosage levels, reduced dosages often provide adequate resolution.

Third, many people requiring CT scans who stay alive long enough to be at risk of slow-onset, perhaps radiological-induced cancers would probably have died without such scans. Like Chemo-therapy and Radio-therapy, or even the humble aspirin, all medical intervention comes with some risk.

Sadly, so does crossing the road. Having once been very-near-missed by a speeding truck that went the wrong side of a traffic island, whites-of-driver's-eyes close, even 'Due Care' needs a seasoning of luck...

( OT: I must point out that pseudo-scientific intervention / non-intervention also comes with risk, which often extends to community... )

Fourth, I must wonder if the Australian data was adequately corrected for cosmic ray exposure on long-haul flights...
http://spaceweather.com/

 

[gleem]

There is no way to prove that radiation causes cancers using this type of study. If they had selected a group of volunteers given a CT who did not need a medically indicated CT as a control they would have something but that is probably not possible.. In any statistical study there are always tacit assumptions about the populations about the method. They massage the data, making correction they deem necessary and then apply their statistical wizardry to produce a number.

It is well know that radiation (especially at high dose rates) can cause cancer. It is also well know that children are more sensitive to the effects of radiation exposure than adults. What is not well known is the quantitative dependence of the number of cancers produced per Sievert. It is also well known that the rate of cancer incidence increases over time after exposure something that is not seen in this study. In the comparison of this study with a previous one for colon cancer the present study show a large increase, but states without further explanation that the comparison is problematic.

As for the exhortation of telling patients and parents to assure the exam is necessary well how many will challenge the recommendation of the physician. It has been know from the introduction of CTs (mid to late '70s) that they use a substantial amount of radiation. Physicians have been continually warned about using CTs unnecessarily e.g. general screening or headaches. Radiologists have been supposedly taught to use the lowest radiation exposure consistent with the objectives of the exam or recommend different modalities when appropriate. Medical physicist are taught to assure the radiation doses for the equipment are as low as reasonably achievable. Manufacturers try to improve efficient use of radiation but you cannot throw out good equipment even if it uses more radiation. All that said in the general medical community changes occur slowly. And then there is the "cover your a..." philosophy at least in the US. It is well known that way more diagnostic/lab tests are performed than necessary.

I would like to see a similar study here in the US to corroborate their results. In the meantime if your physician recommends a CT ask if it is necessary or if another type of modality might do as well (assuming the insurance company agrees). As pointed out every medical procedure or pharmaceutical carries some risk. It has been estimated that medical errors are the third leading cause of death in the US.