CT Radiation in Kids: How Much of a Risk, Really?
CT Radiation in Kids: How Much of a Risk, Really?
Medscape: Can you describe the risk-benefit calculation of CT imaging in a child, as you see it, in the current environment of much lower doses of ionizing radiation?
Dr. Brody: For me, the problem is that the extensive coverage of possible risks from diagnostic imaging --- CT, but also regular radiographs, such as chest radiography, fluoroscopy, and others -- only presents the possible danger, but not the likely benefit. Just the word "radiation" scares people. Physicians and families concentrate on the risk, but not on the benefit. Our article was written to point out that the potential risk is tiny, and the potential benefit usually large.
At our institution, we have seen parents agonizing over whether they should allow their child to have a CT scan, even when the benefit is clear. If there is any risk for a downstream cancer as a result of that image, it is about 1 in 4000. That means that the CT will not lead to later cancer 99.975% of the time. Neither the family nor the doctor should hesitate for a second to do a CT scan that has even a 1 in 100 chance of saving the child harm. I doubt that anyone would suggest a CT scan if it didn't have much more than a 1% chance of making things better for the child.
What about a child with possible appendicitis? Let's say we have done ultrasonography, and we're not sure whether the child has appendicitis. We think the CT has a 25% chance of telling us that the child has appendicitis. What is the risk of not doing the CT?
If we go to surgery right away, we have a 75% chance of doing an unnecessary surgery. If we choose instead to monitor the child, we have the risk for the appendix rupturing and causing a widespread abdominal infection that will require a week or more in the hospital, and the future risk for multiple hospitalizations for adhesions and bowel obstruction.
The CT is 99.975% safe. When the benefits are this clear, no one should be worrying about a 1 in 4000 chance that cancer might occur, especially when the data suggest that if a cancer does occur, it will occur long after the child has become an adult -- probably at 60 or more years of age. We also have to remember that the lifetime risk that a child today will die of cancer without that child ever having CT is 1 in 5.
Maybe even more concerning is that radiologists are doing lower-quality CT scans in order to reduce the dose. Reducing unnecessary dose is, of course, the right thing to do. The hard thing to do is to figure out how much dose is necessary.
In a very concerning study, 6 pediatric radiology departments got together to see whether they could decide how much radiation should be used for abdominal CT in children. Each department contributed CT scans that had been done on patients in their departments who needed to have CT.
The group looked at the quality of the CT scans that were sent in and decided that 6% of them were of such poor quality that they were judged nondiagnostic and could not be used to adequately evaluate for the presence of disease. This means that any child who had CT at one of these hospitals had a 1 in 20 chance that the scan would be of unacceptably low quality.
These were all dedicated pediatric radiology departments, with radiologists who were particularly interested in CT scanning and radiation dose. If the best sites are using too low a dose, how many smaller hospitals are doing CT scans that don't show what they could if only a small additional amount of radiation had been used?
On the basis of the estimate of a 1 in 4000 risk for cancer, cutting the dose in half would reduce the risk to 1 in 8000. To do an inadequate CT scan to decrease risk by this almost infinitesimal amount doesn't make sense. Reducing the dose too far has a large risk for harm with almost no benefit.
Another situation where clinicians may overestimate risk is when a child who has previously had 1 or more CT scans needs another CT scan -- for the same clinical situation, or a different concern. As we've discussed, the risk of obtaining that second CT is extremely small. What's more, the risk from that next CT scan doesn't change whether the child has had 1 or 100 prior CT scans.
Consider this corollary: We know that riding in a car carries the risk of dying in a car accident. The more miles ridden, the higher the risk. But that doesn't mean that the risk is greater the next time you get in a car. No one would say, "My son has ridden many thousands of miles this year; I don't think he should go on the next family trip." Not doing a necessary CT because one has been done before is the same thing. Somehow, though, factoring in that previous imaging, and allowing that to affect our decisions now, does seem to happen with CT.
Medscape: Are there specific conditions for which MRI or ultrasonography is a particularly poor substitute?
Dr. Brody: CT is the best study for acute head trauma -- better than MRI. MRI is nowhere near as good as CT for the lungs, although it is getting better. It would not be an alternative to CT for any of the diseases that are difficult to evaluate on chest radiography and on CT.
I have often heard clinicians say that they would like to order a CT scan because they think the information would be helpful, but they are questioning that decision because of concerns about the radiation. That is a way that you could really hurt your patient. Because if you think CT is going to help you and you're a good clinician, then it probably is going to help you, even if it's negative. You're going to say, "Okay, I was worried about all of these things on my differential. Now, I don't have to worry about one third of them. I can cross them off my list because I would have seen them on the CT. And now, my team can concentrate on the things that remain as likely possibilities."
That is the process that we should have. We should compare the risk of CT with other risks, such as hospitalization risks. We all know that there is a risk for harm from a hospitalization and potential in-hospital error. We need to be certain that families understand that the chance of harm from the CT scan is a lot lower than the chance of harm resulting from being in the hospital for a week.
Medscape: Although you've noted the risks of not obtaining a needed CT, you've also stressed the importance of assuring that the CT is indeed needed. In that context, are there specific conditions that merit particular concern about potential overutilization of ionizing radiation?
Dr. Brody: One of the big situations where it is indeed appropriate to save radiation is when you really don't need a CT scan. There are low-risk situations where we don't have to use a CT scan or other diagnostic imaging tests.
An example of that is in the situation of head trauma. Failure to use available evidence-based assessment algorithms for patient evaluation that include good decision rules for when to obtain a CT after head trauma is a situation that presents a risk for overimaging.
Medscape: Any concluding thoughts for our audience?
Dr. Brody: I would urge clinicians to remember the benefit side of the risk-benefit equation and be certain to share that information with families. You know, we say that if a million children get CT scans, 100 will have a risk of getting cancer. But we don't say that if a million kids receive CT scans, one half of them will avoid unnecessary surgery, 100,000 of them will receive surgery that is better because the surgeon is guided by the CT results, and 300,000 of them won't have to go into the hospital unnecessarily.
It's analogous to the arguments for vaccines. The benefit is theoretical because we did give the patients their immunizations, so they didn't get sick. So we can't point to those children. But of course, we can point to the dramatically smaller number of children who did have some kind of harm associated with that immunization.
These are really difficult situations. One of the things that I think is really important is that, unfortunately, your risk and my risk of dying of cancer is 1 in 5. We're not talking about something that almost never happens, and that might happen because of a diagnostic imaging study. This discussion occurs in the context of a reality that cancer is very common. And reducing risk for future cancers is an important goal for all of healthcare.
Calculating Risk vs Benefit
Medscape: Can you describe the risk-benefit calculation of CT imaging in a child, as you see it, in the current environment of much lower doses of ionizing radiation?
Dr. Brody: For me, the problem is that the extensive coverage of possible risks from diagnostic imaging --- CT, but also regular radiographs, such as chest radiography, fluoroscopy, and others -- only presents the possible danger, but not the likely benefit. Just the word "radiation" scares people. Physicians and families concentrate on the risk, but not on the benefit. Our article was written to point out that the potential risk is tiny, and the potential benefit usually large.
At our institution, we have seen parents agonizing over whether they should allow their child to have a CT scan, even when the benefit is clear. If there is any risk for a downstream cancer as a result of that image, it is about 1 in 4000. That means that the CT will not lead to later cancer 99.975% of the time. Neither the family nor the doctor should hesitate for a second to do a CT scan that has even a 1 in 100 chance of saving the child harm. I doubt that anyone would suggest a CT scan if it didn't have much more than a 1% chance of making things better for the child.
What about a child with possible appendicitis? Let's say we have done ultrasonography, and we're not sure whether the child has appendicitis. We think the CT has a 25% chance of telling us that the child has appendicitis. What is the risk of not doing the CT?
If we go to surgery right away, we have a 75% chance of doing an unnecessary surgery. If we choose instead to monitor the child, we have the risk for the appendix rupturing and causing a widespread abdominal infection that will require a week or more in the hospital, and the future risk for multiple hospitalizations for adhesions and bowel obstruction.
The CT is 99.975% safe. When the benefits are this clear, no one should be worrying about a 1 in 4000 chance that cancer might occur, especially when the data suggest that if a cancer does occur, it will occur long after the child has become an adult -- probably at 60 or more years of age. We also have to remember that the lifetime risk that a child today will die of cancer without that child ever having CT is 1 in 5.
Maybe even more concerning is that radiologists are doing lower-quality CT scans in order to reduce the dose. Reducing unnecessary dose is, of course, the right thing to do. The hard thing to do is to figure out how much dose is necessary.
In a very concerning study, 6 pediatric radiology departments got together to see whether they could decide how much radiation should be used for abdominal CT in children. Each department contributed CT scans that had been done on patients in their departments who needed to have CT.
The group looked at the quality of the CT scans that were sent in and decided that 6% of them were of such poor quality that they were judged nondiagnostic and could not be used to adequately evaluate for the presence of disease. This means that any child who had CT at one of these hospitals had a 1 in 20 chance that the scan would be of unacceptably low quality.
These were all dedicated pediatric radiology departments, with radiologists who were particularly interested in CT scanning and radiation dose. If the best sites are using too low a dose, how many smaller hospitals are doing CT scans that don't show what they could if only a small additional amount of radiation had been used?
On the basis of the estimate of a 1 in 4000 risk for cancer, cutting the dose in half would reduce the risk to 1 in 8000. To do an inadequate CT scan to decrease risk by this almost infinitesimal amount doesn't make sense. Reducing the dose too far has a large risk for harm with almost no benefit.
Another situation where clinicians may overestimate risk is when a child who has previously had 1 or more CT scans needs another CT scan -- for the same clinical situation, or a different concern. As we've discussed, the risk of obtaining that second CT is extremely small. What's more, the risk from that next CT scan doesn't change whether the child has had 1 or 100 prior CT scans.
Consider this corollary: We know that riding in a car carries the risk of dying in a car accident. The more miles ridden, the higher the risk. But that doesn't mean that the risk is greater the next time you get in a car. No one would say, "My son has ridden many thousands of miles this year; I don't think he should go on the next family trip." Not doing a necessary CT because one has been done before is the same thing. Somehow, though, factoring in that previous imaging, and allowing that to affect our decisions now, does seem to happen with CT.
Medscape: Are there specific conditions for which MRI or ultrasonography is a particularly poor substitute?
Dr. Brody: CT is the best study for acute head trauma -- better than MRI. MRI is nowhere near as good as CT for the lungs, although it is getting better. It would not be an alternative to CT for any of the diseases that are difficult to evaluate on chest radiography and on CT.
I have often heard clinicians say that they would like to order a CT scan because they think the information would be helpful, but they are questioning that decision because of concerns about the radiation. That is a way that you could really hurt your patient. Because if you think CT is going to help you and you're a good clinician, then it probably is going to help you, even if it's negative. You're going to say, "Okay, I was worried about all of these things on my differential. Now, I don't have to worry about one third of them. I can cross them off my list because I would have seen them on the CT. And now, my team can concentrate on the things that remain as likely possibilities."
That is the process that we should have. We should compare the risk of CT with other risks, such as hospitalization risks. We all know that there is a risk for harm from a hospitalization and potential in-hospital error. We need to be certain that families understand that the chance of harm from the CT scan is a lot lower than the chance of harm resulting from being in the hospital for a week.
Medscape: Although you've noted the risks of not obtaining a needed CT, you've also stressed the importance of assuring that the CT is indeed needed. In that context, are there specific conditions that merit particular concern about potential overutilization of ionizing radiation?
Dr. Brody: One of the big situations where it is indeed appropriate to save radiation is when you really don't need a CT scan. There are low-risk situations where we don't have to use a CT scan or other diagnostic imaging tests.
An example of that is in the situation of head trauma. Failure to use available evidence-based assessment algorithms for patient evaluation that include good decision rules for when to obtain a CT after head trauma is a situation that presents a risk for overimaging.
Medscape: Any concluding thoughts for our audience?
Dr. Brody: I would urge clinicians to remember the benefit side of the risk-benefit equation and be certain to share that information with families. You know, we say that if a million children get CT scans, 100 will have a risk of getting cancer. But we don't say that if a million kids receive CT scans, one half of them will avoid unnecessary surgery, 100,000 of them will receive surgery that is better because the surgeon is guided by the CT results, and 300,000 of them won't have to go into the hospital unnecessarily.
It's analogous to the arguments for vaccines. The benefit is theoretical because we did give the patients their immunizations, so they didn't get sick. So we can't point to those children. But of course, we can point to the dramatically smaller number of children who did have some kind of harm associated with that immunization.
These are really difficult situations. One of the things that I think is really important is that, unfortunately, your risk and my risk of dying of cancer is 1 in 5. We're not talking about something that almost never happens, and that might happen because of a diagnostic imaging study. This discussion occurs in the context of a reality that cancer is very common. And reducing risk for future cancers is an important goal for all of healthcare.
Source...