A few years ago (2009) Dr. David Brenner from Columbia coauthored a paper that claimed 14,500 prospective cancer deaths would be caused by one year's worth of CT scans. So in the past 20 years, can one assume that the average number of CT scans per year is roughly the same as in Brenner's paper (i.e., the increase in CTs/yr since Brenner's paper would average out with the lower number of CT's/yr 11 years prior to his paper)? And if so, can we assume based on Brenner's calculation that CT has caused, since 1998, 14,500 x 20 = 290,000 actual and prospective cancer deaths? Wouldn't these cancer deaths somehow influence cancer statistics?
The discussion about the increase cancer potential started well before that publication, and so did the response. The number of exams ordered decreased, the amount of exposure for the studies was decreased and programs like the ACR‘s “image gently” changed the landscape.
There will always be debate about the calculation, but there was an immediate change in behavior that will complicate future analysis of actual affect to the earlier calculations.
This paper may help
Chapter Breast Cancer Detection and Classification Using Thermograph...
I guess there will always be debate about that as long as not all the CT-exams are justified. Nobody thinks about the cancer risk increased by a CT scan if that particular scan really is necessary for diagnosing the patient.
You're right, Albertina. But that reduces the decision to a simple binary one: "If you need a CT scan, have it, and if you don't need one, don't have it!" My concern is the first of these choices; i.e., "How MUCH do you need to have a CT scan so that the risk--if there is in fact a risk--makes it worthwhile." These are very complicated issues to figure out since neither benefit nor risk are clearly defined on a statistical basis. Even using the most extensive human radiation effects data we have at the present time, there are very large statistical uncertainties associated with dose risk calculations in the diagnostic range—often larger than +100% (1 SD); i.e., using the LNT hypothesis, the roughly 0.05% probability of dying of cancer due to a single CT scan (producing an assumed effective-dose of 10 mSv), when expressed as a +/-1 SD statistical range, would be zero to 0.1%. I don't think one can make an informed decision of radiation risk with such a large degree of uncertainty! But I come back to my original question: Can we assume based on Brenner's calculation that CT has caused, since 1998, very approximately 290,000 actual and prospective cancer deaths. And if so, wouldn't these cancer deaths somehow impact cancer statistics?
There is no direct epidemiological evidence showing that CT exposures increase the risk of cancer and other diseases. Many epidemiological studies on CT and other diagnostic exposures have indeed reported the increased cancer risk, but these findings are subject to uncertainties in dose (e.g., lack of individual doses) and potential biases (e.g., confounding by indication and reverse causation). These make direct risk estimation of diagnostic exposures unreasonable. The future studies with more relevant information may improve the situation.
The following three recent papers have reported a statistically significant linear dose response at 0-100 mSv for caner risk.
https://www.ncbi.nlm.nih.gov/pubmed/28323979
https://www.ncbi.nlm.nih.gov/pubmed/28319463
https://www.ncbi.nlm.nih.gov/pubmed/26487649
In 2009, the US National Council on Radiation Protection and Measurements published Report No. 160 “Ionizing Radiation Exposure of the Population of the United States”. Report No. 160 assessed non-occupational exposures until 2006, and documented that the average growth rate of the number CT procedures in 1993-2006 is about 10% per year. NCRP Scientific Committee 4-9 “Medical Exposure of Patients in the United States” is preparing a report that provides an update on medical exposures since 2006 until 2016, which will show that the number of CT procedures increased from 2006 to 2012ish, but has leveled off since then. Not only CT but also medical exposures as a whole has also leveled off.
According to the ICRP collective effective dose is not intended as a tool for epidemiological analysis and the prediction of health effects in populations. The analysis by Brenner is strictly collective dose.
Collective dose is based on the LNT in the extreme. There has been no validation of the LNT. Indeed, the orignial basis for the LNT by the BEIR committees was demonstrated to be false.
@ Nobuyuki has clearly stated that there is no epidemiological basis for CT cancer risk. Yet, in a following answer he claims three recent papers have reported a statistically significant linear dose response for cancer at 0-100 mSv. Reading of the papers does not support linearity. The papers report non-linear and threshold effects. More important, the reported results are for modeling data, not cancer data. Not one of the reports shows statistically significant cancers at any of the reported doses. The error in the data preclude claiming quantifiable results.
- Badges
- Science topic