Computed tomography (CT) is a powerful diagnostic tool that allows rapid diagnosis of disease.  CT is widely available in the U.S. and is a mainstay of medical diagnosis.  Estimates state that 85 million CT scans were performed in the U.S. in 2012.  To create images, CT scanners pass ionizing radiation (x-rays) through the body thereby exposing patients to radiation.  Patients who are imaged with CT have a theoretical but widely accepted risk of developing cancer years to decades following radiation exposure.

In contrast, MRI and ultrasound create images without x-rays and have no risk of inducing cancer.  MRI particularly provides equivalent or superior diagnostic ability compared to CT in many clinical scenarios.  Substituting these non-radiation imaging techniques for CT will result in less radiation-induced cancers, especially for children and young adults. While avoiding cancer is important to patients, circumventing the expense of treating these cancers is also beneficial to society.  The financial costs of treating CT-induced cancer have not been determined.  The purpose of this article is to estimate and discuss these costs.

Calculating the costs:

First, we must select an estimate for the rate of cancer induction from CT.  Scientific committees like the National Academy of Sciences’ Committee on the Biological Effects of Ionizing Radiation (BEIR) have published sophisticated estimates of these risks.  The most recent BEIR risk estimates (BEIR VII) are likely the most widely cited and indicate an average risk of radiation-induced cancer incidence at 1.1 percent per 100 mSv of radiation exposure.  For an average CT scan of the abdomen at 5 mSv these BEIR VII estimates correspond to a risk of cancer induction per CT scan of 5.5 in 10,000 (0.055 percent).  Although other, similar estimates exist for the risk of developing cancer after radiation exposure, this estimate will be used in our calculations.

Secondly, the cost of cancer treatment in the U.S. must be estimated.   Data from the Medical Expenditure Panel Survey (MEPS) from the Agency for Healthcare Research and Quality and the National Center for Health Statistics contain nationally representative data on health care payment from government and private insurance for adults over 18 years old.  MEPS data for 2012, the most recent year on record, estimate total direct annual cancer treatment costs for U.S. adults at 87.5 billion U.S. dollars (USD) with a cancer prevalence of approximately 15.5 million adult patients.  Total costs divided by the number of patients corresponds to per patient expenditures of 5,631 USD per year.

Lastly, the aforementioned data must be combined. Multiplying a .055 percent lifetime cancer risk by the 85 million CT scans performed in 2012 equals 47,000 additional cancers induced by CT for that year.  To calculate total direct cancer treatment costs from CT scans in 2012, we multiplied 5,631 USD annually per cancer patient by 47,000 induced cancers. This calculation provides treatment cost estimates in terms of 2012 USD of 260 million annually, 1.3 billion over 5 years and 2.6 billion over a decade.  Estimated annual direct cancer treatment costs for every 1 mSv radiation exposure to the population scanned are estimated at 53 million USD.  Three dollars per CT scan may be spent on future cancer treatment in terms of 2012 USD.

What does this mean?

Each day CT scans save many lives from early and accurate diagnosis of disease.  We emphasize that benefits of CT for appropriately selected patients far exceed small risks of radiation-induced cancer.  Nevertheless, our analysis suggests that population exposure to CT radiation may have downstream costs on the order of hundreds of millions of dollars annually.

The occult cost of radiation from CT can be framed in economic terms as a negative externality. A negative externality is a negative consequence of a good or service that should be reflected in the original cost (in this case CT) to reflect its true value.  For example, the downstream costs of air pollution could be added into the purchase price of a car to accurately represent its true cost.  Similar to cars, CT provides great benefit to society but has downstream costs that are often overlooked.

To illustrate, new CT scanners can reduce patient doses by 50 percent or more compared to older equipment.  However, facilities may profit more from operating older higher-dose CT rather than lower-dose, more expensive scanners as both reimburse the same.  Conversely, no economic incentives reward imaging centers with protocols that provide the lowest possible patient exposure.  Therefore, current policy promotes or propagates rather than corrects for the negative externality of radiation-induced cancer costs.  To address this issue the Protecting Access to Medicare Act of 2014 will reduce payments beginning in 2016 for scans performed on CT equipment with doses exceeding the MITA Smart Dose CT standards.

Abdominopelvic MRI may increasingly substitute for CT as first-line exams for patients with most non-traumatic abdominal pain.  Using our estimates, if an MRI scanner substituted for 5,000 CT scans annually at 5 mSv per scan, cost savings from radiation-induced cancers over a decade would be 145,000 USD 2012 for that scanner.  Although MRI remains more expensive than CT to purchase and operate, these savings may offset a portion of the additional cost of MRI.

A criticism may be that CT-induced cancer costs may occur decades from now but are non-existent at present because cancer requires several years to decades to develop after radiation exposure.  We disagree.  For years, the problem of radiation-induced cancer from medical imaging has been framed as a future event.  Hundreds of millions of CT scans have been performed in the U.S. since the 1990s, thus, we have already been paying for those excess cancers.

Our model is derived from single estimates of radiation risk, cancer treatment costs and number of CT scans performed. More robust analysis could reach slightly different results.  Regardless, the message is that the negative externality of CT-induced cancer should be considered when valuating CT and alternative imaging modalities.

Matthew F. Covington and Phillip H. Kuo are radiologists.

Image credit: Shutterstock.com