Minimizing radiation risks in CT imaging: Medical-legal considerations

By R. James Brenner, MD, JD, FACR, FCLM
pdf path

Dr. Brenner is Director, Breast Imaging, Bay Imaging Consultants, Western Division, and Professor of Radiology, University of California San Francisco, San Francisco, CA

When we talk about potential liability in the area of radiation dose, there are three types of law that we should be concerned about: statutory law, civil law and criminal law. Statutory laws are passed by legislators at the local, state or national congressional levels. Once a statutory law is passed, the implementation of that law is generally left to a regulatory agency. The Mammography Quality Standards Act (MQSA) is an example of a statutory law. I will draw many comparisons to the MQSA in this discussion.

Malpractice is generally a state-oriented procedure, and most of the rules are found in the business and professional codes of most states. In the case of federal laws like MQSA, they can be preemptive of state laws. In this way, federal laws can define the minimum standards of care that need to be promulgated. While much of what was discussed in this supplement involves systems-oriented procedure, the law of malpractice (the law of negligence) is handled on a case-by-case basis.

Understanding civil law

Civil law addresses the relationships between individuals or parties. Criminal law is generally a state matter and is enforced by the government. Civil law typically governs what we do as physicians. Particularly, the law of negligence governs malpractice.

Ordinary negligence (sometimes defined as gross negligence) can also pertain to willful and reckless conduct. In these cases, negligence borders on criminal law. In general, when we do the best we can and it falls short of the standard of care, we are not subject to criminal law, unless the offenses are so egregious, or so reckless, that the state has an interest in protecting its citizens against that kind of conduct.

Considering malpractice, the law of negligence is defined by four elements: duty, breach of duty, causation and damages. And it is under those elements, that when we are accused of performing substandard care, that the court, whether it be a judge or a jury, attempts to make an analysis.

Standards of care

One way we derive standards of care is, as I suggested earlier, statutory standards of care. Sometimes, laws are passed that tell us exactly what we have to do. For my own subspecialty of breast imaging, a good example is the MQSA, which very clearly indicates how we have to report findings in our interpretive report. It is no longer an issue about what experts can agree or disagree upon. MQSA and its proposed rule and its adopted rule from the U.S. Food and Drug Administration has used specific language as to how mammography results need to be reported.

Common law and trial law

That law which is not defined by statute is called common law. And basically, common law is the accumulation of decisions where appellate courts have rendered judgments about how conduct is supposed to be handled by the courts. These decisions are applicable to that particular jurisdiction, although other jurisdictions very often borrow appellate decisions.

For example, a very famous Ohio appellate court decision is followed, in most states, with respect to direct communication when there are unexpected results or urgent results. This is what the American College of Radiology (ACR) used to develop its communication standard. So common law has nothing to do with trial law, but is the summation of appellate court decisions around the country.

When we talk about liability in the field of negligence, when individuals go to trial the jury is asked for a determination of a question of fact. Not law. Simply put, did the defendant comport with the standard of care of what a reasonable and prudent physician would do under similar circumstances?

As I start to introduce these topics, please recall some of the information in terms of the radiology facility being responsible for appropriating an appropriate dose, based on image quality goals, also note that the clinician is responsible for trying to assess whether or not a particular examination should be proposed and actually carried out. Remember, appellate court decisions, as I mentioned earlier, deal with questions of law. But every once in a while, when an appellate court is making a determination on questions of law, they will review the facts of a particular case, and issue what are called dicta, which are comments upon standards of care for which we can sometimes take judicial notice.

How does a decision get to an appellate level? It goes through a trial court decision, and will occasionally be taken to the appellate level. Appellate courts are not interested in reviewing all the facts of every case that happens to get lost and appealed. They are interested in questions of law, because when they make a ruling, it has precedent-setting value, unlike a trial court.

Expert witnesses and guidelines

At the trial level, if a question of fact arises, for instance, whether or not the physician comported with the standard of care, then opposing witnesses will often bring to bear information that helps the jury make that decision.

These expert witnesses glean their information from the literature, and from their own personal experience. They frequently bring the issue of guidelines to court. There are more than 300 published guidelines and critical pathways, some of which are in concert with one another; some of which are inconsistent. By example, there are screening mammography guidelines, where the ACR suggests screening studies every year for women over 40 years of age. The American College of OB/GYN does not have that exact schedule for women between the ages of 40 and 50.

So experts will often point to guidelines as a foundation for their opinions. The guidelines are not necessarily evidence that can be introduced into court, because you cannot cross-examine the individuals who developed the guidelines. However, you can cross-examine the expert witness, who is drawing his or her opinion from those guidelines.

The development of guidelines usually involves a very deliberate process. That process is taking greater form with the combination of efforts between industry and medicine. In my area, the Integrating the Healthcare Enterprise (IHE) gradually brought together vendors and physicists, until mammographers found themselves in a situation where they had trouble with connectivity issues with one digital manufacturer vs. another. Radiologists became involved in this dialogue, and it led to the establishment of certain IHE integration profiles.

Challenges with guidelines

There will likely continue to be a greater cooperative effort in terms of establishing certain kinds of guidelines, such as the ones we are talking about today. Obviously, guidelines are strongest when they are evidence-based. But most of the ACR guidelines, as well as those from other subspecialties, lack the kind of robust data that allow all suggested practice parameters to be evidence-based. So we do the best we can in terms of the level of evidence and the level of reliability.

We have heard different opinions today, for example, the use of a bismuth shield for evaluating or at least protecting the breast from a certain amount of radiation vs. adjusting the mA. There will be differing opinions. Presumably, in a deliberative, guideline-setting process, all of those opinions are brought to bear. Where guidelines are not evidence-based, they default to a consensus-based situation.

There can be alternative consensus statements. With opinion-based or consensus-based guidelines, there is greater likelihood of disagreement, depending on who is sitting on a particular panel and who is drafting the guidelines.

It is interesting that we operate, under an image-gently environment, adhering to the tenets of ALARA (as low as is reasonably available). And the reason I find this important is because the law of negligence is based on reasonableness: Did the physician involved comport with a reasonable approach and act with reasonable conduct? There is always going to be disagreement whether or not a particular study should have been ordered, or whether a particular dose approach was employed. These questions will generally come within the latitude of reasonableness when a judge or jury is being asked to make a decision.

In a legal case of negligence, where guidelines are primarily consensus-based, there is a globally accepted alternative school of thought doctrine. If a particular guideline is issued, and especially if it is only consensus-based and not evidence-based, and you choose to conduct yourself differently, and if you can bring to bear similar evidence, and a rational basis for what you do, then that generally is a pretty strong defense in a case for negligence.

Applications to radiation dose

There are various assumptions one can make about radiation damage. One can take a threshold vs. nonthreshold approach. Or one can take a nonlinear or a linear quadratic approach. When the BEIR (Biological Effects of Ionizing Radiation) data started to grapple with the difficult-to-interpret data from the Japanese experience with the atomic bomb, the analysis was very incomplete, and therefore certain assumptions have to be made.

So if one is taking the position that a patient has had too high of a dose, then “too high” becomes a somewhat subjective determination. The consideration that comes to bear is: what is the threshold dose? Depending on the assumption one makes, one can take the position that a certain degree of radiation is more harmful than another degree of radiation, and it is not a settled issue.

Although dose is an important consideration, we are convened in this forum to raise consciousness regarding this issue. When considering the deliberative process of making clinical decisions, dose is generally subordinate to image quality. Further, if we reduce the dose so much so that we obtain nondiagnostic images, then we have, in effect, wasted the dose. If we cannot get the correct image quality, then dose reduction becomes a somewhat trivial issue. Poor technique can certainly be a liability in terms of the law of negligence for unreasonable image quality. Image quality should be the primary effort, and the dose has to condition the image quality, but the two priorities cannot be reversed.

Late effects of radiation

The next question would concern the teratogenic effect in children, such that perhaps 15 or 20 years after receiving 5 or 6 CT studies they develop cancer. The plaintiff’s attorney would likely call an expert witness to testify that the radiation directly caused the harm.

The court system, in terms of the law of negligence, has to place certain conditions upon that kind of testimony. And probably the two most important rules of evidence are federal rules FRE702 and FRE703. FRE702 states that scientific testimony should be admitted into evidence if the testimony will assist the trier of fact to understand the evidence or to determine a fact in issue. FRE703 excludes expert opinion if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading of the jury.

So one can at least propose that any degree of radiation is harmful, but it is going to be a big step to show whether or not there was a causative relationship between that event, that CT that is at issue, or those 5 or 6 CTs that are at issue, and an actual adverse consequence. However, you should be aware that the nature of what judges allow into evidence has considerably widened in latitude. The old rule used to be that if there was general acceptance, in the relative scientific community, evidence would be admitted; and if there was not, it would be excluded. Judges did not even want juries to hear new and novel pieces of evidence. With the speed and change of information and technology, today courts clearly recognize that that kind of ripening of general acceptance and overwhelming acceptance of certain information is probably too narrowly constricted. And so a great deal of evidence is now allowed in the court, presumably with the judge’s discretion, that can be protected by vigorous cross-examination, presentation of contrary evidence, and careful jury instructions.

That being the situation when we try and think about this concept in tort law and the law of negligence, one really asks oneself, especially on this panel, can a reasonable duty be established (a standard of care) for the X-ray facility and the radiologist, as well as the physician. In other words, can causation be sufficiently established between dose and harm, to reliably issue parameters for imaging? Causation requires, as the California Supreme Court once said, a direct nexus between the conduct and the harm involved.


One of the essential elements in the law of negligence, in showing that a tort has been committed, has to do with foreseeability. This is a doctrine that comes up in the jurisprudence involved with the evaluation of tort law. It seeks to determine if something was reasonably foreseeable. The operative term here is reasonable. Is it foreseeable that certain dose-related examinations will cause adverse effects? One of the examples that was discussed was with respect to the clinician vs. the radiologist. What we do might have effects 20 years later, whereas a clinician, in giving an antibiotic, might get a reaction, but it is not going to have a long-term effect.

There are variations on that theme. When we treat young patients with chemotherapy, we have come to recognize that 20 years later there are secondary malignancies that may occur. One must weigh the equities, the benefits, of treating vs. the potential long-term effects. The issues involved in this kind of reasonable foreseeable analysis, have to do with total body dose, as well as organ-specific dose. How do we then talk to parents when we do not know if the effects will be developmental, genetic, cancer-related or immunosuppressive?

If I know the patient has had 3 CT scans and if I need to do fourth, I have no way of knowing, in most circumstances, that patient may have to undergo 5 more CT scans after the present issue is addressed. In other words, the patient may be subject to multiple traumas or multiple bouts of infection. So if negligence law is partly predicated upon foreseeability, it is a very difficult analysis, at any given moment, for the radiologist or the clinician to foresee what is going to happen down the line.

Informed consent

Another defense to a negligence claim is informed consent. Informed consent is a process, not simply a form that the patient signs. The signed form is evidence, but not proof of the consent. Sometimes policies and procedure manuals that are drafted in a deliberate way may be referenced in the situation of informed consent, but basically informed consent talks about the material risks and the likelihood of those risks that are discussed with the patient, as well as the alternatives that can be discussed. Under certain emergency situations the consent process is not required.

The way to think about informed consent is that the burden of obtaining informed consent is related to the probability of an adverse event and its severity. So when we try to analyze whether or not we should be, on a case-by-case basis, obtaining informed consent from patients, one of the struggles that we have to contend with is that we are not exactly sure what the severity of the event is—namely, what is the severity of the consequence of a particular radiation dose.

Let me return briefly to the recent Sodickson paper which looked at patients with chronic conditions, and found that >50% had 10 CT scans within the past 8 years. Now, most of us would agree that is a large amount of CT scans.1 Notwithstanding our vendors’ attempts to lower the dose, that can be a lot of radiation.

The dose considerations for CT certainly are promoted by the ACR. The indications for CT are sometimes promoted by guidelines as well. But what the ER physician is thinking about, even if he is very concerned about the dose, is that he would rather take the risk of being sued for radiation dose over the potential misdiagnosis.

Under current circumstances, more often than not, the clinician acts to minimize the risk of a potential misdiagnosis. There is very clear tort law in that respect. If the radiologist pushes back about excessive radiation on an equivocal case, the physician is likely to still order the scan even if it is probably not indicated because they are worried about a potential misdiagnosis. The alternative school of thought doctrines provide a defense if the physician is later challenged on the adverse effects of radiation dose.

Government initiatives

I think most physicians do not want legislators dictating how they should practice medicine. But the solution depends on how we define the problem. With screening mammography, the radiology community became engaged in somewhat of a public health effort. Screening mammography is unlike the usual diagnostic paradigms where a patient is ill, sees a clinician, is sent to us for imaging, which contributes to the evaluation of that patient, and thus closes the loop.

About 30 years ago, a survey of Michigan-based facilities found that the dose was too high in screening mammography. As a public health initiative, the federal government decided to control the public safety by issuing certain guidelines. So we now have a maximum dose that is statutorily specified in the MQSA guidelines. Anyone who exceeds that dose will not be approved to perform screening mammography.

But there are gaps in this public health approach. There is still no consideration of cumulative dose from mammography. We only regulate screening and we do not regulate how much radiation is imposed upon diagnostic mammography examinations, or adjunct imaging.

When discussing cumulative dose, we have considered a report card, which would be similar to immunization records. Now that we have a national practitioner data bank, the implementation of that kind of cumulative radiation dose would be helpful.

As technology changes, so do the practice patterns. The government is incentivizing us for electronic medical record use. It has been brought up, earlier today, that this could indeed be part of the medical record. Thus, the clinician who is deciding whether or not to order an extra CT for a trauma victim in an equivocal situation should could turn to a look-up table, so long as it is provided in an easily accessible manner.

So, for example, clinicians who want to quickly assess a risk model for the patients coming in for mammography have a number of Websites to consult for this information. Based on a patient’s history, clinicians can quickly determine a woman’s risk of developing invasive breast cancer. If we had a national data bank for dose accumulation, so long as it was very easy to use for the clinician, it could be part of the decision-making process.

The ACR can also take the lead in establishing appropriateness criteria. But if we simply create a document, many of the initiatives will not be implemented. I would venture to say that most radiologists practicing today do not know what the appropriateness criteria are, in many of the specialties in which they practice. So this is a very grand challenge, and it requires a somewhat grand solution. It is not enough to simply create documents and recommendations.

The issues in radiation dose do not lend themselves to an analysis based on negligence (substandard care), nor do they lend themselves to an informed consent analysis because we are still not clear on the likelihood of an adverse event or the severity of that event. This provides ordering physicians an opportunity to second guess the equities between radiation risk and resolving the clinical status.

Statutory intervention

So instead, I am suggesting that we look back at the areas of law which might apply to us, and as intuitively offensive or resistant as it might sound, it may be a regulatory or administrative solution that is required. The dose parameters can be established for facilities. We are already working on that with the manufacturing community. We are working on that at our own institutions. We can come up, within the college of radiologists, reasonable dose determinations and dose reduction strategies within which we can operate. And it might be best done by self-policing, through accreditation programs, similar to the ones we have today. If we cannot implement this then it can be implemented in a regulatory fashion.


So any regulatory approach, whether it is taken by our specialty colleges or by the government, has to involve the effort of documenting a deliberate decision, especially considering that ordering clinicians are under no external pressure to consider the dose from studies that they order. It is unlikely that ordering habits will change at the order-entry side. Clinicians will not stop to consider dose in the ER or even in the clinical wards. So the only way we can expect change is if the clinician is now required to be deliberate about that decision, because of a regulatory effort to be considerate of the dose. Unfortunately, this is the only way we can create an environment where this very important consideration will be implemented on a general basis.

Chances are that an ER physician, who is still worried about the “hot belly,” is going to get a CT scan, whether or not he checked lookup tables about the appropriateness of the exam. For other situations, where the radiologist is suggesting alternative approaches, where practice parameters have been developed within that community to encourage the physician to consider the alternatives, I think that creates amuch greater chance of imposing, in this legal environment, the kind of conduct that we are advocating.


  1. Sodickson A, Baeyens PF, Andriole KP, et al. Recurrent CT, cumulative radiation exposure, and associated radiation-induced cancer risks from CT of adults. Radiology.2009;251:175-184.


Back To Top

Minimizing radiation risks in CT imaging: Medical-legal considerations.  Appl Radiol. 

December 28, 2009

Copyright © Anderson Publishing 2020