130: Risk Assessment and Risk Communication

All health care professionals confront questions involving risk on a daily basis. In the area of toxicology, these questions may take many forms. An anxious parent with questions about a child’s potentially toxic exposure, an urgent consultation for a critically ill patient in the emergency department or intensive care unit, a request to interpret a laboratory test, media requests for information about environmental public health issues, a response to a hazard-materials situation, and biopreparedness education all involve directed communication of information and recommendations. Toxicologists and Certified Specialists in Poison Information (CSPIs) must establish rapport and provide information, instructions, and when appropriate, reassurance, typically by telephone or in short face-to-face interactions. For CSPIs, attribution of the patient’s complaints to one or more potential exposures and ascertaining the true reason or concern behind a call are also ­difficult given the limited information and time and lack of visual clues that are usually available during a clinical evaluation. All of these situations require a knowledgeable, compassionate, and well-reasoned response.

This chapter focuses on two particular components of this response: risk assessment and risk communication. These principles apply equally to individual calls to poison control centers, interactions with the public and medical professionals in educational outreaches, occupational and environmental exposure evaluations, and supportive roles with other public health agencies in bioterrorism preparedness, environmental public health tracking programs, and research.

In the context of this text, risk assessment is the process of determining the likelihood of toxicity for an individual or group after a perceived exposure to some substance, generally referred to as a xenobiotic. It involves determining the nature and extent of the exposure (ie, xenobiotic, dose, duration, route) and its specific clinical effects, defining an exposure pathway, and assessing the likelihood of effects from a given situation. A published body of knowledge can be applied to some components of risk characterization or assessment. An overview and a number of tools can be accessed through the Web sites of the US Environmental Protection Agency (EPA) and the Agency for Toxic Substance and Disease Registry (ATSDR) of the Centers for Disease Control and Prevention (CDC).^1,9 However, any given risk assessment is often based on incomplete information. This may include such features as uncertainty regarding the exposure xenobiotic or mixture, whether there has been an actual exposure or just proximity to the xenobiotic (completion of an exposure pathway), lack of the exact dose, or unpredictable features such as host factors (underlying medical conditions or genetic polymorphisms) that could modify the response to a potential exposure. Unfortunately, those conducting a risk assessment are affected by their own biases and assumptions in the interpretation of their results, as are the people to whom a risk assessment is communicated. The emotional response to being “poisoned” makes evaluation and attribution even more difficult.

A good example of the practical difficulties involved in a risk assessment is evident from a published description of mass psychogenic illness.¹⁸ In this incident, many individuals at a school complained of odor-triggered symptoms that spread in a so-called “line of sight” transmission with no evident dose–response pattern. Extensive testing identified the possibility of potential sources of exposure, such as dry floor drain traps. However, no actual release was documented, nor was a scenario that would present significant harm identified. Yet symptoms recurred when people returned to the school. This is considered an example of psychogenic illness. The extent of investigation of these events can be profound, highlighting the difficulty in appropriately applying potentially unlimited laboratory technology to a situation. In addition, our ability to assess a “no-risk” situation is limited, as can be noted in the letters to the editor of the journal in which the article was published criticizing the methods or conclusions in this event and “subsequent and comparable” outbreaks.^4,16,23

The response of individuals to uncertainty correlates with their affinity for one component of the negative data paradigm—“the absence of evidence of harm” versus “evidence of absence of harm.” Both of these positions should have at their core the continued evaluation of evidence as it becomes available, with subsequent refinement of a resulting risk assessment. Unfortunately, these potentially converging points on a spectrum of knowledge and research have been polarized in debate and policy as two opposing principles: the Kehoe Principle and the Precautionary Principle. The Kehoe Principle is best summarized as “prove something is harmful before excluding a product with known benefits because of concern about potential, unproven future adverse effects.” A common example of the use of this principle is the continued marketing of a xenobiotic after another member of the class has been removed because of safety issues. More intense scrutiny may be indicated, but a class-wide medication recall without evidence of some level of harm by an individual therapeutic drug is very rare. This principle has been misused in the past to minimize known risks attributable to environmental lead pollution to delay removal of lead additives from gasoline.²⁶ Critics of the Kehoe Principle have suggested that waiting for evidence of harm from a substance results in costly or irreparable damage.

The alternative position of the Precautionary Principle is often summarized as “where there are threats of serious or irreversible damage, lack of full scientific certainty shall not be used as a reason to postpone cost effective measures to prevent environmental degradation.”^21,30 Critics of the Precautionary Principle often cite the lack of attention paid to the “cost” component of the principle, complaining that devotees stifle economic growth and prosperity with unfounded fears rather than reasoned consideration of known data. This principle is commonly extended to potentially harmful situations other than the environment. A common criticism of the Precautionary Principle is the degree to which alternative actions have been evaluated for safety. As an example, concern about thimerosal safety (as a vaccine preservative and source of ethylmercury exposure) in young children led many parents to forego childhood vaccinations. Vaccine manufacturers have removed the thimerosal preservative from most routine childhood vaccines. Although this process was accelerated by the theoretical—and, ultimately, unfounded—concerns regarding this source of mercury exposure in young children, the cost of delayed or omitted vaccination was a number of real and preventable infectious diseases that occurred, including hepatitis B.^5,6,11

Although the Kehoe and Precautionary Principles originated as policy approaches to public health issues, they underlie the automatic or subconscious biases that each individual brings to his or her own personal risk assessment or tolerance. The response to uncertain situations is derived from one’s framing of belief systems and assumptions about life, justice, and eternity.^3,10,14 These underlying world views should be explicitly recognized and addressed in a formal risk assessment. Table 130–1 lists the components of a risk assessment.

It is difficult to translate public health risk assessment done for populations by entities such as the EPA to the individual level. Simplistically, the iterative process of adjusting known noncancer adverse exposure outcome limits in an animal model (eg, lowest observed adverse effect level {LOAEL} or no observed effect level {NOEL}) to a safe level of exposure for all humans (including so-called “sensitive subpopulations”) has been arbitrarily set at repetitive multiplicative factors of 10 for each of these extrapolations. These are called “uncertainty factors” and are used in the absence of specific information about human exposure to identify a conservative human “safe dose.” This would be set at 0.001 times the animal model LOAEL, reflecting a 10-fold reduction in dose for extrapolation from LOAEL to NOEL, another 10-fold reduction for extrapolation from animal to human, and another 10-fold reduction for potentially “sensitive” human populations. These “uncertainty factors” are actually safety factors for the adverse effect of interest. When individuals exceed these limits, they remain protected by very robust safety factors, and are not—as it is often portrayed—exposing themselves to a defined harm. An example of this is the dietary guidelines (also known as “fish health advisories”) for fish consumption based on concern about exposure to methylmercury and polychlorinated biphenyls. These dietary recommendations are based on epidemiologic studies that suggest subtle neuropsychiatric abnormalities in maternal–fetal pairs (in at least some populations) from levels of fish consumption 10 to 100 times that of the “usual” American diet. Clinical mercury toxicity requires still higher levels of consumption. Although it seems reasonable (precautionary) for pregnant women to limit their intake of certain high mercury-containing fish species, it is inappropriate to avoid fish and the nutrients contained therein because of a misplaced fear of mercury exposure. Furthermore, these risk assessments do not apply to nonpregnant women, children, or men, although they are often generalized to all humans.²

The public health modeling concept for cancer health effects risk assessment is even more complex and often misunderstood. In this setting, modeling assumes a linear, no-threshold carcinogenic effect from exposure to a given xenobiotic. Cancer risks from exceedingly small exposures are extrapolated from data in cancer-prone animal models with exposures so large that they are only likely in the experimental setting. This experimental construct ignores the moderating impact of incremental dosing of small amounts over a protracted period of time, and potential metabolic and self-repair mechanisms in humans. The “acceptable risk” in this setting is then taken as “one excess cancer in a population of 1 million exposed individuals.” Although a limitation of this model is explicit in the statement that this estimate represents “a plausible upper bound estimate of risk at low dose where true risk may be lower, including zero,”¹² this important caveat is rarely communicated, resulting in the common response by individuals that “an extra cancer may be acceptable for you, but not when it is my child.”

As this brief review of the principles of regulatory approaches to noncancer and cancer effects demonstrates, risk assessment is often imprecise. Risk characterization for the individual should avoid unfamiliar statistical concepts, aiming instead to communicate the likelihood of significant risk for the exposure actually experienced. The CDC has codified this approach for a community using such escalating terms as “no public health concern,” “public health concern,” “public health threat,” and “immediate risk.” Although there is debate about the general use of these terms in isolation (without explicit explanations), a similar approach with an individual could summarize the risk assessment using the phrases “safe,” “no adverse effects are expected,” “may be of concern—we will need to do some follow-up testing,” and “this might be (or is) a problem—let’s do the following studies or treatments.”

Risk communication consists of an exchange of facts and opinions that allow an individual or a group to make an informed decision regarding a course of action or treatment. Practically, risk communication is a way of translating incomplete knowledge so that individuals can achieve informed decision-making. During a one-on-one interaction with a poisoned individual and his or her family or a caller to the poison center, there is a need to gain the fullest attention or cooperation of the individual. After this has occurred, the discussion is usually focused on the risks and benefits of various treatment options (eg, gastrointestinal decontamination) or possible diagnostic modalities (eg, observation versus neuroimaging versus antidote administration). The group dynamics of environmental exposure risk communication at a public meeting are very different. Federal agencies that interact with communities in “Superfund” sites (eg, the EPA and the ATSDR) have promulgated principles and practical recommendations for risk communication in this setting. Table 130–2 summarizes general principles of risk communication.

Although some of these recommendations are more applicable to longer-term deliberations and interactions, much of the individual communication done by the poison center and medical toxicologists succeeds or fails based on these same principles (Table 130–2). Lacking the opportunity for repeated interactions over time to identify and discuss assumptions and biases, toxicologists need to establish credibility, listen to concerns and empathetically respond, admit areas of insufficient knowledge, and commit to follow-up interactions to convey effectively a risk characterization for an individual based on the available knowledge and experience. The scientific terms, rationale, and any extrapolation from modeling (eg, animal data or case series) should be conveyed in an understandable manner to show that appropriate safety factors are incorporated into areas of uncertainty as a risk-diminishing step. The patient or audience should leave the interaction with a clear understanding of the difference between a short-term risk of symptoms that will resolve or result in serious illness and the degree of certainty about the potential for a long-term consequence.

An example of poor risk communication can be found in the immediate aftermath of the World Trade Center disaster in September 2001. The mass rescue and recovery response was largely voluntary and heroic; however, inadequate attention was paid to the importance of respiratory protection against the heavy particulate and alkaline dust in the early hours after the towers collapsed. The high incidence and persistence of cough and other respiratory symptoms in responders have been attributed to this exposure. Communication of the real risk of respiratory symptoms to early responders would have emphasized the critical importance of appropriate use of personal protective equipment. This would have been balanced against the time-limited possibility of saving lives of those potentially trapped in largely inaccessible locations. Since 2001, a proliferation of other associations to World Trade Center dust exposure (including low birthweight) are reported, which are of doubtful validity.³¹ Appropriate risk communication to people concerned about these reports should emphasize the important role of confounders, the investigational nature of the study hypothesis, and the lack of relevant risk to any given individual.

Effective risk communication must therefore address several questions. After the best information has been obtained about the identification of the xenobiotic and the nature of the exposure, the following must be conveyed:

• Likely magnitude of the risk: This includes information on the process by which the person would be exposed (ie, the exposure pathway), such as airborne inhalation or drinking water delivery via a contaminated plume in the ground water and dose–response, such as: “Does the reported exposure to a particular xenobiotic (amount and duration) approach the exposure amounts reported to cause symptoms?”

• Urgency of the risk must also be conveyed along with practical recommendations for simple actions consistent with the level of urgency.

• The applicability of a risk characterization might also need to be addressed. Are the animal data applicable to humans? Is the exposure something of concern for an individual?

• Uncertainties of the risk assessment: This could include a “worst-case scenario” approach to unknown exposures or uncertainties in the quantity of an absorbed dose. The need for continued observation or follow-up for clinical changes would be expressed here. Individual risk tolerance may vary greatly. The same information may be interpreted differently by risk-averse versus risk-tolerant people. A variety of comparisons or communication techniques may be used to provide an adequate characterization of risk. It is important to remember that the public and even medical professionals have limited ability to understand and incorporate data that rely on numbers and statistics. This health numeracy limitation is even more prevalent than is limited health literacy (Chap. 135).¹⁵

• Management options: In addition to follow-up and repeated evaluations by a medical toxicologist, the range of choices, associated with their relative benefits or risks, should be presented to the individual or group of individuals. A summary recommendation or opinion from the presenter should emphasize specific steps people can take to decrease exposure or potential toxicity if indicated by the level of risk. This last step is important because uncertainty significantly impacts the ability of an individual to take appropriate action. People should not leave the meeting or interaction with the impression that “no one knows what is going on or what we should do.”

Although it has long been recognized that many home-initiated poison center calls concern nontoxic or minimally toxic xenobiotics potentially ingested by children,²⁴ the frequent lack of documented ingestion raises the possibility of under-triage based on misplaced confidence that relies on prior experience. It is generally assumed that the sheer volume of calls provides some reassurance regarding the accuracy of our risk assessment of these xenobiotics, but we should remain cautious in our interpretation of poison center data^5,17,27 (Chap. 136). Moreover, even calls about nontoxic xenobiotics require communication between the caller and CSPI beyond simple substance identification. The importance of risk assessment and communication principles can be seen in the joint position statement²² on the prehospital management of “minimally toxic substances” crafted by the American Association of Poison Control Centers (AAPCC), the American Academy of Clinical Toxicology (AACT), and the American College of Medical Toxicology (ACMT). According to the position statement, a CSPI can make a risk assessment that an exposure is benign or minimally toxic only if the following characteristics are true²²:

• “The information specialist has confidence in the accuracy of the history obtained and the ability to communicate effectively with the caller.”

• “The information specialist has confidence in the identity of the product(s) or substance(s) and a reasonable estimation of the maximum amount involved in the exposure.”

• “The risks of adverse reactions or expected effects are acceptable to both the information specialist and the caller based on available medical literature and clinical experience.”

• “The exposure does not require a health care referral because the worst potential effects are benign and self-limited.”

The position statement further notes that patient disposition decisions can be altered by many additional factors, including intent, environment, presence of symptoms (possibly unrelated to the xenobiotic in question), and ongoing review of current recommendations in the face of more data.²² These points emphasize both the dependence of the CSPI on information derived from the caller and his or her confidence in the level of comprehension of the caller. The caller should understand that his or her exposed child is safe; the conversation with the toxicology experts should alleviate concern as the nature of the assessment process is explained to whatever degree is necessitated by the caller’s risk tolerance.

In the case of a symptomatic patient or a hospital- or physician-­initiated contact to a poison center or medical toxicologist, the caller should expect more than just xenobiotic-related information; he or she also expects knowledge and expertise that will provide reassurance or direction for improving the patient’s health status. However, merely relaying information regarding the diagnosis, course, and predicted outcome is insufficient. There is often another underlying reason for the call. This could be anxiety, uncertainty, or misinformation established by an individual’s previous experiences or knowledge base. A sense of guilt may underlie a parent’s call for an inadvertent exposure occurring when a child was unsupervised. A health care professional may have previously had significant difficulties in the management of a poisoned patient. If these issues are not addressed, then the caller may continue seeking reassurance by repeated calls to the poison control center or by seeking additional input from other sources, such as family, friends, primary physicians, other health care providers. Any variance in the information obtained from these sources may be construed as inconsistencies between supposed experts rather than differences in emphasis with regard to the same information, leading to further uncertainty for the caller. Of further concern, the Internet has become a common source of second opinion for health care. Although many sites are useful, there is no quality control or filter to sort good information from bad or even harmful advice.¹³ Table 130–3 lists some barriers to effective risk assessment and communication.

CSPIs and medical toxicologists frequently encounter callers or individuals at community events or interact with the media regarding public health–related issues, such as heavy metal exposures involving mercury, lead, or arsenic or concerns about “toxic mold,” plasticizers, and other environmental xenobiotics. Often these people are concerned that their symptoms or future personal or family health may be adversely impacted by such exposures. Such supposed exposures are usually poorly documented, sometimes also driven by popular media descriptions or litigation, and the risk is virtually impossible to ascertain during a short telephone or personal interaction. In these situations, the individual is best served by referral to a primary care physician with toxicology consultation or directly to a medical toxicology clinic. The data and perceptions can be completely reviewed and a more appropriate risk assessment communicated in those settings. These interactions are very difficult because they are often emotionally and politically charged.²⁵

In general, the communication of and response to information depend on a preexisting world view and prevailing circumstances. The same possible outcome will be perceived as more or less severe depending on factors other than the nature of the outcome itself. Several authors have characterized the perceived tolerance to different risks, stratified by features such as familiarity and personal control^10,30 (Table 130–4). The emotional response of individuals confronted with these risks is sometimes characterized as “outrage.” One communications specialist has posited that “risk = hazard + outrage.”²⁸ He characterizes situations in which there is a significant hazard but little outrage as requiring “precaution advocacy,” essentially informing the relevant parties of the need for more action or involvement to reduce risk. On the other end of the spectrum is a situation with little hazard but significant outrage, which requires “outrage management” to address fear or anger that is dissociated from the actual hazard posed by the situation. The greater the degree of familiarity with the particular exposure situation and the greater the voluntary nature of the exposure, the less fear or outrage will be expressed for a given adverse outcome, whether this is an appropriate response or not. Although not necessarily applicable to the initial “fight-or-flight” response to an emergency, these concepts are certainly applicable to the aftermath of these events. Of note, although risk communicators use analogies to place exposures into a context familiar to their audience, one must be careful to avoid equating voluntary and involuntary risk assumption or equating those exposures or risks assumed by one segment of the population unequally. An example of this is the use of a smoking risk analogy when talking to a nonsmoker.

Risk communication has become very important in the setting of preparedness for terrorism. Although a great deal of attention and money have been directed to improvement of public health infrastructure, reporting and surveillance mechanisms, and response to perceived and actual terrorist acts, less attention has been directed to the process of communicating risk to the individual.^8,20 Although some countries practice public health emergency drills regularly, the United States has concentrated on development of organizational structures and lines of authority, with attention to the importance of outcome-based exercises only recently emphasized.

Maintaining readiness for catastrophic terrorist events (or natural occurrences such as pandemic influenza or storm related flooding) should use the same risk assessment and communication techniques appropriate for other urgent public health matters. Unfortunately, many factors affect the characterization of risk other than the facts. The importance of presentation style, perception, or the role of the communicator’s own biases is exemplified by these two composite articles describing the same events:

1. Unknown assailants have infiltrated the mail delivery system, resulting in severe illness and death of children, healthy adults, and elderly people throughout the country. The initial symptoms can be nonspecific but rapidly progress to death if treatment is not begun early. The medical community routinely fails to diagnose the conditions early, and the government has no system in place to detect this threat after it occurs. The long-ignored public health system is not prepared to deal with the huge burden of preventing illness in those who may have been or will be exposed. Anyone who receives regular mail may be at risk. Tens of thousands of our citizens are taking prophylactic antibiotics “just in case.” If you receive any unusual packages or see collections of powder that do not have an obvious explanation, call the police. If you develop a fever, cough, chest pain, or unusual rash, which may not be painful, seek medical attention at once. Tune in to your local news station for more information on this burgeoning threat to our nation’s security.

2. A small number of individuals in isolated exposure settings have developed illnesses after bioterrorism events. Most people have survived these exposures, particularly with early and proper medical care. The government has developed a case definition, and medical experts have disseminated information to assist the medical community and public in the early recognition of symptoms and signs that are consistent with this exposure. Prophylactic treatment within days of exposure of those in high-risk professions, such as mail handlers at major postal sorting facilities, prevents illness. Unfortunately, there have been a large number of hoaxes and false alarms about possible terrorist events and a lot of understandable fear in the community about nonspecific symptoms. For more information, contact your local health department or use the CDC Web site: http://www.bt.cdc.gov/agent/anthrax/needtoknow.asp.

Both of these paragraphs describe the 2001 anthrax bioterrorism events within the United States, during which a total of 22 people become ill, of whom five died from anthrax exposure. The first communication suggests that everyone is at risk and the situation is dire; the communication in the second paragraph is that the risk is isolated (a single individual died who was not in what was recognized as an at-risk setting from the seven identified mailings) and there is a plan and process being developed to respond to the threat. Whereas the first is sensationalistic, imparting a helpless victim role to the reader, the second provides a framework in which to assess one’s personal risk and access to sources of reliable information. Both types of reports were prevalent after the 2001 anthrax attacks. Which report seems more complete, accurate, or useful is determined by the assumptions and perspectives of the reader in addition to the message the author wishes to deliver or response desired. Some would say that communicating a high degree of risk is important to gain the attention of the reader and to ensure that no one ignores a warning. However, the lack of a risk perspective prevents the reader from placing this information in context with the myriad other risk communication messages conveyed on a daily basis. In general, risk communication messages that do not provide a context or comparison to generally familiar activities or risks are more prone to misinterpretation or misapplication. As biopreparedness moves from public health infrastructure development and surveillance improvement to planning and response drills, appropriate message development and risk communication to the public need to be emphasized.

• High quality risk assessment and effective risk communication are the hallmarks of a successful interaction between the public and poison centers and between a toxicologist and an individual patient, the media, or the public health community.

• Adherence to general principles includes obtaining the best information possible regarding potential exposures and conveying in an understandable fashion a risk characterization of the hazard, likelihood of a completed exposure pathway (thus, the likelihood of an actual exposure), possible health effects, and treatment options.

• It is important to clarify the difference between public health standards and individual exposure risks, with an understanding of the many psychosocial issues that influence perception.

• Information should be provided in a context that allows the individual to prioritize his or her response based on a factual and balanced presentation with respect to his or her health literacy and health numeracy.

References