Hungerford, Daniel W. DrPH
“The important thing in science is not so much to obtain new facts as to discover new ways of thinking about them.” - -Sir William Bragg, British physicist (1862-1942)
With a great deal of support from other groups and federal agencies, Centers for Disease Control and Prevention (CDC) convened two conferences to discuss the impact of substance use disorders on patients in acute care settings–one in 2001 on alcohol problems among emergency department (ED) patients, and this conference on alcohol and drug problems among trauma center patients. Both conferences aimed to promote understanding of a new conceptual model for substance use disorders and to initiate greater collaboration between clinical physicians and substance-use treatment professionals and researchers. As an epidemiologist at CDC’s National Center for Injury Prevention and Control (Injury Center), my central mission for the past eleven years has been injury prevention and control. I am convinced that it is not possible to successfully accomplish injury control without addressing the nation’s alcohol and drug problems. These proceedings confirm that trauma centers are a prime setting in which to identify and help patients with substance use problems. It is doubtful that any other clinical medical setting has a higher prevalence of patients with alcohol and drug problems than trauma centers.
Note: Readers who want a quasi-interactive understanding of the work ahead and a sense of “being at the conference,” are encouraged to read the detailed discussion summaries for each of the five sessions. Combined, the discussions and the papers in this issue contribute to our understanding of the complexity and importance of alcohol misuse and drug abuse in trauma care settings and the opportunity that trauma centers provide to address substance use problems.
In Alcohol: The Ambiguous Molecule,2 Griffith Edwards describes beverage alcohol’s pervasive role in the history of human societies, as religious sacrament, agent of disease, and recreational drug. No wonder then, that historically, a variety of conceptual models3 have been used to understand how such a simple molecule–two carbons, five hydrogens, and one oxygen atom–can have such a profound effect on individuals and society. Much of the following discussion will refer to alcohol, but some of it will also apply to illegal, over-the-counter, and prescription drugs. The illegality of some drugs can be a complicating factor; however, many concepts regarding behavior, behavior change, screening, and treatment are similar for both illegal and legal substances.
PREVAILING CONCEPTUAL MODELS
People generally have a conceptual model they use to understand illnesses and behavioral conditions, and that model includes notions of etiology and treatment. Even though people may not have consciously thought about how alcohol and drugs affect individuals and society, their conceptual model is susceptible to subtle, but pervasive and prevailing, social beliefs. Physicians and health care professionals are not immune from such social transmission of models of thinking, particularly when dealing with topics in which they have little training. Therefore, a brief description of prevailing models will be presented before introducing the new model alluded to above.
Dispositional Disease Model
Since the middle of the last century, the dispositional disease model, generally called alcoholism by the lay public, has dominated the American medical community’s understanding of alcohol-related problems.4 This model considers alcoholics qualitatively different from other drinkers; the crucial differences are that alcoholics lose control over drinking once they start and that the disease, although incurable, can be suppressed through abstinence. A central tenet of the disease model is that alcoholism is an abnormal, constitutional disposition influenced by enduring biological factors, affecting only part of the population. This model is beneficial because it encourages society to care for alcoholics with compassion. In place of stigma and punishment, the model fosters humane treatment methods and encourages the medical community to become involved in treatment. Yet, the disease model has a negative side–it supports the belief that only a subset of the total population is vulnerable to the harmful effects of beverage alcohol.3,4 Because only a fragment of the population exhibits this constitutional abnormality, perhaps as few as 4%,5,6 the public, in general, seems unconcerned about how much or how often they drink and views alcohol-related problems as not inherent in the use of alcohol.
The beverage-alcohol industry can easily subscribe to the disease model, yet still accepts the reality of addiction to alcohol and supports treatment for alcoholics. However, by focusing on the vulnerability of only a small portion of the population, society can be distracted from broader measures that can reduce the individual and societal harm associated with alcohol misuse. Put in terms of a commonly used public health paradigm–host, agent, and environment–if alcoholics (the host) are considered the problem, society will be less likely to base policies on the fact that under certain circumstances, or when engaging in certain activities, alcohol (the agent) is a potentially dangerous substance, especially for certain high-risk groups such as adolescents or patients with underlying medical conditions. Consequently, society will also be less likely to embrace broad preventive measures designed to create an environment less conducive to misuse and harm, alochol tax increases, restrictions on availability, and anti-drunk driving measures.7
Alcoholics Anonymous (AA) Model
Historically, the dispositional disease model has often been confused with the other prevailing conceptual model in the United States, that being Alcoholics Anonymous (AA).4 Although the organization offers no formal statement on the cause(s) of an individual’s drinking problems, AA, like the dispositional disease model, places the locus of the problem on the individual and prescribes abstinence as the preferred therapy. To the extent that these two models promote humane treatment and encourage individuals to take responsibility for changing their behavior, the models are useful and seem compatible to the public and medical community.
NEW RESEARCH METHODS
Before the 1950s, physicians and others explored a plethora of treatment modalities to treat or cure alcoholism. However, many of these putative treatments were bizarre by modern standards, and few could truly be categorized as research. In a chapter from Alcohol: The Ambiguous Molecule entitled “In the Name of Treatment,”2 Edwards states, “Over the years, what has been done to people with drinking problems in the name of treatment beggars belief.” In the middle of the last century, the world of medicine was being introduced to new empirical research methods that increased the scientific rigor of studies in human populations. In 1948, the first double-blind properly randomized trial evaluated the use of streptomycin for treating tuberculosis. Several years passed before these new methods were applied to addiction treatment studies. When these methods were used, researchers were surprised by their findings and were forced to re-evaluate prevailing treatment practices.
Enter Empirical Research and Brief Interventions
By 1977, randomized trials in the United Kingdom surprised researchers by demonstrating that specialized alcohol treatment consisting of many counseling sessions did not help patients who were misusing alcohol any more than brief, onsite counseling during family clinic visits.9 Starting in 1978, these results were replicated in the United States. In a series of studies, Miller and colleagues showed that “whereas a small dose of counseling appears to be much better than no intervention, increasing the dose does not necessarily yield greater gains.”10 These studies and those of other researchers revealed two other important findings. One, specific measurable therapist behaviors–showing empathy or being confrontational–were predictive of positive and negative patient outcomes, respectively. Two, brief empathic interventions with dependent drinkers (read “alcoholics”), drug addicts (including marijuana abusers), diabetics, and cardiovascular rehab patients showed positive results.10
In 1979, studies showed that brief advice by physicians increased the probability that patients’ smoking cessation efforts would be successful.11 Since then, further trials have confirmed and broadened findings that skillfully and appropriately delivered brief advice provided in different clinical settings (including emergency departments and trauma centers), can significantly increase the likelihood that patients will reduce harmful substance abuse and, in many cases, stop drinking altogether. In 1985, the World Health Organization initiated a collaborative, randomized, brief alcohol-intervention trial in community health centers in 10 countries on five continents. Results showed that 5 and 20 minute interventions significantly decreased average daily alcohol consumption and intensity of drinking, compared with controls.12 Many studies have shown that brief alcohol interventions–some as short as 5 and 10 minutes–reduced injuries requiring emergency department (ED) or trauma visits, length of hospitalization, and alcohol consumption.13–22 More recently, brief intervention has been shown to reduce cocaine and heroin use.23
Recent studies also indicate that screening and brief interventions are cost-effective and can have a prolonged beneficial effect. In a trauma-center cost-benefit analysis, every dollar spent on screening an intervention saved $3.81 in direct injury-related costs.24 In a primary-care benefit-cost analysis, the average benefit per patient was $1,151 and the economic cost of the intervention was $205 per patient, for a benefit-cost ratio of 5.6:1.25 The same research group followed patients 48 months after their interventions and measured three different ways that patients had significantly reduced alcohol use compared with the control group.26 The benefit-cost ratio for future health care costs was estimated to be 4.3:1. Screen-positive subjects from a community-wide screening program in Tromsö, Norway were given a brief alcohol intervention in 1986. Mean serum GGT values for these subjects in 1995, nine years later, were significantly lower than their baseline values and reductions were significantly greater than for a control group.27
Sufficient evidence has accumulated for the US Preventive Services Task Force to recommend “screening and behavioral counseling interventions to reduce alcohol misuse… by adults, including pregnant women, in primary care settings.”28 As evidence of effectiveness mounts in other clinical settings, varied expert and consensus groups suggest that screening and brief interventions be extended to trauma centers.29–33 Recommendations from these proceedings suggest that research funds be broadly allocated. Rather than expend virtually all funds on efficacy trials in a few academic trauma centers, implementation studies on screening and brief intervention methods should be funded and tailored to real world trauma care settings to increase efficiency and effectiveness in those settings.34
Enter Epidemiologic Research
In an effort to evaluate treatment programs, studies enrolled patients who presented for treatment of substance use problems. Using modern scientific techniques, these studies led the way to understanding that brief alcohol and drug counseling could be effective. However, it wasn’t until population-wide studies were implemented that greater understandings of the total societal effects of alcohol were appreciated. In 1961, Kerr White showed studies based only on patients treated in university medical centers are limited and biased in their ability to obtain an accurate impression of health problems in the community because “…in a month, only 0.0013% of the “sick” adults [1 out of 750]…or 0.004% of the [medical] patients [1 out of 250] in a community are referred to university medical centers.”35 This epidemiologic shift in thinking became evident in the late 1950s when epidemiologists at the Alcohol Research Group (ARG) in Berkeley, California, began to collect and analyze data from general-population surveys on drinking behavior.36 In 1969, American Drinking Practices was the first published study “to describe drinking practices and their correlates on a random sample of the US population.”36 In the ensuing four decades, ARG implemented 11 national surveys; analyses based on these surveys helped establish the public health approach to alcohol problems. In the process, this population perspective changed conceptual models about alcohol misuse and its consequences. In the 1970s, the federal government created research institutes to address alcohol and drug issues–the National Institute on Alcohol Abuse and Alcoholism (NIAAA) and the National Institute on Drug Abuse (NIDA). These agencies not only support empirical research on treatment methods, but they also support studies of substance use behavior and associated health and social outcomes in the general population.
Risky and Problem Drinkers
Population-based studies revealed two new groups of drinkers. The first group engages in risky drinking (Table 1).37–43 This group includes drinkers who have not experienced alcohol-related harm, but are drinking at levels empirically shown to elevate their risks for experiencing harm in the future. The second group engages in problem drinking—levels of consumption associated with harm, but not meeting the diagnostic criteria for alcohol abuse or dependence. The pyramid in Figure 1 shows the full spectrum of alcohol use; areas in each category within the pyramid reflect the prevalence of each type of drinker. Before population studies were implemented, the size of the two new categories, risky and problem drinkers, was not understood; researchers were not studying the substance-use problems of these groups, and the treatment community was not addressing their needs. Even though, on average, the problems of individual risky or problem drinkers are less severe than the problems of harmful and dependent drinkers, the number of risky and problem drinkers is vastly larger and creates enormous social, legal, medical, and economic problems. Of all the alcohol-related problems seen in EDs and trauma centers, the majority of these problems are experienced by risky and problem drinkers—not addicted patients.
BROADENING THE BASE OF TREATMENT
In 1990, an Institute of Medicine report summarized the knowledge base in the field of alcohol treatment in a landmark tome called Broadening the Base of Treatment for Alcohol Problems.44 This report devotes a chapter to the issue of “what is being treated.” The authors consciously used “alcohol problems” in the title of their report to convey the importance of treating the full range of negative consequences associated with alcohol use, from risky drinking to addiction. “Broadening the base” in the title emphasizes the same message, the critical need to expand the focus of treatment from individuals with severe, chronic problems to include individuals with acute, intermittent, and mild-to-moderate problems. The report describes “the preventive paradox”: “If the alcohol problems experienced by the population are to be reduced significantly, the distribution of these problems in the population suggests that a principal focus of intervention should be on persons with mild or moderate problems.”44 Even though this statement does not exclude the treatment of addicted individuals as a legitimate concern, the statement is clearly paradoxical from the perspective of prevailing conceptual models, which tend to ignore individuals with mild and moderate problems almost completely. Therefore, a shift toward making these groups a principal focus of intervention activities is not just “broadening the base,” but also a recommendation for a major change in prevailing beliefs.
The report proposes a two-part treatment system. One part is the established specialist treatment system designed to meet the needs of alcohol-dependent individuals. The second part is a new proposal which can operate in a variety of community settings (medical, social service, and workplace). In the community-based part of the treatment system, nonspecialists screen individuals for alcohol problems and provide brief, onsite counseling to most who screen positive; first, to help them become aware of their problem, and second, to motivate them to change their behavior. The balance of screen positive individuals, a much smaller number, has severe problems, a prior history of dependence (addiction), or comorbidity such as liver damage or mental illness.45 This group also receives brief counseling, but the goal is to motivate them to enter the specialist treatment sector where the goal is typically abstinence.
Screening for the Problem
Accurate identification of patients with substance use problems is a required first step before help can be provided. Given the pervasive and subtle nature of the prevailing dispositional disease model, it is not surprising that physicians are predisposed to identify only patients with the most severe substance use problems. Research indicates, however, that physicians are unable to reliably identify patients with alcohol problems, even patients who are alcohol dependent. In one study,46 trauma surgeons, surgical house staff, and emergency department nurses who relied on clinical suspicion to identify intoxicated and alcohol-dependent trauma patients missed 23% of acutely intoxicated patients, almost 33% of severely injured, chemically paralyzed, or intubated patients, and more than 50% of patients with positive self-report screening tests—overall, a very poor showing. Specificity was poor too. Providers often suspected young males and disheveled, uninsured, or low-income patients of intoxication, and falsely identified 26% of such patients as alcoholic. Apparently, obvious signs of intoxication or disheveled appearance are not particularly reliable. An equally important result was the number of patients with alcohol problems who were missed. A statewide study of emergency departments in Tennessee estimated that 27% of adult ED patients need substance abuse treatment, but that only 1% will have a diagnosis of an alcohol- or drug-related problem in their charts.47
Part of the difficulty in identifying patients with alcohol problems is that the target is too narrow; it’s either alcoholism or intoxication. Additionally, a large proportion of substance use problems in trauma and ED patients, whether severe or moderate, are occult. Consequently, screening methods designed to identify patients across the complete severity spectrum are needed. For this reason, a critical element of the community-based part of the proposed treatment system is public health-style screening rather than case finding. Case finding depends on clinical suspicion to identify patients with problems; “suspect” patients are referred to specialists for assessment and diagnosis. Public-health screening applies a uniform, routine screening procedure to a predefined population. The goal is not to diagnose, but to measure level of risk for substance use problems and to initiate a response tailored to the individual’s risk level.
A variety of screening instruments for alcohol problems have been developed, evaluated in clinical settings, and used in routine practice. The Alcohol Use Disorders Identification Test (AUDIT, Table 2) is well studied, reliable, valid, and practical.48,49 It taps three domains–alcohol consumption, alcohol-related harm, and alcohol dependence symptoms–and evaluates the level of risk for alcohol problems. Unlike the CAGE Questionnaire,48 probably the most widely known brief screening instrument for alcohol dependence, each AUDIT question has multiple rather than dichotomous response categories, so the AUDIT provides not only a broader and more continuous measure of risk than the CAGE, but also enough personal information to help interventionists segue from the screening interaction into a brief intervention. Note also that the CAGE is typically used to identify individuals who are alcohol dependent; the AUDIT targets a much broader spectrum of alcohol problems and allows earlier intervention. One problem with the AUDIT is that it has 10 questions and may be too time consuming for effective use in acute care settings. However there are many different screening strategies, many of which are presented by Babor and Kadden in these proceedings.50
THE NEW TREATMENT MODEL
The Institute of Medicine’s report integrated decades of research into a new, carefully articulated treatment and nosologic paradigm, a blueprint for improving treatment across the spectrum of alcohol problems. However, the alcohol problems model is not yet familiar to the general public, the substance-use treatment industry, or physicians. In general, the public’s understanding is still too narrow, with the locus of the problem on addicted individuals. By introducing the concept of alcohol problems and by encouraging screening and brief intervention in community settings, the Institute’s report recast the prevailing conceptual model to include the whole population. Epidemiologists, particularly, have begun to realize that if society could, in the blink of an eye, miraculously “cure” its harmful and dependent drinkers, the “alcohol problem” would still not be solved. In fact, the job would not even be half finished.
Several problems are associated with the more narrowly defined dispositional disease model. The dispositional disease model is dichotomous–either an individual is an alcoholic or not. This dichotomy distracts us from acknowledging that risks associated with excessive drinking are distributed broadly and continuously throughout the population. By focusing only on alcoholics, we ignore the bulk of society’s alcohol problems. The dispositional disease model also obscures the knowledge that alcohol is more than a drug of addiction. It is a neurotoxin—a poison that, after repeated exposure, weakens and kills neurons. Edwards lists five ways in which alcohol is a poison and states that, “What we are seeing here is the capacity of a simple molecule to interfere with, or in some way hijack, the functioning of very complex brain systems.”2 The dispositional disease model completely ignores the importance of public policies that influence an individual’s alcohol consumption. For example, higher beverage alcohol taxes decrease consumption, even for the heaviest drinkers.7 Public policies that restrict access to alcohol–for example, laws that control the density of retail sales outlets and bars, and their hours, or laws that control the sale of alcohol in restaurants–influence when, where, and how much people drink. Another issue arises because the dispositional disease model focuses on addiction and for that reason, it appropriately promotes abstinence. However, focus on addiction detracts from the need to define and disseminate consumption guidelines for the general population.51,52
Despite being a major step forward, the two-part treatment system recommended in the Institute of Medicine’s report does not present a comprehensive public health model. Although the two-part treatment system moves from the dispositional disease model’s almost total focus on the host (the addicted drinker) to enlarge our understanding of the risky nature of the agent (alcohol) for drinkers who are not addicted, the focus is treatment not prevention. It does not describe prevention strategies, which are critical components of a comprehensive approach. The goals of prevention components would be to prevent excessive consumption among all drinkers—addicted or not. It would also highlight the importance of early intervention to prevent risky and problem drinkers from becoming harmful and dependent drinkers.
Although the papers presented at the 2003 conference and published in this issue of The Journal of Trauma address many issues regarding treatment and prevention of substance use disorders in trauma centers, some important questions remain unanswered or in need of clarification, such as, “Exactly what is a motivational intervention?”53; “How do we know which patients are ready to change?”54 and, “Would legalization of drugs solve drug-related problems in trauma centers or society?”55 After the conference, the editors commissioned papers (included in this supplement) in an effort to answer these questions. In addition seven recommendations, carefully edited and approved by the steering committee, emerged from the conference (see page S37–S42).
Society has contracted with trauma surgeons to care for its most severely injured patients, and trauma surgeons have responded by developing the most advanced system of trauma care in the world. The trauma patient population is at extremely high risk for alcohol and drug problems. In many instances, these substance-use problems are responsible, or partially responsible, for the events that precipitated the trauma center admission. These proceedings underscore the magnitude and importance of the opportunity that trauma centers provide in the treatment of these serious problems and help us understand why it is in society’s best interest to implement interventions in medical settings where the prevalence of substance use problems is highest. However, it will take more than trauma surgeons and their professional organizations to knock down policy, legal, professional, financial, and knowledge barriers. Trauma surgeons will need help from substance use researchers, hospital administrators, insurance companies, advocates, and policy makers. Decision makers in federal agencies and foundations must also pay greater attention to this opportunity.
Trauma surgeons, emergency physicians, and nurses who treat patients in acute care settings have a difficult job. They treat society’s most acute and complicated medical problems 24 hours a day, 7 days a week, and realize that alcohol and drug problems are inextricably linked with their daily work. Many voluntarily accept responsibility for addressing these problems. In the recommendations presented in these proceedings, conference participants–mostly trauma surgeons–point the way forward for interventions in trauma centers. The rest of us—those of us whose daily work is outside the trauma center–must accept our responsibility to help address the substance use problems heaped at the trauma center door.