Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Theories and Principles of the Dissemination and Diffusion

Information is transmitted within any given population through two general pheno­mena: dissemination and diffusion. The former refers to any purposeful action in which new information is introduced to members of that population, whereas the latter refers to the non-deliberate process by which members of a population accept or reject disseminated information (Buller 2006). In this section, diffusion is described first as a foundation for understanding dissemination, and in this context the social marketing approach is borrowed and modified as a suggested strategy for maximally efficient dissemination of scientific work (Dearing et al. 2006). Following this discussion, we describe practical approaches to enhance the likelihood of successful dissemination.

1.1 Diffusion of Innovations Theory and the Social Marketing Approach

Perhaps the most important advance in the understanding of human behavior change on a societal level – as opposed to behavioral change in an individual – occurred in 1962 when Everett Rogers published his groundbreaking book Diffusion of Innovations. This theory of social change is presented only in a brief summary format here, but those interested in this topic are referred to Rogers’ book (Rogers 2003) and to Malcolm Gladwell’s The Tipping Point for formal and popularized discussions, respectively. Terms and concepts relevant to this chapter are defined in Panel 30.1.

Fundamental to the Diffusion of Innovations theory is the idea that society has inherent albeit dynamic channels of communication through which a message is transmitted from one entity in society to another. When an individual, group, or organization adopts an innovation – which may take the form of a new idea, behavior, practice, or object – others in society are directly or indirectly exposed to that innovation via one or more communication channels. These individuals may or may not be persuaded to adopt the innovation. The decision of whether to adopt an innovation can be optional, like trying the latest diet; collective, such as a professional association’s approval of clinical guidelines by consensus; or authority-driven, such as laws that ban dumping waste in public water supplies. Early adopters interface with those who have remained undecided about whether to adopt the innovation themselves, and some will perpetuate this cycle of adoption in what’s known as the multiplier or snowball effect (Fig. 30.1). After a party decides to adopt an innovation, the individual makes further decisions about whether to continue or discontinue use of the innovation and how. Ultimately, the steps in diffusion depend on social, ecologic, and systematic factors in society; therefore, dissemination strategies will differ at the individual, organizational, and network levels (Estabrooks and Glasgow 2006; See: RE-AIM Framework section).

Fig. 30.1
figure 1

Dissemination and diffusion of information. Dissemination refers to any purposeful action intended to introduce new information to society. The dissemination process may be highly variable and is represented by the blue square (Step 1). All individuals in society are inherently uncertain regarding whether to adopt never-before encountered information, a state of mind represented by a blue cloud with a question mark. Exposure to new information prompts a decision of whether or not to adopt that information (represented by the conversion of a blue with a question mark to a green cloud with a check mark), a process that depends on the removal of uncertainty (e.g., by logic or incentive). Among the early adopters are “opinion leaders,” who have the capability of interfacing with (Step 2) and influencing others in society to imitate the opinion leader or adopt the information (Step 3). Some of these individuals will be opinion leaders themselves and propagate the pattern just described, a scenario known as the multiplier effect (Step 4). A consequence of the multiplier effect is snowballing imitation and adoption (Step 5), a phase of diffusion that represents the greatest change in the number of imitators and adopters. Note that, although this figure depicts the opinion leader and others as people, this representation is not intended to be exclusive of other entities in society, such as organizations

Full size image

Insight into diffusion dynamics allows for an understanding of how new information, once introduced to society, succeeds or fails to spread amongst its ­constituents. Social marketing is the practice in which opinion leaders – who may be socially influential people, peer-educators, celebrities, authoritative organizations, etc. – are identified and made the point of dissemination. These opinion leaders then leverage the diffusion process by influencing many individuals, some of whom will be parti­cularly socially influential among peers, to adopt an innovation (Fig. 30.1). Dearing et al. (2006) suggests that social marketing principles can be employed for the implementation of campaigns to increase physical activity, the implication of which is that epidemiologists in general can use a social ­marketing-like approach to disseminate research findings. In this context, the epidemiologist’s innovation is typically new information about the benefits or harms of a particular health-related behavior or phenomenon.

2 Practical Considerations for Dissemination Strategies

It is important to consider the dissemination strategy in detail because the entity that or person who disseminates the innovation may influence the success of diffusion. The dissemination of the Behavioral Risk Factor Surveillance System (BRFSS), a low-cost health survey system originally developed in the United States in the 1980’s, is an example of a well-designed, well-controlled dissemination process (Bauman et al. 2006). The BRFSS was launched in 15 states in 1984 and, over the course of a decade, was adopted by the other 35 states and the District of Colombia; by 1998, some US territories in the Caribbean also adopted the BRFSS (Nelson et al. 1998). After successful implementation in the United States, other countries – including Mexico, China, and Russia – made culturally appropriate revisions before deploying the system themselves (McQueen and Puska 2003). However, one should be aware that dissemination and diffusion are highly variable processes and may or may not go according to plan.

The experience of the International Physical Activity Questionnaires (IPAQ), perhaps the most widely used survey tools for the surveillance of physical activity, shows us that dissemination processes are not always so structured, even when dissemination is successful. The IPAQ investigators had planned further research and development of IPAQ before promoting widespread use of the tool; however, the need for the instrument was so strong that research groups throughout the world adopted IPAQ rapidly and early (a more detailed history is found in Bauman et al. 2006). The IPAQ story suggests that some innovations are so needed or desired that the simplest form of dissemination (e.g., publication of a single journal article) is sufficient to trigger widespread diffusion, but such instances are rare and cannot be expected.

2.1 Six-Step Model to Enhance Dissemination of Information

Although there is not just one approach to planning a dissemination strategy, Bauman et al. (2006) proposed a six-step model to enhance dissemination. This model, shown in italics below, is annotated here with questions that may be useful to consider when planning a dissemination strategy. Some of these questions are derived from the RE-AIM Planning Tool (discussed below), whereas others are suggested by the authors. Table 30.1 also provides a framework for thinking about possible dissemination strategies. Collectively, the six-step model, RE-AIM Framework, and Table 30.1 should be helpful in selecting the processes that balance ideality, practicality, and cultural appropriateness.

Table 30.1 Examples of dissemination-based intervention strategies reported for five cancer-­related topics between 1980 and 2004a
Full size table

2.1.1 Step-1: Describe the Innovation and its Rationale, Evidence Base, and International Contexts

  • In full detail and also in just 1–2 sentences, what is the innovation?

  • Why is the innovation important?

  • What evidence suggests that this innovation will succeed?

  • Under what sociodemographic contexts were efficacy and effectiveness studied, and how do these compare to those of target audiences?

For dissemination, having information about both efficacy and effectiveness of the dissemination strategy is ideal, indeed sometimes necessary, but information on effectiveness is usually unavailable. Lack of effectiveness data is especially common when attempting to disseminate innovations to international or underserved populations; useful advice for these situations is available in Cuijpers et al. 2005 and Yancey et al. 2006, respectively.

2.1.2 Step-2: Characterize the Dissemination Strategy

  • Identify the target audience for dissemination and its size.

    • Who should adopt the innovation and why?

    • Who/what are the relevant opinion leaders?

    • Do you hope to reach all members of the target population?

    • What is the sociodemographic breakdown of the target population?

  • Anticipate and describe the sequence, timing, and format of the dissemination strategy. (This may require simultaneous or prior consideration of Step 3.)

    • How might characteristics of relevant opinion leaders influence dissemination?

    • How might characteristics of the target population influence the dissemination strategy?

2.1.3 Step-3: Define the Current Communication Channels Through Which Diffusion Might Take Place

These channels may occur within or across multiple levels of society:

  • Who do the relevant opinion leaders interact or affiliate with?

  • How does the target audience perceive the identified opinion leaders?

  • Are any relevant changes in the communication channels foreseen?

2.1.4 Step-4: Determine the Role of Decision-Makers and Partners that will be Necessary for Dissemination at Various Levels of Society (e.g., Local, National, International, etc.)

  • Which societal sectors, industries, organizations, or government bodies are particularly important for the planned dissemination strategy?

  • What are the political structures in the relevant areas?

  • Who in your social network has direct or indirect connections with identified decision-makers and partners?

2.1.5 Step-5: Identify Factors that Might Impair or Facilitate the Dissemination Strategy

  • Which cross-cultural or political factors might influence the route of dissemination in different locations and how?

  • What barriers might limit your reach to the target population, and how do you plan to overcome them?

2.1.6 Step-6: Create an a priori Plan for Evaluating the Dissemination Process (May Not Be Necessary Depending on the Goals of Dissemination)

  • How will dissemination and subsequent diffusion be assessed?

  • What resources will be required to conduct evaluations?

  • Are the stakeholders in agreement with the evaluation plan?

2.2 The RE-AIM Dissemination Framework

The RE-AIM (Reach, Effectiveness, Adoption, Implementation, Maintenance) framework was proposed by Glasgow et al. (1999) as a practical tool to enhance dissemination. Although championed by the National Cancer Institute (NCI) of the U.S. National Institute’s of Health (NIH), RE-AIM can be used to plan dissemination strategies targeting a range of health issues, not just cancer. Various helpful RE-AIM resources can be downloaded from the NCI’s Cancer Control website (See: Sect. 30.6). Perhaps most notable are the RE-AIM Planning Tool and the RE-AIM Checklist for Study or Intervention Planning, both of which we recommend. To exemplify the utility of the RE-AIM framework, complex dissemination strategies are presented in a generalized form at three levels of society – an individual, an organization/specific location, and a network/population – in Table 30.2.

Table 30.2 Examples of general dissemination-related interventions aimed at various levels of society using the RE-AIM frameworka
Full size table

3 Communicating with the Public and News Media

In order to get the attention of a target constituency – those whom the researcher intends to influence, such as the public, policymakers, or other stakeholders – one must inform the members of that constituency about the innovation. Various routes are available, of which we explore two especially prominent ones: conventional media interviews and the emerging venues provided by online social networks. Reaching out to journalists, media broadcasters (e.g., radio and television), and media relations specialists is uniquely challenging and requires summarizing findings in ways that reflect the current state of the science, not just an isolated conclusion. Advice on interacting with journalists and preparing for interviews is therefore provided below. Much of this information is readily translatable to communicating with the public via social networks, a venue that has enabled companies to interact with and learn about their customers and that should, by extension, be useful for epidemiologists as well.

3.1 Routes of Communication and Getting a Message Across

Communicating scientific health information to the public is especially difficult, as most members will not be well-versed in the subject at hand. When communicating with the public, a researcher can select one or several routes, the best of which will depend on the researcher’s goals of dissemination and characteristics of the target audience. Usually, there is at least one intermediary (e.g., a journalist) but direct communication is increasingly possible with technological innovation (e.g., through blogs and social networking websites). Panel 30.2 lists several common routes of communication, but this list should not be considered all-inclusive, nor should discussion of selected routes be interpreted as an indication of their importance.

How one delivers a message can influence whether or not the public adopts that message (most often through changes in understanding and behavior). Since that message is often relayed through an intermediary, the importance of proper delivery is heightened. Central tenets of effective communication are elucidated below through a discussion of preparing for an interview with a reporter, one of the most common intermediaries that epidemiologists encounter. These tenets may be extrapolated to other scenarios in which a researcher might need to deliver a message to a lay audience.

3.2 Preparing for an Interview with a Journalist

Perhaps the most critical exercise in preparing for an interview with the press is to iden­tify the Single Overriding Communication Objective (SOCO), i.e., the single most important message to deliver about a study or health issue (Dan 2008; See alsoPanel 30.3). Typically, a research project will produce several interesting findings, and although discussing them all is tempting, doing so is almost always counterproductive. For example, if the researcher provides two main findings, then constraints of space and time may force the reporter to select one of the two and thereby miss the main thrust of the work. Simplicity is not often a hallmark of research findings, but it is an imperative for successful communication of one’s findings and interpretations to the public. Consequently, the SOCO must also be rid of jargon wherever possible. Should jargon be necessary, it is useful to be aware of the jargon in advance of an interview and to be prepared to define terms clearly. These two difficult tasks – to find simplicity within complexity and to translate scientific terms into commonly understood language, all without introducing inaccuracies – are critical and therefore deserve pause and consideration before beginning an interview.

One should realize that the job of the reporter is to interpret a story and then to convey it to a broader audience, not to be up-to-date on the scientific field relevant to the interview topic. It is therefore necessary for the researcher to help the reporter understand the essential background information, the purpose of the research, the one main finding, and any major ambiguities. Some researchers insist that the reporter read the related scientific paper before the interview, although the utility of that strategy is dubious to us because it probably has low yield, blurs the line of who is the epidemiologist, and risks alienating the reporter (reading a research paper can be difficult and time consuming even for an epidemiologist!). Alternatively, we suggest providing the reporter with a copy of the paper without insistence and to offer to serve as a resource for them should they have any questions in preparation for the interview.

3.3 During the Media Interview

During an interview – which may occur via e-mail, in online discussion forums, over the phone, or in person – the interviewee’s goal is to turn the interview into an opportunity to relay the SOCO (Dan 2008). An imperative is to communicate the SOCO early and repetitively. If the reporter attempts to discuss a point that deviates too far from the SOCO, one commonly employed approach is to transition back to the SOCO using phrases such as ‘the most important thing to say about that is…,’ ‘the bottom line in those situations is…,’ or ‘the take-home message is….’ This technique is called bridging. Importantly, the phrases used in bridging are also useful for flagging the SOCO and thereby signaling to the interviewer the importance of what will be said next.

To help the reporter understand the context and importance of one’s findings, the researcher will need to convey specific epidemiological terms and statistics that might have multiple or alternative definitions in common parlance or be difficult to understand. For example, the word risk to an epidemiologist signifies a concrete concept defined by an objective numerical value, but to the public risk may be an abstract concept implying a subjective degree of danger or hazard (Loukissas 2011). Given the potential for the reporter or reader to misinterpret the specific meaning of a term or statistic, it is imperative for the researcher not only to select carefully which statistics to discuss but also to phrase the terms associated with those statistics in the clearest way possible.

Typically, an interview will contain the following six basic questions (Loukissas 2011):

  1. 1.

    Why did you do the study?

  2. 2.

    What is the one main result of the study?

  3. 3.

    What challenges did you encounter and/or overcome?

  4. 4.

    What mechanism explains the association?

  5. 5.

    What is the public health message?

  6. 6.

    Where should future research focus?

If these questions are not asked directly, the researcher should attempt as naturally as possible to provide the answers during conversation, or give verbal clues that cause the reporter to ask these questions. It is acceptable for the researcher or reporter to politely redirect the interview to the most important content. Indeed, both parties control the content of the interview. Reporters do so mainly by asking questions, and interviewees do mainly by regulating how questions are answered and for how long.

Other common questions that the researcher should be prepared to answer are:

  1. 7.

    Did anything about the study surprise you?

  2. 8.

    Is there anything you would like to add?

  3. 9.

    Is there anything I did not ask you?

Other useful advice for participating in an interview can be found in summary format in Panel 30.4 (Dan 2008; Loukissas 2011). We refer readers to Dan (2008) for useful advice on how to deal specifically with press conferences and interviews over the radio/podcast or television.

Hint

Either at the beginning or end of the interview, it is common for the interviewee to request the opportunity to check quotes or even read the media article before publication. Some reporters will be accommodating, but this practice should not be expected, nor is it always possible.

3.4 Online Social Networks

Increasingly, social networks have become an outlet for disseminating health policy ideas and objectives. Social networks are a loosely defined group of websites through which users share and disseminate information. They include websites where users set up social profiles (e.g., Facebook), post public updates, follow other users’ updates (e.g., Twitter), review local businesses (Yelp), post their current location and activity (e.g., Foursquare), rate content that they enjoy (e.g., Digg), or post videos and photos (e.g., Youtube and Flickr). Many of these websites overlap in functionality, and the available services are constantly evolving.

In 2006, Facebook began allowing organizations to create profiles. The creation of a profile on a social networking site can achieve many goals. Some of the uses that social network profiles can achieve include:

  • Identifying and gathering contact information for people interested in the organization’s product or policy goals

  • Providing information, including links to pertinent news articles and websites

  • Publicizing upcoming events

  • Recruiting volunteers

  • Providing a forum for discussion and feedback

  • Soliciting donations

A content analysis of 275 nonprofit organizations’ Facebook profiles found that the profiles of nonprofits tend to provide transparent explanations of their purpose. But many of them failed to take advantage of the social aspect of Facebook, rarely offering users opportunities to get involved (Waters et. al. 2009).

Furthermore, social networking can afford an opportunity to identify topics of discussion and possible questions, concerns, or points of intervention for organizations that wish to disseminate particular information. Many for-profit companies regularly monitor mentions of their products on Twitter in order to identify dissatisfied customers and address their concerns. This opportunity exists for more public service-oriented organizations as well. A recent study on the use of the word “antibiotics” on Twitter found hundreds of tweets that contained misinformation or discussed inappropriate uses of antibiotics (Scanfield et. al. 2010). These users could serve as intervention points for organizations searching to publicize a public health message, for instance.

3.5 Making Change in the Real World

It is helpful for the researcher to realize that information, in the traditional sense, is not the only essential prerequisite for action. Health information is often difficult to understand without substantial background. Epidemiologists are trained to think about public health issues from a highly technical, methodological, and intellectual perspective. It is therefore most natural for many epidemiologists to communicate about their work in such terms. However, most people rely on emotions, feelings, instincts, and heuristics to guide their actions. We often form judgments by subconsciously asking: “How do I feel about this?” (Oz 2010). When communicating with the press or public, one should keep this in mind and cater to the strengths of one’s target audience, not to those of one’s colleagues.

4 Communicating with Study Participants and Their Healthcare Providers

The most important stakeholders of any study are its participants. Study findings should be reported to them, perhaps employing the strategies discussed above regarding interviews with reporters. But sometimes the researcher must communicate with particular patients about new personal health information. This task requires a great deal of care and sensitivity.

4.1 New Personal Health Information

Researchers often acquire information on participants’ personal health that was previously unknown to the participants and their health care providers. The new information may be of a diagnostic or prognostic nature. It is standard practice to provide such information to the participant’s indicated primary health care provider, although participants should be given the option to bar this action (unless laws require otherwise). The new information can be quite sensitive, such as the risk of having, acquiring, or transmitting a strongly heritable illness or a potentially stigmatizing illness. Indeed, not all participants want to know if they have certain diseases or disorders. One option is to let the participants decide, as part of the informed consent process, whether they wish to be informed about new personal health information and in which way (White et al. 2008). However, there are rare instances in which laws mandate reporting of some diagnoses to individuals and/or agencies; therefore, it is necessary to identify and follow such laws and to explain these laws to participants during the informed consent process.

Moreover, new personal health information can also be of an acute nature that requires immediate intervention. Each study should have a system of referral and/or treatment of new health issues that may be discovered during the study. These ­provisions relate mainly to anticipated problems such as known side-effects of treatments or measurements, or possible values of health measurements (e.g., a positive HIV test).

4.2 Communicating Overall or Interim Study Results to Participants

To facilitate the communication of overall or interim study results to participants, one may use the strategies for communicating with the public described previously in this chapter. Common strategies for reporting study results are to send a newsletter or to host a special post-study gathering. In doing so, one should be careful to avoid creating anxieties that are out of proportion to the size of the risk (White et al. 2008), and to never reveal personal health information (a breach of confidentiality).

5 Striving for a Desirable Influence on Public Health Policy

Policymakers consist of administrators and elected officials who make decisions that affect laws, rules, and regulations. Ideally, they respond to health issues that have the greatest importance for public health, but in reality, politics often determine the salience of the issues at hand. Epidemiologists have the opportunity to interface with policymakers in order to increase the likelihood that they will effect change on the most important health issues and to intervene in a manner that is supported by empirical evidence. Indeed, epidemiologists can raise awareness of an issue and potentially change policymaker’s priorities.

This discussion raises an important question: to what extent are epidemiologists responsible for ensuring that their research results are ‘translated’ into policy? The answer to this question is not clear-cut and comes with many caveats. One obvious element of the answer is that any such responsibility should be limited by the internal and external validity of the research. ‘Translation’ should ideally not be attempted unless research evidence has turned into scientific knowledge. Scientific knowledge is a majority consensus phenomenon, so health policy decisions need to be informed by expert consensus on scientific issues. In reality, however, the experts often disagree on whether a sufficient basis for action has been reached, on what more evidence is needed, and minority views – whether they are right or wrong – can be influential.

Of course, scientists should be concerned about the consequences of their activities and findings, and therefore communication with policy makers is unavoi­dable. Viewed from the other side, health policy decisions need to be based on a rational system for prioritizing competing health intervention needs, and policy makers are heavily reliant on experts when it comes to understanding a body of research. Creating and maintaining active channels of communication between policy makers and the scientific community is therefore a dual responsibility of the involved ­parties. Scientific experts should be involved in health policy decision-making, but such involvement is contingent on relationships. Dissemination represents an important approach by which researchers can form and build relationships with policy makers. In creating communication channels, researchers should avoid the temptation to increase the impact of their work by becoming policy makers themselves, or by becoming activists who try to pressure public health authorities and politicians in a somewhat desperate way.

Efficient structures of communication between policy makers and the scientific community are needed for giving policies an appropriate evidence base and to avoid that influencing policy becomes merely a matter of who shouts the loudest, talks the smoothest, or has the most political friends. Bounded rationality – a term coined by Herb Simon – is the idea that, in decision making, the rationality of individuals is limited by the information they have at hand, the cognitive limitations of their minds, and the finite amount of time they have to make decisions. Given the complexities of health policy, it could be said that a consulting epidemiologist’s role is to reduce bounded rationality by efficiently communicating the most important information.

5.1 The Right Messages at the Right Time

It is relatively uncommon for a single research study to have any direct positive impact on public health. Conclusions based on just one study are the most susceptible to bias – although a conclusion based on no evidence at all is markedly less reliable unless that conclusion is agnostic – and should be promoted with caution. Without training in epidemiology, those who are exposed to a researcher’s SOCO are much less likely to understand the nuances of object design and statistics than a researcher’s fellow epidemiologists and are, therefore, usually unable to fully appraise a single study’s internal and external validity.

This problem is especially apparent when multiple studies on a similar topic yield seemingly different conclusions. For instance, if Study A and Study B represent the total available evidence on whether exposure to a substance is associated with heart failure, where Study A shows a 20 % increase and Study B shows a 20 % decrease in the risk of developing heart failure, then a group of people might reasonably conclude that there is, on balance, no association between the substance and heart failure. Yet, the two studies would likely suffer from differing degrees of bias, confounding, and other flaws that might make one study less valid than another. These complex features are usually omitted in media coverage of scientific research, thereby making the public vulnerable to misinterpretation of potentially useful health information. If Study A is more valid than Study B, then the public may not recognize the substance’s potential harm.

Even if the internal validity of the study is optimal, there are known instances of authors incorrectly placing and interpreting the evidence in the light of external evidence on the same topic. The external evidence may only be partially mentioned and the author’s own interpretation of the importance of their study may be ­exaggerated. Various stakeholders may then ‘jump on’ the most categorical and striking statements in a discussion section or in the conclusions of a scientific paper. Take away messages (Single Overriding Communication Objectives) can be dangerous if they are distorting simplifications. The result may then well be another wave of unfounded health-related anxieties and another contribution to over-medicalization of society. It is our view that the dissemination of research findings charges the researcher with the responsibility to attempt to protect the public from misinterpretation of potentially useful health information. This may occasionally imply restraining from dissemination of findings of single studies.

The same logic just described also applies to the dissemination of intervention campaigns. In such a scenario, however, there is usually a body of evidence supporting the efficacy or effectiveness of the intervention and sometimes even of the dissemination strategy. Selecting the best estimates to convey to the target audience is a challenging task, especially when multiple studies on a similar topic provide interpretations as dissimilar as Studies A and B above. Meta-analyses may be useful for selecting a best estimate but must be interpreted very carefully and always in the context of possible publication bias (See: Chap. 25).

5.2 Advice for Dissemination Aiming at Public Health Impact

Taking into account the above considerations, the appropriate combination of scientific and ethical concerns translates into the advice list for dissemination of research results in Panel 30.5.

6 Additional Dissemination Resources

This chapter provides several different frameworks to be used as tools for the deve­lopment of dissemination strategies. Though the body of literature on dissemination strategies is relatively sparse and usually unavailable for or irrelevant to a specific dissemination project, some of that body of literature and a wealth of experience have been embedded in many helpful resources. We recommend the following resources for those who wish to extend and deepen their knowledge of dissemination strategies:

In this chapter we discussed theories about how information spreads among groups of people and society, and we leveraged those theories to advise on the development of a dissemination strategy. Dissemination to stakeholders (Chaps. 29 and 30) raises some ethical issues with study reporting. Some of these have already been mentioned, but several major issues around publication/authorship policy and publication bias remain to be discussed in the final chapter.