Ethical Conundrums in Fatality Review Planning,Data Collection,and Reporting

Introduction

Since the late 1990s, communities throughout the United States have used the practice of fatality review to help them assess prevention and intervention strategies for domestic violence. Fatality review teams are multidisciplinary, interprofessional groups assembled to conduct case-level evaluations of preventable deaths. The goal of this practice is to identify ways to prevent similar circumstances from leading to injury or death in the future. In addition to domestic-violence-related deaths (see Websdale, 2003, 2012; Websdale, Town, & Johnson, 1999; Wilson & Websdale, 2006), fatality review has been used to investigate deaths under a variety of circumstances, including hospital deaths (see Wright et al., 2006; Zimmerman et al., 2010), occupational deaths (see Bugeja, Ibrahim, & Brodie, 2010), pregnancy-associated and maternal deaths (see Burch, Noell, Hill, & Delke, 2012; Kalter, Salgado, Babille, Koffi, & Black, 2011), child fatalities (see Christian & Sege, 2010; Dufree & Tilton-Dufree, 1995; Dufree, Tilton-Dufree, & West, 2002; National Center for Child Death Review [NCCDR], 2005), and deaths due to elder abuse and neglect (see Steigel, 2005).

There is limited research on fatality review as a method of inquiry in general and even less on the associated ethical issues. To the extent that the literature engages ethical issues related to fatality review, these discussions focus mainly on maintaining confidentiality in the review process (Hobart, 2005; NCCDR, 2005; Thompson, 2002; Websdale, 2003; Websdale, Sheeran, & Johnson, 1998; Wilson & Websdale, 2006; Zollo, 2008). While this is a key issue for fatality review, the process forces practitioners to wrestle with an extensive range of ethical conundrums. We define ethical conundrums as dilemmas that arise when individuals negotiate competing values to make decisions about a group’s best course of action. Ethical frameworks, with associated formal and informal rules, exist at group and individual levels across a range of contexts and domains. However, there is no concrete ethical framework governing the work of fatality review teams. Each stage of the review process presents a unique set of potentially contested decision-making moments. In the absence of a specific set of ethical principles and rules, teams often have to confront these decisions in ad hoc and largely reactive ways.

The diverse make up of fatality review teams further complicates efforts to confront ethical issues. Fatality review teams are typically composed of a wide range of professionals, academics, and community stakeholders who bring with them the established ethical guidelines from their respective professions. These include frameworks for those working in research, counseling, medicine, and the practice of law, to name a few. Interprofessional work introduces ethical issues that go beyond confidentiality and transcend any one profession (Clark, Cott, & Drinka, 2007). As such, the frameworks of professional ethics may or may not sufficiently guide the process of fatality review and, in some cases, may be in conflict with one another or the perceived mission of the team.

Here, we use our own experiences with these issues to offer a framework that can help teams responding to domestic violence fatalities navigate the unique ethical challenges they confront. The authors are members of a domestic violence fatality review team that has been operating since the late 1990s. Our team originated in a university as a research protocol. Our practice borrowed elements from the model for investigating child deaths developed by Michael Dufree and colleagues at the Los Angeles County Inter-Agency Council on Child Abuse and Neglect (see Dufree et al., 2002; Dufree & Tilton-Dufree, 1995). As the principal investigator, team coordinator, criminologist, emergency medicine physician, and director of a local legal aid domestic violence unit, we bring a range of experience and perspectives to the table. All of us have served on the fatality review team for multiple years, some since the team’s inception. Over the years, we have developed and refined all aspects of our case review strategy. Many of these revisions are the result of ethical conundrums the team has confronted over time.

The current paper represents an effort to apply an ethics-based framework that may help us identify and mitigate a range of ethical issues that can hamper the domestic violence fatality review process. We propose that the AEA Guiding Principles for program evaluation can be a useful tool for evaluating ethical issues in fatality review (AEA, 2004). We do not claim that program evaluation is the only framework through which we can or should view ethical concerns in fatality review. However, we see these principles as a helpful way to begin extending our identification and understanding of, as well as our response to, the potential ethical conundrums associated with planning, data collection, case review, and the reporting of findings from the fatality review process.

In the sections that follow, we briefly describe the application of the fatality review method to the evaluation of domestic violence fatalities. Following this discussion, we examine the practice of fatality review through the framework provided by the AEA Guiding Principles of Evaluation and identify potential ethical conundrums that can result from decision-making points in the fatality review process. We conclude with a summary of ethical concerns, recommendations for ensuring ethical practice, and the identification of potential conundrums that require further study.

Using Fatality Review to Evaluate Deaths Attributable to Domestic Violence

Generically, the fatality review process involves identifying and selecting cases for review, collecting data relevant to each case, reviewing case information by an individual or team with personal expertise in the area, and documenting participant feedback on a set of questions related to the purpose of the inquiry. This process also involves the formulation of recommendations for changes to existing policies or practices, which are intended to mitigate future adverse events.

An examination of published papers, team reports, and best practice manuals reveals commonalities in the practice of fatality review across substantive areas. Similarities include guidance on and options for setting up a team, conducting case reviews, reporting findings, and generating recommendations (see Bugeja et al., 2010; NCCDR, 2005; Steigel, 2005; Wright et al., 2006). This is not surprising, given that the goals of fatality review are similar across the types of preventable death, and the teams involve many of the same community stakeholders.

Efforts to provide guidance to domestic violence fatality review teams began relatively early in the application of the method to domestic-violence-related deaths. The earliest known reviews took place in the early 1990s. By 1998, the Family Violence Department of the National Council of Juvenile and Family Court Judges joined with other agencies engaged in the practice to hold a summit to examine best practices and create recommendations for forming review teams and conducting case reviews. McHardy and Hofford (1999) documented the resulting recommendations in a summary report. In that same year, the U.S. Office of Justice Programs Violence Against Women Office funded the National Domestic Violence Fatality Review Initiative (NDVFRI), which serves as a clearinghouse for team information, archives team publications, and offers technical assistance to domestic violence fatality review teams (NDVFRI, n.d.; Wilson & Websdale, 2006). The resulting documents and the information available on the NDVFRI website show that the use of this method by domestic violence fatality review teams has become more similar over time.

The application of the fatality review method to domestic violence homicide is based on four shared underlying assumptions: (a) No single professional group or agency can end domestic violence, (b) domestic violence deaths may be prevented through reasonable interventions, (c) the domestic violence perpetrator is responsible for the homicide, and (d) the system of professional agencies tasked with domestic violence intervention are accountable for ensuring that their practices lead to increased safety and welfare of victims and the communities they serve (Wilson & Websdale, 2006). These shared assumptions lead to key similarities, particularly the interdisciplinary nature of review teams and the team goals of victim safety and homicide prevention. However, the practical details of the review process also vary as a function of specific team objectives, structure, and composition. We contend that differences in team organization and methods of inquiry reinforce the importance of broad ethical guidelines that teams can adapt to their unique needs.

Fatality Review as Evaluation

Evaluation refers to the systematic assessment of policies or programs against a set of understood and typically documented standards and expectations (Weiss, 1998; Wholey, Hatry, & Newcomer, 1994). Evaluation aims to assess and identify ways to improve program functioning and outcomes. These goals equate to the two common types of evaluation: outcome and process. Outcome evaluations answer the question, “Are we getting the desired result?” Process evaluations answer the question, “Is the policy or program functioning according to plan?” Process evaluation may also yield an assessment of outcomes but is more focused on the operation of the program.

The activities of evaluation are similar to those of other types of research (Weiss, 1998). Generally, the evaluator begins with a set of questions derived from the objectives, procedures, and expected outcomes of a policy or program. The evaluator then collects information on program implementation and outcomes, including relevant details of individuals involved in the implementation and the individuals, groups, or communities that serve as either the target or control group for the program under study. With the relevant data in hand, the evaluator then analyzes the information and prepares a report of findings. The agency or program under evaluation uses the findings as feedback for decision making around issues related to program continuation and improvement.

Fatality review shares a number of commonalities with the field of program evaluation and its ethical considerations. It is most similar to process evaluation. Teams evaluate deaths in the context of a prevention system with the goal of identifying intervention points where the system is not functioning well and formulating potential solutions. Fatality review teams collect information from the systems that have the most involvement in the case. For domestic violence homicide reviews, data collection may include documents from the police department and interviews with responding officers for the homicide and previous calls for service, court documents related to prior protection orders or custody proceedings, information available from the prosecuting attorney, and/or other known systems of which case parties were in contact. Teams analyze the information and identify system failures or areas in need of improvement. The nature of these activities (similarities in purpose, objectives, data collection, and analysis) are such that we expect to find a crossover between the ethical dilemmas faced by program evaluators and those encountered during the process of fatality review.

There are a number of ethical frameworks that could be applied to the practice of fatality review. Domestic violence fatality review is situated in the practice of community prevention and intervention services, civil and criminal justice system interventions, and public health interventions. Within these fields, practitioners provide services to domestic violence victims and offenders in a manner consistent with the ethical principles of their respective professions. Examples of relevant ethical frameworks may include, but are not limited to, principles related to the practice of public health (Public Health Leadership Institute, 2002) or social work (National Association of Social Workers, 2008). However, the involvement of these same practitioners in domestic violence fatality reviews changes their role to one of evaluator and, at times, subject. The study of deceased persons is not considered human subjects research. However, as noted above, fatality review also involves research activities that may bring the practice into the realm of biomedical and behavioral research ethics detailed in the Belmont Report (see National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979). Fatality review findings and recommendations are intended to inform the policies and practices of agencies that provide services to domestic violence victims and offenders. These findings may be aggregated over time to assess patterns related to domestic violence homicide or access to service delivery systems. The goal of this practice is not to provide generalizable data but rather to provide timely input on policy and practice in community services.

Ethics in Evaluation

In 2004, the AEA ratified a set of five guiding principles to align and add clarity to the varied purposes and techniques of the profession of evaluation. The five principles are (a) systematic inquiry, (b) competence, (c) integrity/honesty, (d) respect for people, and (e) responsibilities for general and public welfare (AEA, 2004). The need for principles, specifically for the practice of evaluation, as distinct from general research principles, stems in part from the fact that evaluation often occurs outside of the research profession. In addition, the information generated by an evaluation is often designed to meet a specific operational need rather than to produce generalizable knowledge or test a new technology.

The adoption of guiding principles or “evaluation ethics” is not intended to constrain evaluators or dictate their practice but rather to provide a uniform set of standards for addressing ethical issues that arise from program evaluation activities. Adherence to this broad set of ethical guidelines should also ensure quality evaluation practices and outcomes. The AEA expands on the principles for evaluation professionals, provides examples in various online and printed materials, and offers training on the implementation of the principles. In addition, there is an emerging literature related to the adherence to these principles during the conduct of evaluation practice (for a review of this work, see Morris, 2011).

Applying the Principles of Evaluation to Fatality Review

The purpose of the following discussion is to open a dialog on ethical conundrums in the practice of fatality review. As we have shown, the products of this work are the culmination of numerous decisions made by funding agencies, staff, and team members. These decisions affect not only the practice of fatality review but also individual participants, stakeholder agencies, and the broader community. We examine components of the fatality review process through the lens of the AEA Guiding Principles of Evaluation to identify ethical conundrums that the review teams may confront. We also aim to show how these principles can help guide responses to a range of ethical concerns. Where possible, we cite published literature on fatality review as a method of inquiry. However, the concerns we focus on and the strategies we offer in large part stem from our collective experience as active, long-standing members of a successful domestic violence death review team. In the following sections, we define each principle, identify areas of the fatality review process that correspond to similar areas of evaluation, and summarize some of the ways in which failing to acknowledge the ethical implications of decision making in fatality review can hinder the process and its intended outcomes.

Systematic Inquiry

The first principle calls for evaluators to conduct systematic, data-based inquiries (AEA, 2004). Systematic inquiry guards against problems in data collection that stem from an evaluator’s decisions regarding the design and execution of an evaluation, and helps lay the groundwork for producing conclusions that are connected to empirical observations (Morris, 2011). This principle applies particularly to fatality review case selection and the collection of case information but also has implications for how teams generate recommendations. Domestic violence fatality review teams often use standardized data collection and reporting forms that facilitate systematic inquiry. However, there are a number of decision-making points in the process that may have implications for ethical practice.

Choosing which deaths to review depends on a number of factors and often requires compromise and negotiation between team members and stakeholders. The definition of a specific type of death not only differs from team to team but also may be a point of disagreement within a team (Websdale, 2003, 2012). For example, our team began by reviewing homicides involving female victims killed by their male intimate partners. The seemingly specific case definition derived from our inception as a research project. However, as our process matured and we became more practice-focused, we realized that our identification of system issues and recommendations were limited to those specific scenarios and did not represent the range of cases confronted by system agencies. As such, we began including male homicide victims, perpetrator deaths, cases with same-sex intimate partners, and those where a bystander to the intimate partner violence was killed. Sources from the National Domestic Violence Fatality Review Initiative suggest that examining a broad range of homicides and other types of deaths attributable to domestic violence can help us better understand domestic violence and community response (National Domestic Violence Fatality Review Initiative, 2002). By broadening our scope, we were able to present a more complete picture of domestic violence homicide in our jurisdiction. No team has limitless resources, but some teams operate without funding or the necessary resources to review all or even most deaths. Regardless of circumstances, teams must make decisions about which and how many cases to review (Websdale, 2012).

Disagreements about case selection can lead to ethical debates over such things as why certain deaths are prioritized over others. Biased reporting may also become a problem if the results of a limited set of case reviews are extrapolated beyond the reference cases. These disagreements become less likely and easier to mitigate when definitions of target cases are clear and well-documented in team policy. Similarly, when case selection conforms to scientific principles of inquiry, teams can have more confidence in the validity of the results and recommendations that emerge from the process. The principles of systematic inquiry can guard against dilemmas related to case selection by creating a clear set of rules or criteria for case inclusion. Clear decision rules and reporting case selection criteria can also mitigate a key ethical concern regarding policy recommendations based on random or nonrepresentative samples of relevant deaths under investigation.

Another area germane to systematic inquiry in domestic violence fatality review is associated with team procedures for accessing and collecting case data. Once cases are identified, the team must assemble a physical or virtual case file that documents the relevant history, incident details, and system or agency responses. This collection of data also invokes ethical considerations. Fatality reviews may involve not only the retrieval of public records but also negotiating agreements to access confidential records maintained by system actors and agencies (McHardy & Hofford, 1999; Thompson, 2002). These agreements are an important starting point for ensuring ethical acquisition of data. Team members and staff involved in data collection must adhere to the agency’s policies as well as the ethical guidelines of their own profession. Regardless of the format, it is also important that data collection strategies are consistent with the regulatory environments of state and federal law (Thompson, 2002). In addition, disagreements among members may arise, if for example, a participating agency chooses to withhold information other members view as critical to the case review. The lack of access to specific information (e.g., children’s school records) may be frustrating for the team, as they try to piece together case information and assess prevention and intervention efforts. However, team staff, members, and stakeholders need to be aware of the limitations of data access and disclosure in the team’s work, overall.

It is also important that teams are both thorough and consistent in data collection across cases to the extent possible. Without such consistency, teams again confront the ethical concern that recommendations are more idiosyncratic than generalizable because they may be biased toward those for which the team has access to the broadest array of information. For example, a team may include the direct testimony of case-involved system actors in some reviews but not others. This is especially relevant for teams that function at regional or state levels, because system actors with direct case involvement may not be available to attend meetings. To the extent that cases will differ in the level of available data, teams need to be cognizant of these gaps when generalizing their findings and conclusions and making policy recommendations.

Competence

The second principle calls for evaluators to provide competent performance to stakeholders (AEA, 2004). In the evaluation literature, competence is broadly interpreted as referring to the appropriateness of the background, training, and experience of the evaluator (Morris, 2011). Domestic violence fatality review teams operate as an evaluation body. Whether or not they are engaged in project design or data collection, members represent their agency or profession as system experts during case review. Moreover, members are called on to make judgments about the functioning of prevention and intervention systems. Issues of competence in the practice of domestic violence fatality review are relevant to decisions about team composition, member participation, and training in the fatality review process.

The broad range of members on fatality review teams is an important strength of the technique and allows for comprehensive evaluation and recommendations that are rooted in real-life experiences of system actors and affected community members. The representativeness of the team can also help to ensure competent performance to stakeholders. In some instances, the composition of the team is explicitly specified by statute. However, most of these statutes describe broad categories for the representation of “victim service programs” or “members of the business community” or incorporate language to allow the agency of responsibility some leeway to include other members according to the team’s needs (e.g., see Florida Statutes § 741-316; Mont. Code Ann. § 2-15-2017; Tenn. Code Ann. § 36-3-624). Others from outside the systems of prevention and intervention might also be included in the case review meeting. For example, representatives from marginalized populations disproportionately affected by the type of fatality under review or individuals with personal experience to share with the team may be invited to participate as members or on an ad hoc basis (Websdale et al., 1998).

From time to time, teams will encounter cases that fall outside the scope of expertise represented by the membership. For example, our team sometimes reviews cases involving parties who live in an environment with different systems for prevention and intervention such as a college campus, a military base, or on tribal land. Assessing these systems by the standards of another community may not yield accurate insights about system functioning and lead to a misrepresentation of system function and needs. In such instances, we try to invite relevant experts from those communities to the team meeting to help us understand the unique community dynamics that might influence the case and our recommendations.

Simply ensuring a representative team, however, is not enough to meet the ethical concerns that motivate “competence” as a guiding principle. It may be safe to assume that participants are all competent in their professional roles. However, the expansion of their duties to include systematic evaluation of the range of issues that frame domestic violence homicides may be outside of their expertise. All team members come to the table with prior experiences that may include biases not directly related to the case review. This increases the importance of having clear guidelines for case review procedures and participation. In addition, relevant to competence is consistent, stable team membership. An ad hoc and inconsistent involvement of key participants can lead to recommendations that are superficial and that do not reflect well on the important work of fatality review, thereby limiting competent performance to stakeholders.

Integrity and Honesty

The third principle calls for evaluators to display integrity and honesty in their own behavior, and attempt to ensure the integrity and honesty of the entire evaluation process (AEA, 2004). Evaluators define integrity and honesty not only in terms of individual conduct but also as transparency of all aspects of planning, operations, deriving outcomes, reporting conflicts of interest, and providing a truthful report of findings (Morris, 2011). This principle highlights a number of possible ethical issues for fatality review teams, including the adherence of participants to the policies and procedures of the case review process, recognition of members’ conflicts of interest, and presentation of findings consistent with the observations derived from reviewed homicides.

The sensitive and often confidential nature of the data that fatality review teams rely on necessitates a commitment to integrity and honesty. This means that fatality review teams should use appropriate methods of collection and use of data. The benefits to the public good do not create an exemption to the legal and intended guidelines established by the data owner or review subject—and teams should avoid using an “any means necessary” approach to data collection. The review process must also encourage and expect ethically sound responses to situations that laws and regulations may not address. For example, sensitive data from decedents may be protected under confidentiality policies at the local, state, and federal level, but family members, advocates, or other system agents may disclose protected information to further the interests of the review process (Thompson, 2002). In such cases, it is even more important that team members keep any details that stakeholders share in strict confidence. Inappropriate access to information, even when technically legal, may violate the rights of privacy and integrity of individual subjects of the review. Furthermore, relationships with participating agencies could become strained if they perceive the team as sidestepping agency policy on data access and security.

Another set of ethical concerns related to the principle of integrity and honesty stem from the fact that team members often find themselves in the role of subjects of review and participants in the review process. This is especially the case for those who represent law enforcement, judiciary, victim advocates, and social service agencies in community-based reviews. These members work in agencies that have responded to a particular case under review, or perhaps even more problematic, that may shoulder some responsibility for not having responded or for having responded in ways that the team views as inadequate. The purpose of having stakeholders at the table is to engage in an honest assessment of the system. Such an assessment may result in the type of self-evaluation that can be difficult but, over time, can improve agency and broader system functioning. For example, stakeholder team members can encourage their agencies to bolster their strengths and address their weaknesses in ways that can improve overall responses.

In addition, some teams review cases with pending litigation. Statutes authorizing fatality review teams often prohibit the use of review data and findings from subpoena (Websdale, 2003, 2012). Should members working on the case or in agencies involved in the case recuse themselves from the case review? Recusal could result in a loss of potentially valuable case information; however, participation of case-involved parties may also lead members to guard case information. This duality raises the potential for conflict of interest and requires member commitment to the integrity and purpose of the fatality review process.

The principle of integrity and honesty also applies to the domestic violence fatality review process for reporting findings and producing recommendations. A primary factor that differentiates the fatality review process from traditional program evaluation is that almost all fatality review teams exist to make formal recommendations and even advocate for change, not simply to provide high-quality data and analysis for others to use. Recommendations regarding agency policy or criminal law have the potential to result in real consequences in the lives of community members. The principle of integrity and honesty necessitates that any advocacy be in support of recommendations based on case review data, not in support of particular agencies or procedures or to support particular political agendas. The evaluation literature suggests that this means evaluators should also avoid or correct misuse of findings by stakeholders and others in the community (Morris, 2011).

Respect for People

The fourth principle proposes that evaluators respect the security, dignity, and self-worth of respondents, program participants, clients, and other evaluation stakeholders (AEA, 2004). This principle incorporates the standards for the ethical treatment of human subjects including respect, beneficence, and justice. In evaluation, these principles extend to the intended beneficiaries of a program or policy, persons involved in implementation, agencies, and other stakeholders in the community (Morris, 2011). Domestic violence fatality review teams function in an environment that complicates anonymity and confidentiality of the subjects of the case review. Moreover, the concept of justice requires teams to consider the ways in which their work may affect the distribution of services to the general population.

Homicide case reviews are fraught with the potential for ethical conundrums related to respect for personal autonomy and privacy. Each review examines the intimate details of the victim’s life and death. In addition, case reviews document the victim’s relationship with the offender and details of the offender’s life (and sometimes death). When specific information is disclosed as part of the review, confidentiality during the entire review process is paramount for deceased and surviving review subjects, including friends, family, neighbors, and other witnesses to the crime (Thompson, 2002; Websdale, 2003; Websdale et al., 1998; Wilson & Websdale, 2006). While a great deal of the review is composed from available public sources, teams may confront questions about the use of information that is not directly relevant to the case review. Constructing the case timeline also involves putting the victim and offender’s lives in motion (Websdale, 2012). This activity often brings other living persons into the case review. This may include surviving family members, friends, other acquaintances, and system actors who responded to the homicide or to prior system contacts. As we indicated earlier, some of these actors may also be fatality review team members. These persons are often identified in public record documents, and, in some communities, it may be difficult to conceal identities. The practice of fatality review requires balancing the “public need to know” with the privacy of individuals (Thompson, 2002). Teams should be judicious with information about identifiable system agents and survivors, who could potentially be involved in services offered by stakeholder agencies either at the time of the review or sometime in the future.

Confidentiality and attention to the relevance of information should also be afforded to offenders, who may be only suspects at the time of the review or acquitted of crimes. Most teams choose to review closed homicide cases to avoid conflicts of interest and liabilities for system actors participating in the review that may be associated with the determination of charges, prosecution, and sentencing (Websdale, 2003). However, living offenders are also potential clients of stakeholder agencies. This may be particularly sensitive in small jurisdictions with local review teams.

The prospect for fatality review to foster honest participation and generate insights about how systems aid and fail victims of domestic violence also requires respect for team members and the agencies they represent. The interprofessional nature of domestic violence fatality reviews can result in competing or adversarial agencies working together, and often includes an unavoidable imbalance of power within team composition (Websdale, 2012). System improvement can only be successful under the auspices of mutual respect for individuals, and significant effort to fully understand the broader mission of each agency and the constraints it faces due to such things as funding, community support, political oversight and influence, and competing public services. A respectful but candid evaluation of system components is among the desired results of fatality review. Confidentiality should also extend to the review team members and the agencies they represent.

Respect for persons also extends to the consideration of the impact of review findings and recommendations on the population of persons for whom future prevention and intervention efforts are aimed. For this reason, it is important to consider the population when identifying cases for review. A well-thought-out and documented plan for selecting cases will help teams avoid reifying the existing service populations. The concepts of beneficence and justice also require generating benefits for individuals and/or society that are commensurate with the burden of research participation (National Commission on Human Subjects of Biomedical and Behavioral Research, 1979). For homicide deaths (and other types of fatalities), fatality review yields no direct benefit for review subjects. The alternative is to generate benefits for the broader population. For our team, the beneficiary population includes persons at risk for injury or death due to domestic violence. The lack of a well-defined population may lead to biased case selection, thereby reducing the applicability of recommendations aimed to improve system issues that affect the beneficiary population.

Responsibilities for General and Public Welfare

The fifth principle calls for evaluators to articulate and take into account the diversity of general and public interests and values that may be related to the evaluation (AEA, 2004). Evaluators frequently face difficulties in reconciling conflicting views on what constitutes “public good” among stakeholders (Morris, 2011). On the surface, the mission of fatality review centers on improving public health and safety, focusing almost entirely on prevention of future cases of premature death. However, the interprofessional nature of the process, combined with the focus on a broad set of systems of prevention and intervention, can yield conflicting recommendations and difficulty in identifying targets to be held accountable for system functions.

In most teams, there are no individual benefits to the decedent or the circle of family and friends from a review of the system of response. Furthermore, the benefits to society from system review can be divergent and difficult to align. Recommendations that call for an increased incarceration of offenders to prevent escalation, for example, do not align with the need to address prison overcrowding or recommendations to rely more heavily on court-mandated offender programs. Teams can address public interest concerns by relying on well-qualified and experienced personnel in the review process who recognize the tensions inherent in system change and who advocate for sensible, measured approaches to improvement.

A bigger challenge to the ethical and successful use of fatality review outcomes, and a factor that sets community-based fatality review apart from traditional program evaluation, is the fact that no one is in charge of the system under evaluation. Program evaluations review components of a single program, or a set of programs that are under the control of an entity that maintains the ability to act on the results of the evaluation. Conversely, domestic violence fatality review conducts evaluation in a vacuum of overall system control. McHardy and Hofford (1999) document more than 40 agency types or occupational groups recommended for consideration on domestic violence fatality review teams. Some agencies are governmental and others are nongovernmental, some are for-profit, others are nonprofit. These agencies make up a system in the sense that they are components of the service delivery system frequently encountered by members of the population at risk for death due to abuse and neglect. Of course, individual components of the system are overseen by authorities appropriate to the mission of the agency. However, system improvement is more than the sum of individual parts. Fatality review examines the intersection of sometimes unrelated components of a large and complex system, and the interconnectedness of those components as they converge and interact to address a common problem. This does not discount the application of the ethical constructs of traditional program evaluation to fatality review, but does create a need for further discussion regarding the ethical dilemmas inherent in fatality review and how to support participants in their desire to maintain the highest ethical standards throughout the process.

Discussion and Recommendations

When addressing the application of the AEA Guiding Principles to the practice of evaluation, Michael Morris (2011) writes that most articles reporting evaluation findings have little or nothing to say about ethical issues, but “such silence does not necessarily mean that ethical considerations played no significant role in the evaluations; it could just mean that no noteworthy ethical challenges were encountered” (p. 145). The effort that has gone into the design and implementation of the practice of fatality review helps to mitigate many of the potential pitfalls we discuss in this paper. However, as we move forward in the use of this practice, with new practitioners and a constantly changing landscape of information availability and access, it seems prudent to lay out these issues in a public forum and allow practitioners and researchers alike to examine this method of inquiry and engage in a conversation about the ethical practice of fatality review.

There are numerous decision-making moments in team formation, organization, and the practice of domestic violence fatality review. Given the interprofessional nature of these teams, it is important to consider the ways in which these decisions may lead to conflict or affect the team’s process. One of the primary insights of the current work is that the applicability and credibility of fatality review findings and recommendations are affected by decisions made at every stage of the process. The National Domestic Violence Fatality Review Initiative (2002) highlights the need for teams to be honest about the practice of fatality review and the impact of this activity in the community. Specifically, the body advises teams to assess the appropriateness of recommendations and track system changes consistent with team recommendations. In addition, the needs and preferences of members and their agencies may diverge in marked ways. This is often the case in interprofessional work. Teams should make an effort to not only avoid forced consensus but also acknowledge these differences in ways that improve our understanding of the overall system of prevention and intervention.

Below, we summarize some of the key ethical conundrums identified in this work. Using the AEA Guiding Principles, guidance from previous works on domestic violence fatality review, and our own experiences, we also provide recommendations for ways to address ethical dilemmas and avoid the misrepresentation of team findings and recommendations. We organize the discussion into two broad areas of team process: team organization and case review methodology.

Individuals and agencies organizing fatality review teams have to make a variety of decisions related to team organization and personnel. These include assigning team leadership responsibilities, making decisions about team composition, and determining which agencies and individuals should participate in case review meetings. Even established teams will face decisions related to replacing personnel. The exclusion or marginalization of stakeholders from the process may result in an inaccurate portrayal of the prevention or intervention efforts during the case review and in the team’s findings and recommendations.

Clearly, a range of ethical concerns is at play when working to provide information for a broad range of stakeholders. To ensure competent performance to stakeholders, teams should be representative of relevant professions and agencies. Moreover, teams should have a mechanism for adding new stakeholder members when needed. Ideally, state statutes specifying requirements for participation should be written broadly enough to allow the appointing agency or the team the latitude to add new members from agencies not specified in the legislation. When new or unrepresented interests appear in cases, teams should look for ways to incorporate experts in those areas or at a minimum gather and present relevant information about the system agency during the case review. The principle of integrity and honesty suggests that teams should be transparent about which stakeholder agencies are represented in the review process and make note of how the absence of representation may affect the team’s recommendations. Our team instituted an annual process evaluation to track team performance measures on membership, system representation, and meeting participation (NMIPVDRT, 2012). The process evaluation report is separate from the report of findings but is provided to team members and is publicly available.

Teams may have a strong commitment to stakeholder representation, but that does not necessarily translate into representative participation. Teams must work to ensure regular and meaningful member participation. We have found that participation issues are less troublesome when members have a clear understanding of the case review process. This includes orienting members to the policies, procedures, and the goals of the review process and providing members with regular training and educational opportunities (McHardy & Hofford, 1999). Member participation may also be influenced by the effectiveness of team leadership. The principle of integrity and honesty suggests that it is important to have team leaders who maintain open communication, ensure compliance with policies and procedures for all participants, and adhere to the mission of the review (McHardy & Hofford, 1999). Team leaders can also encourage respect for persons in meetings by guiding members through a structured discussion process that focuses on the details of the case review and avoids pushing members to agree on all aspects of the case. Personal attacks and agency shaming might decrease the willingness of stakeholders to participate in fatality review. Teams should be mindful of the treatment of members and participants and make an effort to ensure voluntariness of participation. The provision of confidentiality for case review materials and member participation in meetings also serves to ensure that review participants can engage in candid and constructive system evaluation (Thompson, 2002; Websdale, 2003; Websdale et al., 1998; Wilson & Websdale, 2006). Like other domestic violence fatality review teams, our team requires a signed confidentiality agreement. We also take a few minutes at each case review meeting to remind members about the provisions of the agreement.

Once the teams are organized, there are numerous decisions to be made about the case review method. These include selecting and identifying cases for review, collecting case data, and deciding on a strategy for reviewing cases and generating recommendations. In addition to confronting limitations in the case review process, teams may also experience internal and external disagreements about data access and sharing. Most teams use the case review process to generate aggregate case findings and highlight systemic system issues. A lack of systematic design and consistency in application can lead to biased results and undermine the team’s overall goal of improving prevention and intervention for domestic violence.

One of the primary goals of fatality review is the prevention of injury and death in our communities. However, the applicability of fatality review findings to policy makers and practitioners requires a clear understanding of the population for whom these findings are relevant. Defining the population is a starting point for ensuring that cases selected for review are representative of the deaths we are seeking to prevent. We recognize that not all teams have the resources to review all deaths that fall into the population. The principle of systematic inquiry would suggest that, in such cases, teams should develop and document selection criteria and ensure that findings and recommendations are presented within the context of case selection.

Systematic inquiry can also ensure that teams collect data and review cases in ways that encourage adherence to other principles of ethical practice. The principle of respect for people suggests that questions examined in fatality review be essential to the purpose of the review and seek to negate, not fall subject to, invalid stereotypes held by the public and even team members. A plan for standardized data collection should include a scope of inquiry, which can help teams guard against overly intrusive, invalid, or irrelevant lines of investigation.

Small teams and those informally organized or underfunded may face difficulties with regard to access to information and collecting standardized data (Websdale, 2003). All teams face some limitation with regard to these issues, but resources should not preclude a coherently designed and consistently applied fatality review process with a well-documented, standardized methodology. The principle of integrity and honesty suggests that it is incumbent on fatality review teams to follow the planned course of action and to disclose any deviation from the team’s procedures to avoid misrepresenting findings. Teams should work to make the activities of fatality review as transparent as possible. Most teams do not produce research reports. As such, team reports are generally lacking details of the method of fatality review. Over time, our team has developed progressively more complex documentation and practices related to the case review method. We have written policies and procedures, which outline our scope of work. In addition, our annual process evaluation documents data sources, problems with data access, case selection criteria, and case review procedures. Written and publicly available protocols can provide guidance for the team and provide stakeholders with the proper context for interpreting findings and recommendations.

It is not our intent to suggest a single model of ethical principles for conducting domestic violence fatality review. Rather, we hope this exercise will open a public dialogue on ethics in fatality review that will lead to a better understanding of how parallel and sometimes competing stakeholder values shape this practice. Indeed, as is evidenced in our discussion, the ethical principles that guide evaluation are broad and can be implemented in a variety of ways to accommodate different team processes and procedures. The key, as we see it, is that teams are cognizant of the potential ethical conundrums embedded in the fatality review process and that they aim to mitigate these by adhering to clearly articulated policies that guide all aspects of the process from case definition and selection to team member participation and interaction. The team should construct these policies in collaboration with stakeholders in a way that gives thoughtful consideration to conflicting values related to the practice of fatality review. We suggest that the ethical principles guiding evaluation are pertinent here because they offer an established framework for developing team processes that can reduce the likelihood of ethical breaches and help resolve potential ethical conundrums.

Fatality review teams face unique challenges that are not fully addressed by the principles of evaluation. Teams rely on members to perform dual roles: evaluator and system expert. This may lead to conflict between members during the course of team activities. The development of standards for participation may differ across professions and types of review teams. Teams should be reflexive with regard to ethical tensions arising from the duality of roles and establish a mechanism for documenting and evaluating these tensions for further consideration. The target of the study is another area of divergence. Program evaluation is typically applied to a single program, which limits concerns about the impact of the review and ownership of the findings. Domestic violence fatality review teams conduct evaluation in a vacuum of overall system control. This characteristic raises unique concerns about the report of findings, the appropriateness of recommendations, and the need to ensure that the process contributes to the public good in a representative way. Teams should begin analyzing the ways in which review findings and recommendations affect system agents and system functioning as a whole. Teams need to be cognizant of the fact that potentially positive changes to one element of the system might have deleterious effects on other system components, and they must, therefore, minimize this potential to the extent possible.

More empirical research on the methods of fatality review across substantive areas and jurisdictions is warranted. This should include attention to the ways in which teams navigate ethical dilemmas and look for ways to increase transparency and documentation of the process used to generate findings and recommendations. Research on the actual experience of ethical conundrums and how teams handle these issues should be conducted in a way that provides teams with anonymity and confidentiality. Moreover, we could learn more about ethical issues related to the practice of fatality review by consulting other professional guidelines for ethical practice, such as those of public health practitioners, social workers, and criminal justice system professions. The interprofessional nature of this work requires not only that we safeguard confidentiality but also that we acknowledge the unique ethical conundrums of the practice and research components of fatality review.