confidentiality_article.pdf | |
File Size: | 1640 kb |
File Type: |
Confidentiality in direct social-work practice: Inevitable challenges and ethical dilemmas
By: Kathleen Millstein
Families in Society 81.3 (May/Jun 2000): 270-282.
Abstract (summary)
The objective of this study was to learn more about practitioners' ethical dilemmas In maintaining client confidentiality. A mailed, anonymous survey of experienced social-work practitioners was conducted to gain an understanding of how practitioners address confidentiality issues in their work with clients and to identify specific areas in which practitioners experience ethical dilemmas related to confidentiality, the factors contributing to these dilemmas, and the resources used by practitioners to resolve them.
Headnote
Social workers have long concerned themselves with confidentiality and its importance to practice. In 1922, social workers created their Code of Ethics, a major precept of which is the protection of confidentiality, defined as the regulation, both legal and ethical, that protects the client's rights of privacy. Social workers are confronting many ethical issues related to confidentiality, such as increasing demands for accountability, mandated duty-to-protect or -warn provisions, expanding court involvement in professional decision making, and widening access to information in records through the expanded use of computer technology. The objective of this study was to learn more about practitioners' ethical dilemmas in maintaining client confidentiality. A mailed, anonymous survey of experienced social-work practitioners was conducted to gain an understanding of how practitioners address confidentiality issues in their work with clients and to identify specific areas in which practitioners experience ethical dilemmas related to confidentiality, the factors contributing to these dilemmas, and the resources used by practitioners to resolve them. Major issues of practitioner concern and specific areas in which practitioners felt a need for more resources, education, and policy clarification were identified.
CONFIDENTIALITY CAN BE DESCRIBED as the regulation, both legal and ethical, that protects the client's rights of privacy. It refers "broadly to legal rules and ethical standards that protect an individual from unauthorized disclosure of information" (Smith-Bell & Winslade, 1994, p. 184). Confidentiality creates boundaries around the client/worker relationship. It also creates boundaries around information outside of that relationship, including information that can be communicated within, between, and among agencies (Rhodes, 1992).
As social workers, we define and balance our obligation to safeguard confidential client information with the rights of the individual, organization, and community to access that information (Bollas & Sundelson, 1995; Goldstein, 1989; McMahon & Knowles, 1995; Saakvitne & Abrahamson, 1994; Strein & Hershenson, 1991 ). We maintain a dual focus on "individual well-being and the well-being of society" (National Association of Social Workers [NASW], 1996, p. 1). In practice, we are seldom able to offer our clients "absolute confidentiality" in which "what data learned or observed by a social worker stay with that individual and are never passed on to anyone" (Wilson, 1978, p. 3). Instead, we offer a "relative" contract for confidentiality. We promise that information that is shared beyond the confines of the client-social worker relationship is shared responsibly, in the interest of providing client services, and with client understanding and consent consistent with legal mandates (Strom-Gottfried, 1998; Dickson, 1998).
Social workers today confront many ethical issues related to confidentiality. Our ability to protect confidentiality is diminished by increasing demands for accountability, mandated duty-to-protect or -warn provisions, expanding court involvement in professional decision making, and widening access to client record information through the requirements of thirdparty payers and the expanded use of computer technology. These complexities are reflected in the fact that "Privacy and Confidentiality" is now the longest section in the newly revised Code of Ethics (NASW, 1996).
The managed-health-care environment in which we find ourselves does not create all of our ethical dilemmas. Even before the days of managed care, social workers were unable to offer their clients unlimited confidentiality (McGuinn, 1938; Oyen, 1982). Managed care merely focuses our attention on ethical and practice issues that have always existed in our roles as professionals: How do we define our responsibilities to our clients, to ourselves, to society, and to funding sources when it comes to maintaining confidentiality? What do we tell and not tell our clients about the limits to confidentiality? Social workers struggle to define the nature of, and limits to, confidentiality in their varied and complex practice settings.
Given the challenges we face, it is striking that there is so little social-work research about ethical issues related to confidentiality in social-work practice. Research from the fields of psychology and psychiatry informs us that clients value confidentiality from their therapists (Schmid, Appelbaum, Roth, & Lidz, 1983; Appelbaum, Kapen, Walters, Lidz, & Roth, 1984; McGuire, Toal, & Blau, 1985), feel that it is important in therapy relationships (Miller & Thelen, 1986), and are more willing to seek and to remain in mentalhealth treatment if confidentiality is assured (Appelbaum et al., 1984; Miller & Thelen, 1986). In conjunction with the clients' perspectives, it is a belief of many practitioners that confidentiality is an important factor in maintaining a helping relationship (Jagim, Whittman, & Noll, 1978; Lindenthal & Thomas, 1980, 1986; Lindenthal, Amaranto, Jordan, & Wepman, 1984; Lindenthal, Jordan, Lentz, & Thomas, 1988; Otto, Ogloff, & Small, 1991).
There is also a small body of literature that challenges the value of confidentiality. Oyen (1982) notes that the principle of confidentiality derives from and supports the view of social problems as based in the individual. Keeping client information confidential implies that those in need of mental-health and social services have something to hide and must be protected from the intrusion of others. In addition, society is not provided with the information it needs for a fuller, more comprehensive understanding of clients' problems within the broader social matrix. In this view, protecting client confidentiality maintains the status quo of an individual-based construction of social problems.
A number of studies have attempted to ascertain under what conditions mental-health practitioners might breach confidentiality (summarized in Lindenthal & Thomas, 1986). These studies, conducted primarily by psychologists using samples that do not include social workers, use a series of vignettes representing hypothetical mental-health-practice dilemmas related to breaching confidentiality (Lako & Lindenthal, 1991; Lindenthal & Thomas, 1982a, 1982b; Lindenthal et al., 1984, 1988). Each vignette is followed by a questionnaire that attempts to assess what practitioners might do in that situation. They do not identify what practitioners define as an ethical dilemma, the nature of dilemmas that they actually experience, or their actions. Only one study (Lindenthal et al., 1988) focused on social workers' management of confidentiality; it found that social workers were more likely than other professionals to report that they will breach confidentiality when confronting specific clinical situations.
The absence of studies on social workers and social-work practice and the heavy reliance on analogue research, such as Lindenthal and colleague's, leave our profession with little knowledge of how we maintain confidentiality in work with our clients. To develop a better understanding of how social workers address confidentiality issues in direct practice with clients, this study used an exploratory-descriptive research design (Fortune & Reid, 1999). The study was conducted by 10 students under faculty direction as part of a year-long research seminar in the second year of the masters of social work (MSW) program at Simmons College School of Social Work. An anonymous survey sought information about social-work practitioners' beliefs, practice actions, and policies regarding confidentiality. Furthermore, we wanted to identify the ethical dilemmas respondents faced in maintaining client confidentiality, and the ways that they resolved these dilemmas. We defined an ethical dilemma as "a situation in which a practitioner feels pulled in two directions and isn't sure what to do." This definition was drawn from Rhodes' (1986) definition of an ethical dilemma as a "choice in which any alternative results in an undesirable action" (p. xii).
Method
A purposive, nonprobability sample of 372 experienced social-work practitioners was used. The sample included all field supervisors at Simmons College School of Social Work for the 1997/1998 academic year (n = 245) and all graduates of the class of 1992 (n = 127) (1992 graduates who were also field advisors were counted in the alumni sample). Simmons College School of Social Work is a single-- concentration masters-degree program specializing in direct practice with individuals, families, and groups. Both groups, field supervisors and practitioners five years post-graduation, represent a sample of experienced MSW social workers. By using the class of 1992, it was our intention to include a group of more recent practitioners whose post-masters practice experience has been in the current managed-care environment. While clearly limited in generalizability, this method of sample selection offered the researchers ready access to a reliable sampling frame and the opportunity to conduct an initial exploration of the topic in the nine-month framework of the student research experience.
Three hundred and seventy-two questionnaires were mailed. Nine were returned due to incorrect addresses; all of these were in the alumni sample. In addition, one supervisor who was interviewed as part of instrument development was excluded from the sample. Of the 362 possible questionnaires, 153 were returned. One of the returned questionnaires was not validly filled out, yielding 152 (42%) for analysis. Ninety-eight (64.5 % ) of the respondents were supervisors and fifty-four (35.5%) were alumni of Simmons College School of Social Work. The response rate was 41 % for supervisors and 43 % for alumni.
The self-administered, anonymous, structured questionnaire was in five parts: ( 1 ) background information about the respondents; (2) practitioner beliefs about confidentiality in practice; (3) how the respondents address issues of confidentiality with clients; (4) the policies on confidentiality in the practitioner's practice setting; and (5) areas in which respondents have experienced ethical dilemmas and how they handled them. The questionnaire was designed to take approximately 20 minutes to complete.
As part of the development of the questionnaire, six hour-long interviews were conducted with experienced practitioners, with social-work educators and researchers, with an ethical theorist, and with a social worker/lawyer. The purpose of the interviews was to refine and narrow the study's focus and questions and to develop items to include in the questionnaire. Five of the six interviewees also volunteered to pretest the final questionnaire. The questionnaire was further pitoted by seven Simmons College School of Social Work faculty and at a meeting of the Education Committee of the NASW Ethics Commission. The questionnaire was mailed out with a cover letter explaining the study's purpose; three weeks after the initial mailing, a reminder postcard was sent to the entire sample. Questionnaire data were analyzed using descriptive statistics. Responses to open-ended questions on the questionnaire were analyzed using thematic content analysis (Miles & Huberman, 1994).
Findings
All respondents were MSW social workers who had either graduated from, or were field supervisors for, a graduate school of social work that offers a single direct-practice concentration. They were predominately Caucasian (91%), female (84%), experienced (median number of years in human service was "11 to 15 years"; number of years in human services includes all human-service experience, including experience obtained prior to receiving their MSW), and mature (average age, 42).
Ninety-four percent of the respondents were currently employed and worked in a diverse group of settings, with the largest percentage in outpatient mentalhealth clinics (20%). The next most frequently identified practice settings were general hospitals ( 10 % ), schools ( 10 % ), community health centers (9%), and psychiatric hospitals (8%). The least frequently reported employment settings were private practice (4%), health-maintenance organizations (HMOs) (4%), criminal-justice system (3%), and the State Departments of Social Services (2%) and Youth Services ( 1 % ). Respondents had been employed in their current practice setting for an average of 6.2 years.
The greatest percentage of their direct-service time was spent with Caucasian clients (average, 66%), followed by African-American (18%), Latino(a)/Hispanic (9%), and Asian clients (3%). The respondents worked with a predominantly adult population, with an average percentage of 57% of their practice time spent with adults. An average of 19.1 % of their practice time was spent with adolescents, 15.9% with children, and 6.9 % with clients older than age 65. When asked to indicate all the theoretical orientations that inform their practice, over half the respondents (53%) espoused a psychodynamic theoretical orientation, followed by eclectic (52%), client-centered (48%), behavioral (48 % ), systemic (43 % ), and short-term (34%). They used multiple interventions, with individual (97%), crisis (74%), group (70%), family (69%) and case-management (60%) as the most frequently indicated modalities.
Beliefs about Confidentiality
Social workers were asked to state their degree of agreement with seven statements that address beliefs about professional, ethical, and legal obligations to maintain confidentiality; about client expectations about confidentiality; and about the threat third-party payers pose to maintaining confidentiality. Respondents were in greatest agreement (strongly agreed) that confidentiality is necessary for maintaining a therapeutic relationship ( 73 % ), and that a social worker has a professional/ethical (88.7%) and legal (85.4%) responsibility to keep information concerning a client confidential except under circumstances prescribed by Massachusetts and federal law.
There were fewer respondents who strongly agreed that state ( 71.3 % ) and federal (69.5 % ) courts must respect the confidentiality of social worker/client communications. The fewest responses of strongly agree were to the statements that most clients expect that communications with social workers will remain confidential (48.3%) and that the requirements of third-party payers threaten the confidentiality of the therapeutic relationship (47%).
It is of note that, while 73.5 % strongly agreed that they believe that confidentiality is necessary for the therapeutic relationship, only 48.3% of the same sample also strongly agreed that they believe that clients expect that confidentiality will be maintained.
The greatest range of opinion came in response to the statement that third-party payers threaten the confidentiality of the therapeutic relationship. The largest number of respondents who strongly agreed were employed in outpatient mental-health clinics (25%), followed by schools (15%), HMOs (7%), and residential treatment (7%). Those least likely to strongly agree were employed in Department of Youth Services (2%) and the criminal-justice system (2%). The majority of those who were undecided worked in general hospitals (75%). Bivariate analyses of the relationships between characteristics of the respondents such as age, experience, theoretical orientation, preferred modes of intervention, practice setting, and client characteristics and specific beliefs did not yield significant findings.
Addressing Confidentiality with Clients
We asked 10 questions about when, how, and under what special circumstances social workers address confidentiality issues with clients. The response options were "always," "sometimes," "never," and "not applicable" (Table 1).
A third of the sample reported that they always inform clients about confidentiality prior to beginning the first session/client contact. More than 50% reported that they always inform clients during the first session/client contact. Further analysis indicated that 80% of those who always inform about confidentiality prior to the first session also inform again in the first session/client contact (X^sup 2^ = 48.9, df = 16, p < .001 ). These data suggest that most of the social workers in our sample are informing clients early in their work and at more than one time. Yet, 6% of the respondents reported always waiting for an issue to arise before they inform clients about confidentiality.
While no statistically significant relationships were found between practice setting and respondents' reported practices in informing clients about confidentiality, the largest percentage of those who reported that they always inform clients prior to the first meeting worked in mental-health (30%) or health (21 % ) settings. This pattern was also evident for those who informed clients in the first session (24% mentalhealth, 22% health). Respondents who spent a greater percentage of their time working with elders were somewhat less likely to wait until a situation arises to inform (r = .30, p < .001). In addition, the practice of informing clients about confidentiality in the first session was related to respondents beliefs that the social worker has a legal obligation to keep client information confidential (X^sup 2^ = 20.7, df = 8, p < .008 ) and that clients expect that communication with the social worker will be confidential (X^sup 2^ = 16.5, df = 6, p < .01).
In the second category, which addressed how clients are informed about confidentiality, practitioners were asked if they informed their clients verbally, in writing, or with a written form that the client signs. The findings indicate that 79.3% of the respondents reported they always use verbal discussion to inform their clients about confidentiality. Fewer than half of the respondents always inform their clients in writing or use a written form that the client signs. Practice setting was related to whether a respondent informed a client about confidentiality in writing (X^sup 2^ = 56.6, df = 40, p < .04) and had the client sign a form (X^sup 2^ = 71.9, df = 40, p < .001), with these practices more prevalent in health and mental-health settings. There were no significant differences between the experienced supervisors and the alumni for any of these practice behaviors, suggesting that these behaviors are not related to level of experience.
In the final category of practice actions, respondents were asked about three "special circumstances" regarding release of confidential information and when confidentiality is limited because a client is dangerous to himself or others. Eighty-nine percent of the respondents reported that they always have clients sign a form for release of client information when information is requested by other agencies. When asked about informing clients of third-party payer access to client information, of the 66.9% who found this issue applicable, 39.9% responded "always." The largest percentage of the 45 respondents for whom this question was not applicable worked in school settings (22 % ), followed by child welfare ( 18 % ). The responses to the final special circumstance show that 86.7% of the sample always inform clients about the limits to confidentiality should the clients be a danger to themselves or others.
Agency Policies and Resources
While there is no way to know in this study whether respondents' agencies have policies about which the respondents are unaware, a majority of respondents (85%, n = 126) reported that their agencies have policies on confidentiality, most of which are written (87%, n = 111). These written policies are not always shared with the clients. Only 43% inform clients of limits to confidentiality in writing (Table 1). There is a relationship between a respondent's practice setting and whether the setting has a policy (X^sup 2^ = 37.2, df = 8, p < .001 ) and whether it is written (X^sup 2^ = 23.9, df = 8, p < .002). Private practice, fee-for-service, and school settings are less likely to have policies, and these three settings along with criminal-justice and college-counseling settings are less likely to have written policies. Not having a policy was also weakly related to waiting until after the first client contact to inform clients about confidentiality (r = -.25, p < .009).
Seventy-five percent (n = 97) of respondents reported that they were generally satisfied with their agencies' policies regarding confidentiality and thought that they provided sufficient information for their clients to understand their rights and agency responsibilities. Of the 25% who were either not satisfied or unsure, several respondents identified the need for a policy written for clients and others stated the need for clients to be informed about insurer access to client records. There was no relationship between type of practice setting or theoretical orientation and satisfaction.
With regard to other agency resources on confidentiality, 85% of respondents had used their supervision to discuss confidentiality at least once in the past year, with an average of 3.4 times. Nearly half of all respondents (46%) had at least one in-service training on confidentiality in the past year. Finally, 40% of respondents reported that they have an ethics committee in their practice settings. Nine questions were asked about specific components included in agency policies in the respondent's primary work setting. Yes/no responses were requested for each component (Table 2). The policy components most frequently identified were those clearly mandated by law, such as release of information to individuals and agencies outside of the practice setting, legal limits to confidentiality imposed by duty to protect, and mandated reporting statutes. Less frequently reported were components related to release of client records within the agency, and release of client information and/or records to insurers. Even less frequently reported were policies related to discussion of case material with service providers in the practice setting, in case conferences, and in supervision. The least frequently reported policy concerned access to client computer records.
The number of specific policies each respondent reported was counted. The modal number of policy components was five, and the median and mean (SD = 1.8) were six. Higher numbers were correlated with respondents' satisfaction with their agency's policy (r = .29, p < .007), suggesting that respondents were more satisfied with more comprehensive policies. In addition, there was a strong relationship between the number of policy components and whether confidentiality issues had been discussed in the past year in supervision (r = .59, p < .001 ).
It is of interest that, just as there were few relationships between beliefs and practices, there were also few relationships between practice actions and specific policies. For example, it might be expected that there would be a relationship between working in an agency with a policy on mandated reporting and respondents informing clients about limits to confidentiality should they be a danger to self or others. This was not the case.
Ethical Dilemmas Related to Confidentiality
Identification of Dilemmas. A major objective of this study was to identify the areas and circumstances in which social workers experience ethical dilemmas related to confidentiality. From our reading, from class discussions, and from the interviews that were conducted as part of constructing the questionnaire, we identified nine major categories in which social workers could experience ethical dilemmas, and then we asked specific questions in each of these categories (for example, "I have experienced an ethical dilemma related to client confidentiality when verbally presenting case material... to a supervisor within my practice setting; at a case conference or training within my practice setting; to a supervisor outside of my practice setting; at a case conference or training outside of my practice setting" ). The nine categories were ethical dilemmas related to ( 1 ) presenting case material in case conferences and supervision (four questions), (2) collaboration (three questions), (3) discussing a client with family members and nonproviders either in the client's life or in the practitioner's life (three questions), (4) confidentiality of records (eight questions), (5) insurers (three questions), (6) working with at-risk groups (six questions), (7) specific client problems (14 questions), (8) specific age groups (four questions), and (9) working with clients who are similar or different from the practitioner (six questions).
There were a total of 52 questions. Each of the questions received a positive response from at least one respondent. For 18 of the 52 questions, 25 % or more of our respondents indicated that they had experienced an ethical dilemma related to maintaining confidentiality. It is these 18 questions and the categories under which they were presented that are summarized in Table 3. Collaborative relationships both inside and outside of the agency, and with family members and friends in the client's life, are a major area in which practitioners are reporting difficulty maintaining confidentiality. These dilemmas were not related to particular practice settings. Respondents frequently reported ethical dilemmas related to insurers, especially concerning requests to justify/extend treatment, work with clients who are potentially court involved, and writing for client records. Specific atrisk groups (e.g., clients exhibiting nonsuicidal self-destructive behavior, clients engaged in illegal activity) and client problems (e.g., trauma, mental illness, child welfare) also offered ethical challenges for our sample.
The seventh most frequently responded question addressed ethical issues when a client provides information about another mental-health professional's unethical behavior. Almost a third of the respondents had at least one client report a practitioner's unethical behavior. The least frequently reported areas in which ethical dilemmas were experienced were presentation of case material for publication (4%) and work with clients with employment issues (4%).
In an effort to understand the degree to which respondents were experiencing dilemmas, we added the number of dilemmas each respondent had identified. The average number of dilemmas was 11 (SD = 6.9). In addition, there was a positive relationship between the number of dilemmas experienced and the number of times our respondents discussed ethical issues related to confidentiality in supervision during the past year (r = .37, p < .01).
Handling Dilemmas. Finally, we asked the social workers in our sample to indicate how they handled their ethical dilemmas when they occurred. Table 4 lists the 13 options offered in response to this question. While there was a distribution of responses across the 13 options, the respondents overwhelmingly reported that they handled their ethical dilemmas through consultation with supervisors. This finding is consistent with earlier data about the importance of supervision as a resource for practitioners. Practitioners in our sample also used colleagues as resources. In addition, 55% of respondents discussed ethical dilemmas with the client. Only 6.6% responded that they did not realize that the situation was a dilemma until it was resolved. These respondents were somewhat more likely to report that their agency did not have a policy on confidentiality (r = -211, p < .05 ). With few exceptions, ways of handling dilemmas were not related to practice setting, theoretical orientation, years in current practice setting, policies, or practices.
We also posed an open-ended question about the areas in which respondents would like further knowledge to help them identify and resolve their ethical dilemmas. Forty-nine percent of the total sample responded (n = 74). The most frequently identified areas in which practitioners wanted information were concerning HMOs and insurers (17.6%) and legal issues, especially court subpoenas and privileged communication (20.3 % ). The next most frequently identified area involved what to tell families when working with children, adolescents, and vulnerable adults (14.7%).
To get a very general sense of client concerns about confidentiality, we asked respondents whether their clients had expressed concerns about confidentiality during the past year. Ten percent reported that their clients frequently expressed concerns, and an additional 56% reported that their clients sometimes expressed concerns. Only 4% had not had a client express concern during the past year.
Discussion
The social workers in this sample were in agreement about their professional, legal, and ethical responsibilities to maintain client confidentiality. There was comparably less agreement about state and federal courts' obligations to respect client confidentiality. Unfortunately, we did not have a question that ascertained the respondents' level of knowledge of state and federal regulations. As a result, these findings may be more reflective of their knowledge of the laws regulating privileged communications-those communications that cannot be shared in a court of law (Dickson, 1998)-than of their beliefs about the nature of state and federal responsibility (Jagim et al., 1978; Otto et al., 1991 ).
As shown by our respondents, most social workers also continue to believe that confidentiality is important for a helping relationship. Yet, they are no longer confident that their clients expect their conversations with practitioners to be confidential. It may be that workers believe that clients have become realistic about the possibility of "absolute" confidentiality. However, the differences in these beliefs may be due to the wording of our question. We asked about client expectations; it would be interesting to speculate about how respondents would have answered if we had used "believe" or "value" instead of "expect." Several studies (Schmidt et al., 1983; Appelbaum et al., 1984; McGuire et al., 1985) have reported that the majority of clients value confidentiality and believe that it should be maintained (Miller & Thelen, 1986). The greatest range of opinion came in response to the statement about the threats of third-party payers to client confidentiality. This finding is difficult to interpret; it may be a result of respondents working in practice settings in which insurance is not an issue.
Most of the social workers in our sample report that, in practice, they are verbally informing clients about confidentiality early in the work. They are functioning in concert with the Code of Ethics, which states that "social workers should discuss with clients and other interested parties the nature of confidentiality and limitations of clients' rights of confidentiality" (NASW, 1996, p. 11). A third of the respondents always inform clients prior to the first session/client contact. This figure is lower than data from Somberg et al.'s ( 1993 ) study of psychologists, in which 59.8 reported that they inform clients prior to the first session/client contact. Some are informing clients about confidentiality at more than one time. Appelbaum et al. (1987) describe this as a "process model" in which clients "receive information over time in a fashion that allows it to be contemplated, shared and assimilated" (p. 157). This model has also been suggested in recent social-work literature (Strom-Gottfried, 1998; Somberg et al., 1993) and in the Code of Ethics, which recommends informing clients "as soon as possible in the social-worker client relationship and as needed throughout the course of the relationship" (NASW, 1996, p. 11 ). There is still, however, a small percentage who are not informing clients until an issue arises, and these data are potentially worrisome. If practitioners are waiting to inform clients about confidentiality until an issue arises, there is an increased chance of the practitioner experiencing ethical dilemmas regarding confidentiality. The delay could mean that the client has been sharing information without being aware that there are inherent limitations to confidentiality in the client/practitioner relationship.
It is noteworthy that fewer than half of the respondents always inform their clients in writing or use a written form that the client signs. These respondents were more likely to work in mental-health or health settings and less likely to work in settings where there are less formal boundaries around roles and confidentiality, such as schools. Although verbal discussion is valuable and the Code of Ethics does not prescribe beyond "discussion," most current wisdom suggests the importance of having clients receive written material that they sign (Strom-Gottfried, 1998; Davidson & Davidson, 1996; Dickson, 1998). Given that this is the current practice wisdom, we need to better understand what makes it difficult for social workers to carry out this practice. There is a gap between theory and practice.
Most respondents report that their agencies have policies that address issues of confidentiality. Many of these agencies also have written policies. These policies most frequently include components mandated by law. At the time of this study in 1998, it seems striking that there would be any agencies, especially the types of agencies identified by our respondents, without policies. Ideally, all agencies would have policies and these policies would be written (Corcoran & Vandiver, 1996). The least frequently reported policy was one related to client access to computer records. This is an area of increasing concern (Gelman, Pollack, & Weiner, 1999; Rock & Congress, 1999), and many agencies are just beginning to struggle with the complex issues of computer records and managed-care access to these records.
Whether respondents reported that their agency had policies about confidentiality was directly related to practice actions, such as informing clients about confidentiality in writing and waiting until after the first client contact to inform. It was also those respondents whose agencies did not have a confidentiality policy who were more likely not to realize that a situation was a dilemma until it was resolved. In addition, the type of practice setting is an important factor in determining whether an agency has a policy, with private practice, fee-for-service, and school settings less likely to have formal policies. It is also the factor that most directly relates to social workers' practices concerning confidentiality.
The data about the amount of in-service training our respondents were receiving concerning ethical issues were encouraging. The finding that two-fifths of the respondents' agencies had ethics committees is especially noteworthy, because these committees are a relatively new agency resource. They offer an institutional structure for anticipating and resolving ethical dilemmas (Csikai & Sales, 1998).
However, in spite of the agency policies, education, and support, experienced practitioners in our sample were struggling in their practices with a number of dilemmas related to confidentiality. These ethical dilemmas are "expectable, not exceptional" (thanks to Stefan Krug for this comment). Ninetyeight percent of our respondents reported experiencing at least one ethical dilemma, and the average number of dilemmas experienced was 11. In addition, our data would suggest that social workers feel that confidentiality is a client concern.
The greatest number of ethical dilemmas with confidentiality arose when working in collaborative relationships within and outside agencies, and with client families and others in their lives. While it is not surprising that working with others on behalf of our clients would present social workers with problems about what can be said to whom, the frequency with which such collaborative dilemmas occurs is impressive. Half of the respondents have experienced a dilemma working with family members, and 42% have experienced a dilemma while working on a treatment team outside of their agency. This raises important questions, especially because these areas are not frequently addressed in agency informed-consent policies and not well defined in the Code of Ethics. With diminished resources, there is increased reliance on informal caregivers (Petrila & Sadoff, 1992), on coordinated case-management models of service delivery, and on court involvement. Meeting the information demands of third-party payers is also an area of concern (Davidson & Davidson, 1996; Strom-Gottfried, 1998). Social workers are often in the position of providing just enough information to justify needed services without violating clients' rights to privacy, especially when working with multiple systems and with clients at risk due to trauma, mental illness, and nonsuicidal self-destructive behavior.
These dilemmas represent areas of practice that are directly related to our profession's complex, Janus-- like role. We negotiate complex systems on behalf of our clients, all the while struggling to maintain their autonomy and self-determination. We have a mission as agents of social change and must attend to both "individual well-being in a social context and the wellbeing of society" (NASW, 1996, p. 1). In addition, challenges to dyadic assumptions about confidentiality and violation of these assumptions are inherent to the structure of our work with involuntary clients. A major question rests in how we define confidentiality in social-work practice. The concept of confidentiality derives from the physician-patient model, which carries the underlying assumption of a dyadic, one-to-one interaction (Shuman, 1985; Strein & Hershenson, 1991; Millstein, Dare-Winters, & Sullivan, 1994). Whereas this dyadic model may fit certain segments of mental-health practice, much of social-work practice not in mental-health settings is nondyadic. Group work, family and caregiver interventions, coordination of services, client advocacy, multidisciplinary team practice, and case management are common forms of our jobs as social workers. They all require collaboration. The traditional notion of confidentiality provides little guidance to social workers in these areas (Backlar, 1996; Millstein et al., 1994), yet much of the literature and research are framed within these assumptions. As Sabin (1997) wrote, "[We] have sometimes confused the confidentiality preconditions for successful therapy conducted under the confessional model with the standards appropriate to the entire spectrum of mental health care" (p. 41 ). Contemporary social-work practice is broad and diverse. Different practice areas may require different confidentiality guidelines. Our data support the importance of practice setting in understanding the challenges practitioners face.
In addition, as Oyen (1982) suggests, there may even be a need for a radical reexamination of the importance of confidentiality to social-work practice. Social workers in the daily discharge of their duties often experience dilemmas regarding confidentiality that are based on the competing values of individual versus collective good. As one respondent wrote, "The rights of the individual versus the rights of the community are of grave concern to me at this time"; he went on to describe his struggles with legislation requiring criminal background checks for social workers and community-reporting mechanisms for convicted sexual offenders.
It was unanticipated that ethical dilemmas concerning a client providing information about another mental-health professional's unethical behavior would be so frequently reported. It is a difficult finding to interpret in the context of our study. While the 1996 revisions to the Code of Ethics discuss this issue extensively in standard 2.11 and while there is an excellent explication of its implications in Reamer's review (1998), unethical professional behavior and the challenges it presents for colleagues remain areas that have not been well examined in the social-work literature on confidentiality.
Our study suggests that practitioners are facing significant ethical challenges as they attempt to maintain confidentiality in their work. Practice guidelines in these changing times are evolving at the same time that we continue to make decisions within our own understanding of ethical practice. A respondent wrote, "How and when does one know they've made the right decision in an unclear situation where there may be no defined rule or mandate?"
Social workers need places and opportunities to discuss their concerns (Fleck-Henderson, 1991 ). Some respondents specified their concerns as follows: "I do wish that...conversations about ethics were more common conversation pieces in the office setting" and "the issues around confidentiality seem simple, but in fact they are quite complex. I felt supported more in educational settings and feel less supported and more pressured now to know the answers."
It appears from the data that it is not professionals in ethics or law to whom practitioners turn when faced with an ethical dilemma. Practitioners are using supervisors and colleagues as their main resources for handling their ethical dilemmas. These findings are consistent with those of Landau's (in press) recent survey study of Israeli social workers, which found that 45 % of her respondents resolve their ethical dilemmas in supervision. This need for supervisory and collegial resources comes at a time of diminished emphasis on and availability of supervision and conference time. In addition, supervisors have not necessarily been trained to offer this support to workers. It is also of concern that discussion of case material in supervision is not a component of most agency policies on confidentiality. Clients have a right to understand how and in what ways their privacy is maintained within the agency (Strom-Gottfried, 1998). While the Code of Ethics specifies that "social workers should not disclose identifying information when discussing clients for teaching or training purposes unless the client has consented to disclosure of confidential information" (NASW, 1996, p. 12), this standard is often not applied to supervision.
In addition to supervisors and colleagues, workers are using a broad armamentarium of ways to resolve their dilemmas; the more dilemmas they report, the more ways they use in trying to resolve them. The number of respondents who had discussed their dilemmas with their clients also impressed us. It would appear that our younger, less experienced alumni subsample are more likely than the supervisors to discuss their dilemmas with their clients (supervisors were significantly older [t = 2.7, p < .001] and had more grouped years of experience in human service [X^sup 2^ = 32.5, df = 7, p < .001] than the alumni). Seventy percent of the alumni, as compared with 46% of the supervisors, discussed the dilemma with the client (X^sup 2^ = 8.05, df = 1, p < .005). What does this mean? Is it a sign that workers are empowering their clients by involving them in the process of resolving dilemmas? These data suggest important questions about how clients are being involved in the process of discussing confidentiality.
Somewhat disappointing, although consistent with other studies (Holland & Kilpatrick, 1991; Kugelman, 1992; Landau, in press), is the low percentage of respondents who referred to the NASW Code of Ethics as they attempted to handle their dilemmas. The respondents in our sample, especially the 1992 graduates, had been "taught" the Code of Ethics in a number of foundation and advanced courses, yet somehow were not using it as a resource in working though their dilemmas (Loewenberg & Dolgoff, 1992). Goldstein (1998) has raised issues about how we as social workers teach and learn about ethics. He urges that we move beyond a focus on substantive knowledge about the Code of Ethics to a more experiential pedagogy. Teaching about the Code is not enough. There is a need for dialogic and ontologic learning about ethics that occurs in the "here and now in the immediacy of the human encounter" (p. 250).
Implications for Practice and Research
This study serves as an initial effort to identify social workers' beliefs, practice actions, and agency policies regarding confidentiality. It is also a preliminary examination of the ethical dilemmas that workers report facing and how they handle them. With a relatively small and specific sample of experienced masters-level respondents, our findings cannot be generalized to all social workers. However, a profile of these 152 workers' experiences does begin to emerge and suggest implications for practice, research, and practitioner dialogue. These implications include recognition that ethical dilemmas related to confidentiality are not exceptional, but expectable; the need for comprehensive written policies about limits to confidentiality that extend beyond legally determined requirements to include collaborative relationships, third-party payers, and discussion of case material in supervision; review and discussion of written policies with each client as standard procedure, not only at the beginning of treatment but also throughout the course of the therapeutic work; support and training for supervisors on ethical dilemmas related to confidentiality and on the processes of identifying and resolving ethical dilemmas; training on ethical dilemmas related to confidentiality focusing on the areas in which practitioners indicate need for knowledge, such as collaborative relationships, legal issues, and managed care; forums for discussion of the nature of confidentiality and of ethical issues; and further research conducted by social workers on the nature of ethical dilemmas related to confidentiality in social-work practice and the process of identifying and resolving these dilemmas.
While there are clear challenges presented by managed care and the information requirements that third-party payers present, these are not the only challenges we face. Many of our challenges relate to who we are as social workers, what we do, where we work, and with whom. We need to better understand the ethical questions that social workers experience in maintaining client confidentiality, the nature of their dilemmas, and the process of resolution. This study serves as a preliminary step in that process.
Footnote
Author's note: Earlier versions of this paper were presented at the Joint World Congress of the International Federation of Social Workers and the International Association of Schools of Social Work, Jerusalem, Israel, July 7, 1998, and at the Council on Social Work Education's 45th Annual Program Meeting San Francisco, California, March 13, 1999. This research was conducted as part of a year-long research seminar in the second year of the MSW program at Simmons College School of Social Work. The following students were co-researchers for this study: Cristin Bengtson, Emily Fenster, Wellington Johnson, Kathleen Kelts, Helena Levine, Carolyn MacDuffie, Anna Rubley, David Sandier, Pamela Sinotte, and Lea Tolin.
References
Appelbaum, E S., Kapen, G., Walters, B., Lidz, C., & Roth, L. H. (1984). Confidentiality: An empirical test of the utilitarian perspective. Bulletin of the American Association of Psychiatry and the Law, 12, 109-115.
Appelbaum, P S., Lidz, C., & MeiseL, A. (1987). Informed consent: Legal theory and clinical practice. New York: Oxford University Press.
Backlar, P (1996). Ethics in community mental health care: Confidentiality and common sense. Community Mental Health Journal, 32(6), 513-518.
Bollas, C., & Sundelson, D. (1995). The new informants: The betrayal of confidentiality in psychoanalysis and psychotherapy. Northvale, NJ: Jason Aronson.
Corcoran, K., & Vandiver, V (1996). Maneuvering the maze of managed care: Skills for mental health practitioners. New York: Free Press.
Csikai, E., & Sales, E. (1998). The emerging social work role in hospital ethics committees: A comparison of social work and chair perspectives. Social Work, 34(2), 233-242.
Davidson, J., & Davidson, T. (1996). Confidentiality and managed care: Ethical and legal concerns. Health & Social Work, 22(3), 208-215.
Dickson, D. (1998). Confidentiality and privacy in social work. New York: Free Press.
Fleck-Henderson, A. (1991). Moral reasoning in social work practice. Social Service Review, 65(3), 186-202.
Fortune, A., & Reid, W (1999). Research in social work (3rd ed.). New York: Columbia University Press.
Gelman, R., Pollack, D., & Weiner, A. (1999). Confidentiality of social work records in the computer age. Social Work, 44(3), 243-252.
Goldstein, H. (1998). Education for ethical dilemmas in social work practice. Families in Society 79(3), 241-253.
Goldstein, R. (1989). When doctors divulge: Is there a "threat from within" to psychiatric confidentiality? Journal of Forensic Sciences, 34(2), 433-438.
Holland, T., & Kilpatrick, A. (1991). Ethical issues in social work: Toward a grounded theory of professional ethics. Social Work, 36(2), 138-144.
Jagim, R. D., Whittman, W D., & Noll, J. O. ( 1978, August). Mental health professionals' attitudes toward confidentiality privilege, and third party disclosure. Professional Psychology 458-466.
Kugelman, W (1992). Social work ethics in the practice arena: A qualitative study. Social Work in Health Care, 17(4), 59-80.
Lako, C. J., & Lindenthal, J. J. (1991). The management of confidentiality in general medical practice: A comparative study in the USA and the Netherlands. Social Science Medicine, 32(2), 153-157.
Landau, R. (in press). Ethical judgment, code of ethics and supervision in ethical decision making in social work: Findings from an Israeli sample. Journal of Applied Social Sciences.
Lindenthal, J. J., & Thomas, C. S. (1980). A comparative study of the handling of confidentiality. The Journal of Nervous and Mental Disease, 168(6), 361-369.
Lindenthal, J. J., & Thomas C. (1982a). Consumers, clinicians, and confidentiality. Social Science and Medicine, 16(3), 333-335. Lindenthal, J. J., & Thomas, C. S. (19826). Psychiatrists, the pub
lic, and confidentiality. The Journal of Nervous and Mental Disease, 170(6), 319-323.
Lindenthal, J. J., & Thomas, C. S. (1986). Some roots, dilemmas and research into confidentiality. In L. Taucredi (Ed.), Ethical issues in epidemiologic research (pp. 139-157). New Brunswick, NJ: University Press.
Lindenthal, J. J., Amaranto, E. A., Jordan, T J., & Wepman, B. J. (1984). Decisions about confidentiality in medical student mental health settings. Journal of Counseling Psychology 31(4), 572-575.
Lindenthal, J., Jordan, T., Lentz, J., & Thomas, C. (1988). Social worker's management of confidentiality. Social Work, 33(2), 157-158.
Loewenberg, F, & Dolgoff, R. (1992). Ethical decisions for social work practice. Itasca, IL: EE. Peacock Publishers.
McGuinn, W (1938). The professional secret in social work. Boston College School of Social Work Studies, 1(1), 1-31. McGuire, J., Toal, P, & Blau, B. ( 1985). The adult client's concep
tion of confidentiality in the therapeutic relationship. Professional Psychology: Research and Practice, 16(3), 375-384. McMahon, M., & Knowles, A. (1995). Confidentiality in psycho
logical practice: A decrepit concept? Australian Pschologist, 30(3), 164-168.
Miles, M., & Huberman, A. (1994). Qualitative data analysis. Thousand Oaks, CA: Sage Publications.
Miller, D., & Thelen, M. (1986). Knowledge and beliefs about confidentiality in psychotherapy. Professional Psychology: Research and Practice, 16(1), 15-19.
Millstein, K., Dare-Winters, K., & Sullivan, S. (1994). The power of silence: Ethical dilemmas of informed consent in practice evaluation. Clinical Social Work Journal, 22(3), 317-329.
National Association of Social Workers. (1996). Code of ethics. Washington, DC: NASW Press.
Nicolai, K., & Scott, N. (1994). Provision of confidentiality information and its relation to child abuse reporting. Professional Psychology: Research and Practice, 25(2), 154-160.
Otto, R., Ogloff, J., & Small, M. ( 1991 ). Confidentiality and informed consent in psychotherapy: Clinicians' knowledge and practices in Florida and Nebraska. Forensic Reports, 4, 379389.
Oyen, E. (1982). The social functions of confidentiality. Current Sociology 30(2), 1-37.
Petrila, J., & Sadoff, R. (1992). Confidentiality and the family as caregiver. Hospital and Community Psychiatry 43(2), 136139.
Reamer, E (1998). Ethical standards in social work: A critical review of the NASW Code of Ethics. Washington, DC: NASW Press.
Rock, B., & Congress, E. (1999). The new confidentiality for the 21st century in a managed care environment. Social Work, 44(3), 253-262.
Rhodes, M. (1986). Ethical dilemmas in social work. Milwaukee: Family Service America.
Rhodes, M. (1992). Social work challenges: The boundaries of ethics. Families in Society 73( 1 ), 40-47.
Saakvitne, K., & Abrahamson, D. (1994). The impact of managed care on the therapeutic relationship. Psychoanalysis and Psychotherapy, 11(2), 181-199.
Sabin, J. E. (1997). What confidentiality standards should we advocate for in mental health care, and how should we do it? Psychiatric Services, 48(1), 35-36,41.
Schmid, D., Appelbaum, P, Roth, L. H., & Lidz C. (1983). Confidentiality in psychiatry: A study of the patients' view. Hospital and Community Psychiatry 34(4), 353-355.
Shuman, D. (1985). The origins of the physician-patient privilege and professional secret. Southwestern Law Review, 39, 661687.
Somberg, D., Stone, G., & Claiborn, B. ( 1993 ). Informed consent: Therapists' beliefs and practices. Professional Psychology-Research and Practice, 24(2), 153-159.
Smith-Bell, M., & Winslade, W (1994). Privacy, confidentiality, and privilege in psychotherapeutic relationships. American Journal of Orthopsychiatry, 64(2), 180-193.
Strom-Gottfried, K. (1998). Informed consent meets managed care. Health and Social Work, 23(1), 25-33.
Strein, W, & Hershenson, D. (1991). Confidentiality in nondyadic counseling situations. Journal of Counseling and Development, 69, 312-316.
Wilson, S. (1978). Confidentiality in social work. New York: Free Press.
Winslade, W, & Ross, J. (1985). Privacy, confidentiality and autonomy in psychotherapy. Nebraska Law Review, 64, 678636.
AuthorAffiliation
Kathleen Millstein is associate professor, Simmons College School of Social Work 51 Commonwealth Avenue, Boston, MA 02116.
Original manuscript received: July 14, 1999 Revised: January 10, 2000
Accepted: January 19, 2000
Copyright Manticore Publishers May/Jun 2000
By: Kathleen Millstein
Families in Society 81.3 (May/Jun 2000): 270-282.
Abstract (summary)
The objective of this study was to learn more about practitioners' ethical dilemmas In maintaining client confidentiality. A mailed, anonymous survey of experienced social-work practitioners was conducted to gain an understanding of how practitioners address confidentiality issues in their work with clients and to identify specific areas in which practitioners experience ethical dilemmas related to confidentiality, the factors contributing to these dilemmas, and the resources used by practitioners to resolve them.
Headnote
Social workers have long concerned themselves with confidentiality and its importance to practice. In 1922, social workers created their Code of Ethics, a major precept of which is the protection of confidentiality, defined as the regulation, both legal and ethical, that protects the client's rights of privacy. Social workers are confronting many ethical issues related to confidentiality, such as increasing demands for accountability, mandated duty-to-protect or -warn provisions, expanding court involvement in professional decision making, and widening access to information in records through the expanded use of computer technology. The objective of this study was to learn more about practitioners' ethical dilemmas in maintaining client confidentiality. A mailed, anonymous survey of experienced social-work practitioners was conducted to gain an understanding of how practitioners address confidentiality issues in their work with clients and to identify specific areas in which practitioners experience ethical dilemmas related to confidentiality, the factors contributing to these dilemmas, and the resources used by practitioners to resolve them. Major issues of practitioner concern and specific areas in which practitioners felt a need for more resources, education, and policy clarification were identified.
CONFIDENTIALITY CAN BE DESCRIBED as the regulation, both legal and ethical, that protects the client's rights of privacy. It refers "broadly to legal rules and ethical standards that protect an individual from unauthorized disclosure of information" (Smith-Bell & Winslade, 1994, p. 184). Confidentiality creates boundaries around the client/worker relationship. It also creates boundaries around information outside of that relationship, including information that can be communicated within, between, and among agencies (Rhodes, 1992).
As social workers, we define and balance our obligation to safeguard confidential client information with the rights of the individual, organization, and community to access that information (Bollas & Sundelson, 1995; Goldstein, 1989; McMahon & Knowles, 1995; Saakvitne & Abrahamson, 1994; Strein & Hershenson, 1991 ). We maintain a dual focus on "individual well-being and the well-being of society" (National Association of Social Workers [NASW], 1996, p. 1). In practice, we are seldom able to offer our clients "absolute confidentiality" in which "what data learned or observed by a social worker stay with that individual and are never passed on to anyone" (Wilson, 1978, p. 3). Instead, we offer a "relative" contract for confidentiality. We promise that information that is shared beyond the confines of the client-social worker relationship is shared responsibly, in the interest of providing client services, and with client understanding and consent consistent with legal mandates (Strom-Gottfried, 1998; Dickson, 1998).
Social workers today confront many ethical issues related to confidentiality. Our ability to protect confidentiality is diminished by increasing demands for accountability, mandated duty-to-protect or -warn provisions, expanding court involvement in professional decision making, and widening access to client record information through the requirements of thirdparty payers and the expanded use of computer technology. These complexities are reflected in the fact that "Privacy and Confidentiality" is now the longest section in the newly revised Code of Ethics (NASW, 1996).
The managed-health-care environment in which we find ourselves does not create all of our ethical dilemmas. Even before the days of managed care, social workers were unable to offer their clients unlimited confidentiality (McGuinn, 1938; Oyen, 1982). Managed care merely focuses our attention on ethical and practice issues that have always existed in our roles as professionals: How do we define our responsibilities to our clients, to ourselves, to society, and to funding sources when it comes to maintaining confidentiality? What do we tell and not tell our clients about the limits to confidentiality? Social workers struggle to define the nature of, and limits to, confidentiality in their varied and complex practice settings.
Given the challenges we face, it is striking that there is so little social-work research about ethical issues related to confidentiality in social-work practice. Research from the fields of psychology and psychiatry informs us that clients value confidentiality from their therapists (Schmid, Appelbaum, Roth, & Lidz, 1983; Appelbaum, Kapen, Walters, Lidz, & Roth, 1984; McGuire, Toal, & Blau, 1985), feel that it is important in therapy relationships (Miller & Thelen, 1986), and are more willing to seek and to remain in mentalhealth treatment if confidentiality is assured (Appelbaum et al., 1984; Miller & Thelen, 1986). In conjunction with the clients' perspectives, it is a belief of many practitioners that confidentiality is an important factor in maintaining a helping relationship (Jagim, Whittman, & Noll, 1978; Lindenthal & Thomas, 1980, 1986; Lindenthal, Amaranto, Jordan, & Wepman, 1984; Lindenthal, Jordan, Lentz, & Thomas, 1988; Otto, Ogloff, & Small, 1991).
There is also a small body of literature that challenges the value of confidentiality. Oyen (1982) notes that the principle of confidentiality derives from and supports the view of social problems as based in the individual. Keeping client information confidential implies that those in need of mental-health and social services have something to hide and must be protected from the intrusion of others. In addition, society is not provided with the information it needs for a fuller, more comprehensive understanding of clients' problems within the broader social matrix. In this view, protecting client confidentiality maintains the status quo of an individual-based construction of social problems.
A number of studies have attempted to ascertain under what conditions mental-health practitioners might breach confidentiality (summarized in Lindenthal & Thomas, 1986). These studies, conducted primarily by psychologists using samples that do not include social workers, use a series of vignettes representing hypothetical mental-health-practice dilemmas related to breaching confidentiality (Lako & Lindenthal, 1991; Lindenthal & Thomas, 1982a, 1982b; Lindenthal et al., 1984, 1988). Each vignette is followed by a questionnaire that attempts to assess what practitioners might do in that situation. They do not identify what practitioners define as an ethical dilemma, the nature of dilemmas that they actually experience, or their actions. Only one study (Lindenthal et al., 1988) focused on social workers' management of confidentiality; it found that social workers were more likely than other professionals to report that they will breach confidentiality when confronting specific clinical situations.
The absence of studies on social workers and social-work practice and the heavy reliance on analogue research, such as Lindenthal and colleague's, leave our profession with little knowledge of how we maintain confidentiality in work with our clients. To develop a better understanding of how social workers address confidentiality issues in direct practice with clients, this study used an exploratory-descriptive research design (Fortune & Reid, 1999). The study was conducted by 10 students under faculty direction as part of a year-long research seminar in the second year of the masters of social work (MSW) program at Simmons College School of Social Work. An anonymous survey sought information about social-work practitioners' beliefs, practice actions, and policies regarding confidentiality. Furthermore, we wanted to identify the ethical dilemmas respondents faced in maintaining client confidentiality, and the ways that they resolved these dilemmas. We defined an ethical dilemma as "a situation in which a practitioner feels pulled in two directions and isn't sure what to do." This definition was drawn from Rhodes' (1986) definition of an ethical dilemma as a "choice in which any alternative results in an undesirable action" (p. xii).
Method
A purposive, nonprobability sample of 372 experienced social-work practitioners was used. The sample included all field supervisors at Simmons College School of Social Work for the 1997/1998 academic year (n = 245) and all graduates of the class of 1992 (n = 127) (1992 graduates who were also field advisors were counted in the alumni sample). Simmons College School of Social Work is a single-- concentration masters-degree program specializing in direct practice with individuals, families, and groups. Both groups, field supervisors and practitioners five years post-graduation, represent a sample of experienced MSW social workers. By using the class of 1992, it was our intention to include a group of more recent practitioners whose post-masters practice experience has been in the current managed-care environment. While clearly limited in generalizability, this method of sample selection offered the researchers ready access to a reliable sampling frame and the opportunity to conduct an initial exploration of the topic in the nine-month framework of the student research experience.
Three hundred and seventy-two questionnaires were mailed. Nine were returned due to incorrect addresses; all of these were in the alumni sample. In addition, one supervisor who was interviewed as part of instrument development was excluded from the sample. Of the 362 possible questionnaires, 153 were returned. One of the returned questionnaires was not validly filled out, yielding 152 (42%) for analysis. Ninety-eight (64.5 % ) of the respondents were supervisors and fifty-four (35.5%) were alumni of Simmons College School of Social Work. The response rate was 41 % for supervisors and 43 % for alumni.
The self-administered, anonymous, structured questionnaire was in five parts: ( 1 ) background information about the respondents; (2) practitioner beliefs about confidentiality in practice; (3) how the respondents address issues of confidentiality with clients; (4) the policies on confidentiality in the practitioner's practice setting; and (5) areas in which respondents have experienced ethical dilemmas and how they handled them. The questionnaire was designed to take approximately 20 minutes to complete.
As part of the development of the questionnaire, six hour-long interviews were conducted with experienced practitioners, with social-work educators and researchers, with an ethical theorist, and with a social worker/lawyer. The purpose of the interviews was to refine and narrow the study's focus and questions and to develop items to include in the questionnaire. Five of the six interviewees also volunteered to pretest the final questionnaire. The questionnaire was further pitoted by seven Simmons College School of Social Work faculty and at a meeting of the Education Committee of the NASW Ethics Commission. The questionnaire was mailed out with a cover letter explaining the study's purpose; three weeks after the initial mailing, a reminder postcard was sent to the entire sample. Questionnaire data were analyzed using descriptive statistics. Responses to open-ended questions on the questionnaire were analyzed using thematic content analysis (Miles & Huberman, 1994).
Findings
All respondents were MSW social workers who had either graduated from, or were field supervisors for, a graduate school of social work that offers a single direct-practice concentration. They were predominately Caucasian (91%), female (84%), experienced (median number of years in human service was "11 to 15 years"; number of years in human services includes all human-service experience, including experience obtained prior to receiving their MSW), and mature (average age, 42).
Ninety-four percent of the respondents were currently employed and worked in a diverse group of settings, with the largest percentage in outpatient mentalhealth clinics (20%). The next most frequently identified practice settings were general hospitals ( 10 % ), schools ( 10 % ), community health centers (9%), and psychiatric hospitals (8%). The least frequently reported employment settings were private practice (4%), health-maintenance organizations (HMOs) (4%), criminal-justice system (3%), and the State Departments of Social Services (2%) and Youth Services ( 1 % ). Respondents had been employed in their current practice setting for an average of 6.2 years.
The greatest percentage of their direct-service time was spent with Caucasian clients (average, 66%), followed by African-American (18%), Latino(a)/Hispanic (9%), and Asian clients (3%). The respondents worked with a predominantly adult population, with an average percentage of 57% of their practice time spent with adults. An average of 19.1 % of their practice time was spent with adolescents, 15.9% with children, and 6.9 % with clients older than age 65. When asked to indicate all the theoretical orientations that inform their practice, over half the respondents (53%) espoused a psychodynamic theoretical orientation, followed by eclectic (52%), client-centered (48%), behavioral (48 % ), systemic (43 % ), and short-term (34%). They used multiple interventions, with individual (97%), crisis (74%), group (70%), family (69%) and case-management (60%) as the most frequently indicated modalities.
Beliefs about Confidentiality
Social workers were asked to state their degree of agreement with seven statements that address beliefs about professional, ethical, and legal obligations to maintain confidentiality; about client expectations about confidentiality; and about the threat third-party payers pose to maintaining confidentiality. Respondents were in greatest agreement (strongly agreed) that confidentiality is necessary for maintaining a therapeutic relationship ( 73 % ), and that a social worker has a professional/ethical (88.7%) and legal (85.4%) responsibility to keep information concerning a client confidential except under circumstances prescribed by Massachusetts and federal law.
There were fewer respondents who strongly agreed that state ( 71.3 % ) and federal (69.5 % ) courts must respect the confidentiality of social worker/client communications. The fewest responses of strongly agree were to the statements that most clients expect that communications with social workers will remain confidential (48.3%) and that the requirements of third-party payers threaten the confidentiality of the therapeutic relationship (47%).
It is of note that, while 73.5 % strongly agreed that they believe that confidentiality is necessary for the therapeutic relationship, only 48.3% of the same sample also strongly agreed that they believe that clients expect that confidentiality will be maintained.
The greatest range of opinion came in response to the statement that third-party payers threaten the confidentiality of the therapeutic relationship. The largest number of respondents who strongly agreed were employed in outpatient mental-health clinics (25%), followed by schools (15%), HMOs (7%), and residential treatment (7%). Those least likely to strongly agree were employed in Department of Youth Services (2%) and the criminal-justice system (2%). The majority of those who were undecided worked in general hospitals (75%). Bivariate analyses of the relationships between characteristics of the respondents such as age, experience, theoretical orientation, preferred modes of intervention, practice setting, and client characteristics and specific beliefs did not yield significant findings.
Addressing Confidentiality with Clients
We asked 10 questions about when, how, and under what special circumstances social workers address confidentiality issues with clients. The response options were "always," "sometimes," "never," and "not applicable" (Table 1).
A third of the sample reported that they always inform clients about confidentiality prior to beginning the first session/client contact. More than 50% reported that they always inform clients during the first session/client contact. Further analysis indicated that 80% of those who always inform about confidentiality prior to the first session also inform again in the first session/client contact (X^sup 2^ = 48.9, df = 16, p < .001 ). These data suggest that most of the social workers in our sample are informing clients early in their work and at more than one time. Yet, 6% of the respondents reported always waiting for an issue to arise before they inform clients about confidentiality.
While no statistically significant relationships were found between practice setting and respondents' reported practices in informing clients about confidentiality, the largest percentage of those who reported that they always inform clients prior to the first meeting worked in mental-health (30%) or health (21 % ) settings. This pattern was also evident for those who informed clients in the first session (24% mentalhealth, 22% health). Respondents who spent a greater percentage of their time working with elders were somewhat less likely to wait until a situation arises to inform (r = .30, p < .001). In addition, the practice of informing clients about confidentiality in the first session was related to respondents beliefs that the social worker has a legal obligation to keep client information confidential (X^sup 2^ = 20.7, df = 8, p < .008 ) and that clients expect that communication with the social worker will be confidential (X^sup 2^ = 16.5, df = 6, p < .01).
In the second category, which addressed how clients are informed about confidentiality, practitioners were asked if they informed their clients verbally, in writing, or with a written form that the client signs. The findings indicate that 79.3% of the respondents reported they always use verbal discussion to inform their clients about confidentiality. Fewer than half of the respondents always inform their clients in writing or use a written form that the client signs. Practice setting was related to whether a respondent informed a client about confidentiality in writing (X^sup 2^ = 56.6, df = 40, p < .04) and had the client sign a form (X^sup 2^ = 71.9, df = 40, p < .001), with these practices more prevalent in health and mental-health settings. There were no significant differences between the experienced supervisors and the alumni for any of these practice behaviors, suggesting that these behaviors are not related to level of experience.
In the final category of practice actions, respondents were asked about three "special circumstances" regarding release of confidential information and when confidentiality is limited because a client is dangerous to himself or others. Eighty-nine percent of the respondents reported that they always have clients sign a form for release of client information when information is requested by other agencies. When asked about informing clients of third-party payer access to client information, of the 66.9% who found this issue applicable, 39.9% responded "always." The largest percentage of the 45 respondents for whom this question was not applicable worked in school settings (22 % ), followed by child welfare ( 18 % ). The responses to the final special circumstance show that 86.7% of the sample always inform clients about the limits to confidentiality should the clients be a danger to themselves or others.
Agency Policies and Resources
While there is no way to know in this study whether respondents' agencies have policies about which the respondents are unaware, a majority of respondents (85%, n = 126) reported that their agencies have policies on confidentiality, most of which are written (87%, n = 111). These written policies are not always shared with the clients. Only 43% inform clients of limits to confidentiality in writing (Table 1). There is a relationship between a respondent's practice setting and whether the setting has a policy (X^sup 2^ = 37.2, df = 8, p < .001 ) and whether it is written (X^sup 2^ = 23.9, df = 8, p < .002). Private practice, fee-for-service, and school settings are less likely to have policies, and these three settings along with criminal-justice and college-counseling settings are less likely to have written policies. Not having a policy was also weakly related to waiting until after the first client contact to inform clients about confidentiality (r = -.25, p < .009).
Seventy-five percent (n = 97) of respondents reported that they were generally satisfied with their agencies' policies regarding confidentiality and thought that they provided sufficient information for their clients to understand their rights and agency responsibilities. Of the 25% who were either not satisfied or unsure, several respondents identified the need for a policy written for clients and others stated the need for clients to be informed about insurer access to client records. There was no relationship between type of practice setting or theoretical orientation and satisfaction.
With regard to other agency resources on confidentiality, 85% of respondents had used their supervision to discuss confidentiality at least once in the past year, with an average of 3.4 times. Nearly half of all respondents (46%) had at least one in-service training on confidentiality in the past year. Finally, 40% of respondents reported that they have an ethics committee in their practice settings. Nine questions were asked about specific components included in agency policies in the respondent's primary work setting. Yes/no responses were requested for each component (Table 2). The policy components most frequently identified were those clearly mandated by law, such as release of information to individuals and agencies outside of the practice setting, legal limits to confidentiality imposed by duty to protect, and mandated reporting statutes. Less frequently reported were components related to release of client records within the agency, and release of client information and/or records to insurers. Even less frequently reported were policies related to discussion of case material with service providers in the practice setting, in case conferences, and in supervision. The least frequently reported policy concerned access to client computer records.
The number of specific policies each respondent reported was counted. The modal number of policy components was five, and the median and mean (SD = 1.8) were six. Higher numbers were correlated with respondents' satisfaction with their agency's policy (r = .29, p < .007), suggesting that respondents were more satisfied with more comprehensive policies. In addition, there was a strong relationship between the number of policy components and whether confidentiality issues had been discussed in the past year in supervision (r = .59, p < .001 ).
It is of interest that, just as there were few relationships between beliefs and practices, there were also few relationships between practice actions and specific policies. For example, it might be expected that there would be a relationship between working in an agency with a policy on mandated reporting and respondents informing clients about limits to confidentiality should they be a danger to self or others. This was not the case.
Ethical Dilemmas Related to Confidentiality
Identification of Dilemmas. A major objective of this study was to identify the areas and circumstances in which social workers experience ethical dilemmas related to confidentiality. From our reading, from class discussions, and from the interviews that were conducted as part of constructing the questionnaire, we identified nine major categories in which social workers could experience ethical dilemmas, and then we asked specific questions in each of these categories (for example, "I have experienced an ethical dilemma related to client confidentiality when verbally presenting case material... to a supervisor within my practice setting; at a case conference or training within my practice setting; to a supervisor outside of my practice setting; at a case conference or training outside of my practice setting" ). The nine categories were ethical dilemmas related to ( 1 ) presenting case material in case conferences and supervision (four questions), (2) collaboration (three questions), (3) discussing a client with family members and nonproviders either in the client's life or in the practitioner's life (three questions), (4) confidentiality of records (eight questions), (5) insurers (three questions), (6) working with at-risk groups (six questions), (7) specific client problems (14 questions), (8) specific age groups (four questions), and (9) working with clients who are similar or different from the practitioner (six questions).
There were a total of 52 questions. Each of the questions received a positive response from at least one respondent. For 18 of the 52 questions, 25 % or more of our respondents indicated that they had experienced an ethical dilemma related to maintaining confidentiality. It is these 18 questions and the categories under which they were presented that are summarized in Table 3. Collaborative relationships both inside and outside of the agency, and with family members and friends in the client's life, are a major area in which practitioners are reporting difficulty maintaining confidentiality. These dilemmas were not related to particular practice settings. Respondents frequently reported ethical dilemmas related to insurers, especially concerning requests to justify/extend treatment, work with clients who are potentially court involved, and writing for client records. Specific atrisk groups (e.g., clients exhibiting nonsuicidal self-destructive behavior, clients engaged in illegal activity) and client problems (e.g., trauma, mental illness, child welfare) also offered ethical challenges for our sample.
The seventh most frequently responded question addressed ethical issues when a client provides information about another mental-health professional's unethical behavior. Almost a third of the respondents had at least one client report a practitioner's unethical behavior. The least frequently reported areas in which ethical dilemmas were experienced were presentation of case material for publication (4%) and work with clients with employment issues (4%).
In an effort to understand the degree to which respondents were experiencing dilemmas, we added the number of dilemmas each respondent had identified. The average number of dilemmas was 11 (SD = 6.9). In addition, there was a positive relationship between the number of dilemmas experienced and the number of times our respondents discussed ethical issues related to confidentiality in supervision during the past year (r = .37, p < .01).
Handling Dilemmas. Finally, we asked the social workers in our sample to indicate how they handled their ethical dilemmas when they occurred. Table 4 lists the 13 options offered in response to this question. While there was a distribution of responses across the 13 options, the respondents overwhelmingly reported that they handled their ethical dilemmas through consultation with supervisors. This finding is consistent with earlier data about the importance of supervision as a resource for practitioners. Practitioners in our sample also used colleagues as resources. In addition, 55% of respondents discussed ethical dilemmas with the client. Only 6.6% responded that they did not realize that the situation was a dilemma until it was resolved. These respondents were somewhat more likely to report that their agency did not have a policy on confidentiality (r = -211, p < .05 ). With few exceptions, ways of handling dilemmas were not related to practice setting, theoretical orientation, years in current practice setting, policies, or practices.
We also posed an open-ended question about the areas in which respondents would like further knowledge to help them identify and resolve their ethical dilemmas. Forty-nine percent of the total sample responded (n = 74). The most frequently identified areas in which practitioners wanted information were concerning HMOs and insurers (17.6%) and legal issues, especially court subpoenas and privileged communication (20.3 % ). The next most frequently identified area involved what to tell families when working with children, adolescents, and vulnerable adults (14.7%).
To get a very general sense of client concerns about confidentiality, we asked respondents whether their clients had expressed concerns about confidentiality during the past year. Ten percent reported that their clients frequently expressed concerns, and an additional 56% reported that their clients sometimes expressed concerns. Only 4% had not had a client express concern during the past year.
Discussion
The social workers in this sample were in agreement about their professional, legal, and ethical responsibilities to maintain client confidentiality. There was comparably less agreement about state and federal courts' obligations to respect client confidentiality. Unfortunately, we did not have a question that ascertained the respondents' level of knowledge of state and federal regulations. As a result, these findings may be more reflective of their knowledge of the laws regulating privileged communications-those communications that cannot be shared in a court of law (Dickson, 1998)-than of their beliefs about the nature of state and federal responsibility (Jagim et al., 1978; Otto et al., 1991 ).
As shown by our respondents, most social workers also continue to believe that confidentiality is important for a helping relationship. Yet, they are no longer confident that their clients expect their conversations with practitioners to be confidential. It may be that workers believe that clients have become realistic about the possibility of "absolute" confidentiality. However, the differences in these beliefs may be due to the wording of our question. We asked about client expectations; it would be interesting to speculate about how respondents would have answered if we had used "believe" or "value" instead of "expect." Several studies (Schmidt et al., 1983; Appelbaum et al., 1984; McGuire et al., 1985) have reported that the majority of clients value confidentiality and believe that it should be maintained (Miller & Thelen, 1986). The greatest range of opinion came in response to the statement about the threats of third-party payers to client confidentiality. This finding is difficult to interpret; it may be a result of respondents working in practice settings in which insurance is not an issue.
Most of the social workers in our sample report that, in practice, they are verbally informing clients about confidentiality early in the work. They are functioning in concert with the Code of Ethics, which states that "social workers should discuss with clients and other interested parties the nature of confidentiality and limitations of clients' rights of confidentiality" (NASW, 1996, p. 11). A third of the respondents always inform clients prior to the first session/client contact. This figure is lower than data from Somberg et al.'s ( 1993 ) study of psychologists, in which 59.8 reported that they inform clients prior to the first session/client contact. Some are informing clients about confidentiality at more than one time. Appelbaum et al. (1987) describe this as a "process model" in which clients "receive information over time in a fashion that allows it to be contemplated, shared and assimilated" (p. 157). This model has also been suggested in recent social-work literature (Strom-Gottfried, 1998; Somberg et al., 1993) and in the Code of Ethics, which recommends informing clients "as soon as possible in the social-worker client relationship and as needed throughout the course of the relationship" (NASW, 1996, p. 11 ). There is still, however, a small percentage who are not informing clients until an issue arises, and these data are potentially worrisome. If practitioners are waiting to inform clients about confidentiality until an issue arises, there is an increased chance of the practitioner experiencing ethical dilemmas regarding confidentiality. The delay could mean that the client has been sharing information without being aware that there are inherent limitations to confidentiality in the client/practitioner relationship.
It is noteworthy that fewer than half of the respondents always inform their clients in writing or use a written form that the client signs. These respondents were more likely to work in mental-health or health settings and less likely to work in settings where there are less formal boundaries around roles and confidentiality, such as schools. Although verbal discussion is valuable and the Code of Ethics does not prescribe beyond "discussion," most current wisdom suggests the importance of having clients receive written material that they sign (Strom-Gottfried, 1998; Davidson & Davidson, 1996; Dickson, 1998). Given that this is the current practice wisdom, we need to better understand what makes it difficult for social workers to carry out this practice. There is a gap between theory and practice.
Most respondents report that their agencies have policies that address issues of confidentiality. Many of these agencies also have written policies. These policies most frequently include components mandated by law. At the time of this study in 1998, it seems striking that there would be any agencies, especially the types of agencies identified by our respondents, without policies. Ideally, all agencies would have policies and these policies would be written (Corcoran & Vandiver, 1996). The least frequently reported policy was one related to client access to computer records. This is an area of increasing concern (Gelman, Pollack, & Weiner, 1999; Rock & Congress, 1999), and many agencies are just beginning to struggle with the complex issues of computer records and managed-care access to these records.
Whether respondents reported that their agency had policies about confidentiality was directly related to practice actions, such as informing clients about confidentiality in writing and waiting until after the first client contact to inform. It was also those respondents whose agencies did not have a confidentiality policy who were more likely not to realize that a situation was a dilemma until it was resolved. In addition, the type of practice setting is an important factor in determining whether an agency has a policy, with private practice, fee-for-service, and school settings less likely to have formal policies. It is also the factor that most directly relates to social workers' practices concerning confidentiality.
The data about the amount of in-service training our respondents were receiving concerning ethical issues were encouraging. The finding that two-fifths of the respondents' agencies had ethics committees is especially noteworthy, because these committees are a relatively new agency resource. They offer an institutional structure for anticipating and resolving ethical dilemmas (Csikai & Sales, 1998).
However, in spite of the agency policies, education, and support, experienced practitioners in our sample were struggling in their practices with a number of dilemmas related to confidentiality. These ethical dilemmas are "expectable, not exceptional" (thanks to Stefan Krug for this comment). Ninetyeight percent of our respondents reported experiencing at least one ethical dilemma, and the average number of dilemmas experienced was 11. In addition, our data would suggest that social workers feel that confidentiality is a client concern.
The greatest number of ethical dilemmas with confidentiality arose when working in collaborative relationships within and outside agencies, and with client families and others in their lives. While it is not surprising that working with others on behalf of our clients would present social workers with problems about what can be said to whom, the frequency with which such collaborative dilemmas occurs is impressive. Half of the respondents have experienced a dilemma working with family members, and 42% have experienced a dilemma while working on a treatment team outside of their agency. This raises important questions, especially because these areas are not frequently addressed in agency informed-consent policies and not well defined in the Code of Ethics. With diminished resources, there is increased reliance on informal caregivers (Petrila & Sadoff, 1992), on coordinated case-management models of service delivery, and on court involvement. Meeting the information demands of third-party payers is also an area of concern (Davidson & Davidson, 1996; Strom-Gottfried, 1998). Social workers are often in the position of providing just enough information to justify needed services without violating clients' rights to privacy, especially when working with multiple systems and with clients at risk due to trauma, mental illness, and nonsuicidal self-destructive behavior.
These dilemmas represent areas of practice that are directly related to our profession's complex, Janus-- like role. We negotiate complex systems on behalf of our clients, all the while struggling to maintain their autonomy and self-determination. We have a mission as agents of social change and must attend to both "individual well-being in a social context and the wellbeing of society" (NASW, 1996, p. 1). In addition, challenges to dyadic assumptions about confidentiality and violation of these assumptions are inherent to the structure of our work with involuntary clients. A major question rests in how we define confidentiality in social-work practice. The concept of confidentiality derives from the physician-patient model, which carries the underlying assumption of a dyadic, one-to-one interaction (Shuman, 1985; Strein & Hershenson, 1991; Millstein, Dare-Winters, & Sullivan, 1994). Whereas this dyadic model may fit certain segments of mental-health practice, much of social-work practice not in mental-health settings is nondyadic. Group work, family and caregiver interventions, coordination of services, client advocacy, multidisciplinary team practice, and case management are common forms of our jobs as social workers. They all require collaboration. The traditional notion of confidentiality provides little guidance to social workers in these areas (Backlar, 1996; Millstein et al., 1994), yet much of the literature and research are framed within these assumptions. As Sabin (1997) wrote, "[We] have sometimes confused the confidentiality preconditions for successful therapy conducted under the confessional model with the standards appropriate to the entire spectrum of mental health care" (p. 41 ). Contemporary social-work practice is broad and diverse. Different practice areas may require different confidentiality guidelines. Our data support the importance of practice setting in understanding the challenges practitioners face.
In addition, as Oyen (1982) suggests, there may even be a need for a radical reexamination of the importance of confidentiality to social-work practice. Social workers in the daily discharge of their duties often experience dilemmas regarding confidentiality that are based on the competing values of individual versus collective good. As one respondent wrote, "The rights of the individual versus the rights of the community are of grave concern to me at this time"; he went on to describe his struggles with legislation requiring criminal background checks for social workers and community-reporting mechanisms for convicted sexual offenders.
It was unanticipated that ethical dilemmas concerning a client providing information about another mental-health professional's unethical behavior would be so frequently reported. It is a difficult finding to interpret in the context of our study. While the 1996 revisions to the Code of Ethics discuss this issue extensively in standard 2.11 and while there is an excellent explication of its implications in Reamer's review (1998), unethical professional behavior and the challenges it presents for colleagues remain areas that have not been well examined in the social-work literature on confidentiality.
Our study suggests that practitioners are facing significant ethical challenges as they attempt to maintain confidentiality in their work. Practice guidelines in these changing times are evolving at the same time that we continue to make decisions within our own understanding of ethical practice. A respondent wrote, "How and when does one know they've made the right decision in an unclear situation where there may be no defined rule or mandate?"
Social workers need places and opportunities to discuss their concerns (Fleck-Henderson, 1991 ). Some respondents specified their concerns as follows: "I do wish that...conversations about ethics were more common conversation pieces in the office setting" and "the issues around confidentiality seem simple, but in fact they are quite complex. I felt supported more in educational settings and feel less supported and more pressured now to know the answers."
It appears from the data that it is not professionals in ethics or law to whom practitioners turn when faced with an ethical dilemma. Practitioners are using supervisors and colleagues as their main resources for handling their ethical dilemmas. These findings are consistent with those of Landau's (in press) recent survey study of Israeli social workers, which found that 45 % of her respondents resolve their ethical dilemmas in supervision. This need for supervisory and collegial resources comes at a time of diminished emphasis on and availability of supervision and conference time. In addition, supervisors have not necessarily been trained to offer this support to workers. It is also of concern that discussion of case material in supervision is not a component of most agency policies on confidentiality. Clients have a right to understand how and in what ways their privacy is maintained within the agency (Strom-Gottfried, 1998). While the Code of Ethics specifies that "social workers should not disclose identifying information when discussing clients for teaching or training purposes unless the client has consented to disclosure of confidential information" (NASW, 1996, p. 12), this standard is often not applied to supervision.
In addition to supervisors and colleagues, workers are using a broad armamentarium of ways to resolve their dilemmas; the more dilemmas they report, the more ways they use in trying to resolve them. The number of respondents who had discussed their dilemmas with their clients also impressed us. It would appear that our younger, less experienced alumni subsample are more likely than the supervisors to discuss their dilemmas with their clients (supervisors were significantly older [t = 2.7, p < .001] and had more grouped years of experience in human service [X^sup 2^ = 32.5, df = 7, p < .001] than the alumni). Seventy percent of the alumni, as compared with 46% of the supervisors, discussed the dilemma with the client (X^sup 2^ = 8.05, df = 1, p < .005). What does this mean? Is it a sign that workers are empowering their clients by involving them in the process of resolving dilemmas? These data suggest important questions about how clients are being involved in the process of discussing confidentiality.
Somewhat disappointing, although consistent with other studies (Holland & Kilpatrick, 1991; Kugelman, 1992; Landau, in press), is the low percentage of respondents who referred to the NASW Code of Ethics as they attempted to handle their dilemmas. The respondents in our sample, especially the 1992 graduates, had been "taught" the Code of Ethics in a number of foundation and advanced courses, yet somehow were not using it as a resource in working though their dilemmas (Loewenberg & Dolgoff, 1992). Goldstein (1998) has raised issues about how we as social workers teach and learn about ethics. He urges that we move beyond a focus on substantive knowledge about the Code of Ethics to a more experiential pedagogy. Teaching about the Code is not enough. There is a need for dialogic and ontologic learning about ethics that occurs in the "here and now in the immediacy of the human encounter" (p. 250).
Implications for Practice and Research
This study serves as an initial effort to identify social workers' beliefs, practice actions, and agency policies regarding confidentiality. It is also a preliminary examination of the ethical dilemmas that workers report facing and how they handle them. With a relatively small and specific sample of experienced masters-level respondents, our findings cannot be generalized to all social workers. However, a profile of these 152 workers' experiences does begin to emerge and suggest implications for practice, research, and practitioner dialogue. These implications include recognition that ethical dilemmas related to confidentiality are not exceptional, but expectable; the need for comprehensive written policies about limits to confidentiality that extend beyond legally determined requirements to include collaborative relationships, third-party payers, and discussion of case material in supervision; review and discussion of written policies with each client as standard procedure, not only at the beginning of treatment but also throughout the course of the therapeutic work; support and training for supervisors on ethical dilemmas related to confidentiality and on the processes of identifying and resolving ethical dilemmas; training on ethical dilemmas related to confidentiality focusing on the areas in which practitioners indicate need for knowledge, such as collaborative relationships, legal issues, and managed care; forums for discussion of the nature of confidentiality and of ethical issues; and further research conducted by social workers on the nature of ethical dilemmas related to confidentiality in social-work practice and the process of identifying and resolving these dilemmas.
While there are clear challenges presented by managed care and the information requirements that third-party payers present, these are not the only challenges we face. Many of our challenges relate to who we are as social workers, what we do, where we work, and with whom. We need to better understand the ethical questions that social workers experience in maintaining client confidentiality, the nature of their dilemmas, and the process of resolution. This study serves as a preliminary step in that process.
Footnote
Author's note: Earlier versions of this paper were presented at the Joint World Congress of the International Federation of Social Workers and the International Association of Schools of Social Work, Jerusalem, Israel, July 7, 1998, and at the Council on Social Work Education's 45th Annual Program Meeting San Francisco, California, March 13, 1999. This research was conducted as part of a year-long research seminar in the second year of the MSW program at Simmons College School of Social Work. The following students were co-researchers for this study: Cristin Bengtson, Emily Fenster, Wellington Johnson, Kathleen Kelts, Helena Levine, Carolyn MacDuffie, Anna Rubley, David Sandier, Pamela Sinotte, and Lea Tolin.
References
Appelbaum, E S., Kapen, G., Walters, B., Lidz, C., & Roth, L. H. (1984). Confidentiality: An empirical test of the utilitarian perspective. Bulletin of the American Association of Psychiatry and the Law, 12, 109-115.
Appelbaum, P S., Lidz, C., & MeiseL, A. (1987). Informed consent: Legal theory and clinical practice. New York: Oxford University Press.
Backlar, P (1996). Ethics in community mental health care: Confidentiality and common sense. Community Mental Health Journal, 32(6), 513-518.
Bollas, C., & Sundelson, D. (1995). The new informants: The betrayal of confidentiality in psychoanalysis and psychotherapy. Northvale, NJ: Jason Aronson.
Corcoran, K., & Vandiver, V (1996). Maneuvering the maze of managed care: Skills for mental health practitioners. New York: Free Press.
Csikai, E., & Sales, E. (1998). The emerging social work role in hospital ethics committees: A comparison of social work and chair perspectives. Social Work, 34(2), 233-242.
Davidson, J., & Davidson, T. (1996). Confidentiality and managed care: Ethical and legal concerns. Health & Social Work, 22(3), 208-215.
Dickson, D. (1998). Confidentiality and privacy in social work. New York: Free Press.
Fleck-Henderson, A. (1991). Moral reasoning in social work practice. Social Service Review, 65(3), 186-202.
Fortune, A., & Reid, W (1999). Research in social work (3rd ed.). New York: Columbia University Press.
Gelman, R., Pollack, D., & Weiner, A. (1999). Confidentiality of social work records in the computer age. Social Work, 44(3), 243-252.
Goldstein, H. (1998). Education for ethical dilemmas in social work practice. Families in Society 79(3), 241-253.
Goldstein, R. (1989). When doctors divulge: Is there a "threat from within" to psychiatric confidentiality? Journal of Forensic Sciences, 34(2), 433-438.
Holland, T., & Kilpatrick, A. (1991). Ethical issues in social work: Toward a grounded theory of professional ethics. Social Work, 36(2), 138-144.
Jagim, R. D., Whittman, W D., & Noll, J. O. ( 1978, August). Mental health professionals' attitudes toward confidentiality privilege, and third party disclosure. Professional Psychology 458-466.
Kugelman, W (1992). Social work ethics in the practice arena: A qualitative study. Social Work in Health Care, 17(4), 59-80.
Lako, C. J., & Lindenthal, J. J. (1991). The management of confidentiality in general medical practice: A comparative study in the USA and the Netherlands. Social Science Medicine, 32(2), 153-157.
Landau, R. (in press). Ethical judgment, code of ethics and supervision in ethical decision making in social work: Findings from an Israeli sample. Journal of Applied Social Sciences.
Lindenthal, J. J., & Thomas, C. S. (1980). A comparative study of the handling of confidentiality. The Journal of Nervous and Mental Disease, 168(6), 361-369.
Lindenthal, J. J., & Thomas C. (1982a). Consumers, clinicians, and confidentiality. Social Science and Medicine, 16(3), 333-335. Lindenthal, J. J., & Thomas, C. S. (19826). Psychiatrists, the pub
lic, and confidentiality. The Journal of Nervous and Mental Disease, 170(6), 319-323.
Lindenthal, J. J., & Thomas, C. S. (1986). Some roots, dilemmas and research into confidentiality. In L. Taucredi (Ed.), Ethical issues in epidemiologic research (pp. 139-157). New Brunswick, NJ: University Press.
Lindenthal, J. J., Amaranto, E. A., Jordan, T J., & Wepman, B. J. (1984). Decisions about confidentiality in medical student mental health settings. Journal of Counseling Psychology 31(4), 572-575.
Lindenthal, J., Jordan, T., Lentz, J., & Thomas, C. (1988). Social worker's management of confidentiality. Social Work, 33(2), 157-158.
Loewenberg, F, & Dolgoff, R. (1992). Ethical decisions for social work practice. Itasca, IL: EE. Peacock Publishers.
McGuinn, W (1938). The professional secret in social work. Boston College School of Social Work Studies, 1(1), 1-31. McGuire, J., Toal, P, & Blau, B. ( 1985). The adult client's concep
tion of confidentiality in the therapeutic relationship. Professional Psychology: Research and Practice, 16(3), 375-384. McMahon, M., & Knowles, A. (1995). Confidentiality in psycho
logical practice: A decrepit concept? Australian Pschologist, 30(3), 164-168.
Miles, M., & Huberman, A. (1994). Qualitative data analysis. Thousand Oaks, CA: Sage Publications.
Miller, D., & Thelen, M. (1986). Knowledge and beliefs about confidentiality in psychotherapy. Professional Psychology: Research and Practice, 16(1), 15-19.
Millstein, K., Dare-Winters, K., & Sullivan, S. (1994). The power of silence: Ethical dilemmas of informed consent in practice evaluation. Clinical Social Work Journal, 22(3), 317-329.
National Association of Social Workers. (1996). Code of ethics. Washington, DC: NASW Press.
Nicolai, K., & Scott, N. (1994). Provision of confidentiality information and its relation to child abuse reporting. Professional Psychology: Research and Practice, 25(2), 154-160.
Otto, R., Ogloff, J., & Small, M. ( 1991 ). Confidentiality and informed consent in psychotherapy: Clinicians' knowledge and practices in Florida and Nebraska. Forensic Reports, 4, 379389.
Oyen, E. (1982). The social functions of confidentiality. Current Sociology 30(2), 1-37.
Petrila, J., & Sadoff, R. (1992). Confidentiality and the family as caregiver. Hospital and Community Psychiatry 43(2), 136139.
Reamer, E (1998). Ethical standards in social work: A critical review of the NASW Code of Ethics. Washington, DC: NASW Press.
Rock, B., & Congress, E. (1999). The new confidentiality for the 21st century in a managed care environment. Social Work, 44(3), 253-262.
Rhodes, M. (1986). Ethical dilemmas in social work. Milwaukee: Family Service America.
Rhodes, M. (1992). Social work challenges: The boundaries of ethics. Families in Society 73( 1 ), 40-47.
Saakvitne, K., & Abrahamson, D. (1994). The impact of managed care on the therapeutic relationship. Psychoanalysis and Psychotherapy, 11(2), 181-199.
Sabin, J. E. (1997). What confidentiality standards should we advocate for in mental health care, and how should we do it? Psychiatric Services, 48(1), 35-36,41.
Schmid, D., Appelbaum, P, Roth, L. H., & Lidz C. (1983). Confidentiality in psychiatry: A study of the patients' view. Hospital and Community Psychiatry 34(4), 353-355.
Shuman, D. (1985). The origins of the physician-patient privilege and professional secret. Southwestern Law Review, 39, 661687.
Somberg, D., Stone, G., & Claiborn, B. ( 1993 ). Informed consent: Therapists' beliefs and practices. Professional Psychology-Research and Practice, 24(2), 153-159.
Smith-Bell, M., & Winslade, W (1994). Privacy, confidentiality, and privilege in psychotherapeutic relationships. American Journal of Orthopsychiatry, 64(2), 180-193.
Strom-Gottfried, K. (1998). Informed consent meets managed care. Health and Social Work, 23(1), 25-33.
Strein, W, & Hershenson, D. (1991). Confidentiality in nondyadic counseling situations. Journal of Counseling and Development, 69, 312-316.
Wilson, S. (1978). Confidentiality in social work. New York: Free Press.
Winslade, W, & Ross, J. (1985). Privacy, confidentiality and autonomy in psychotherapy. Nebraska Law Review, 64, 678636.
AuthorAffiliation
Kathleen Millstein is associate professor, Simmons College School of Social Work 51 Commonwealth Avenue, Boston, MA 02116.
Original manuscript received: July 14, 1999 Revised: January 10, 2000
Accepted: January 19, 2000
Copyright Manticore Publishers May/Jun 2000