Should Institutions that Commit Patients also be Gatekeepers to Information about Civil Commitment?
Implications for Research and Policy
Abstract—Public policies must be designed to meet widely shared values and objectives, but also to enable the evaluation of said policies. However, in the justice-mental health field, there is generally a dearth of evidence demonstrating how laws and policies meet their stated objectives. We use civil commitment research in Québec, Canada, to illustrate how legal provisions and institutional discretion made research in this area almost impossible; how this prevented evaluating the operation of pertinent laws and policies; and how this occurred despite the fact that social research agencies, governmental authorities, community resources, user groups, and health and justice organizations officially encouraged and awaited such research and evaluation efforts. We conclude by suggesting ways to enhance the accountability and evaluability of government interventions in the law-psychiatry interface.
Résumé—Les politiques publiques devraient non seulement être conçues avec l'intention de correspondre à des objectifs et des valeurs largement partagés, mais aussi de sorte à ce qu'il soit possible d'évaluer leur application. Cependant, dans le champ de la justice-santé mentale, il y a généralement un manque de telles démonstrations. Nous utilisons le cas de la recherche sur les hospitalisations psychiatriques involontaires au Québec pour illustrer comment certains articles de lois et le pouvoir discrétionnaire des institutions ont rendu la recherche presque impraticable dans ce domaine; comment, en conséquence, l'évaluation de l'application des lois pertinentes était empêchée; et comment ceci s'est produit en dépit du fait que la recherche et l'évaluation étaient officiellement considérés utiles, répondant aux besoins et attentes des organisations de recherche sociale, des autorités gouvernementales, des ressources communautaires, des groupes de clients, ainsi que des instances des milieux de la santé et de la justice. En conclusion, nous suggérons des orientations visant à améliorer la responsabilité et l'évaluabilité des interventions gouvernementales dans l'interface entre justice et psychiatrie.
Key Words—social science research, civil commitment, confidentiality, ethics committees, policy, politics
It is well known that in the mental health system as a whole certain perspectives — paternalism, the disease model, a view of the "mentally ill" as presenting dangers for society by virtue of their illness, and the appropriateness of using psychiatry as a means of social control — are so pervasive that competing voices are rarely heard or actively discouraged (Boyle, 1990; Newnes, Holmes & Dunn, 1999; Scheff, 1999). This is particularly evident in scientific research that could lead to a questioning of the assumptions, value bases, empirical support, and logical rationality supporting these entrenched perspectives.
Hence we have virtually no research on mental patient  powerlessness (McCubbin & Dalgard, 1999), but much on how to achieve patient compliance with treatment regimen; few researchers investigating the social determinants of well-being, but thriving research programs trying to identify biological correlates of psychological distress (Breggin P. R. & Breggin G. R., 1998); almost no study of how society hurts people already psychologically distressed, but many studies focused on the dangerousness of the mentally ill (Hiday, 1997). And, more pertinent to this article, we have extremely few studies of how psychiatry achieves in practice its paternalistic and social control functions.
This article aims to contribute to opening up discussion about what research on the mental health system is impeded, how it is impeded, and what might be done about it. The article focuses on one theme, civil commitment, and one type of barrier, controls over the information required to generate scientific analysis. It examines how some fieldwork conditions and questionable administrative and legal constraints over access to information about what is happening during civil commitment processes can impede socially useful research in this area, and consequently provide obstacles to implementing and evaluating awaited judicial reforms and mental health policies.
Science is not justified entirely by its "usefulness." However, the lack of research on what is actually happening in civil commitment has major implications for current policy orientations in mental health which researchers, various stakeholders (to differing extents), and governments state they support. Understanding what goes on when patients are committed is essential for trying to answer several high priority questions, such as:
• Rights Protection. Does civil commitment avoid unnecessary deprivations of liberty? Do persons subjected to it have adequate due process, including the right to represent their own views and interests (e.g., with the help of advocates or lawyers)? Is the evocation by legal and medical authorities of confidentiality rights of persons subject to civil commitment made in the interests of those persons or rather in the interests of those committing them? How can we protect individual rights to privacy while also advancing the rights to liberty of persons subject to coercion in the mental health system?
• Community Care. Are processes and policies regarding civil commitment congruent with efforts to reduce institutionalization and enhance normalization, social integration, and the transfer of care and support from asylums and hospitals to community? Are "community care" alternatives taken seriously in civil commitment hearings, and actively sought or demanded by judges or tribunal members? Do the professional parties most influential in these hearings have sufficient expertise to ascertain what a severely psychologically distressed person needs to live in the community — beyond control and drugs? Are the available community resources actually sufficient to prevent "unjustifiable" commitments?
Using civil commitment research in Québec as an illustration, we first touch on general problems associated with social science research in hospital and judicial environments. Second, we describe adverse effects of rules and laws ostensibly designed to protect patient confidentiality, but which also advertently or inadvertently serve to impede research. Third, we speculate on why it is difficult to carry out independent social science research on civil commitment. Fourth, we describe how the lack of such research impairs the capacity to evaluate the concrete results of reforms supported in the discourse of most major actors — aimed at enhancing patient rights and empowerment. Finally, we suggest ways to enhance the accountability of policies and programs in this law-psychiatry interface.
This article is not about practice and
provides no advice to practitioners. Rather, it examines barriers to independent
social science research in a field where practitioners — psychiatrists,
psychologists, social workers, physicians, nurses, and others — have an
exceptional amount of power. They exert this power directly over the people
subjected to their treatments ("surplus power-fullness" according to Prilleltensky,
1999), but also over the ability of non-practitioners and science to
know what practitioners and institutions are actually doing. In this environment,
where practitioners and their administrators actually decide what is "useful,"
our discussion takes as its starting point that social science is not only
useful, but essential for understanding what our social institutions 
are up to — and often precisely because its justification does not rely
only upon its "usefulness" (Seidman, 1978).
On Being a Stranger in Strange Lands:
Social Science Research at the Interface of Law and Psychiatry
Prior to any data gathering, fieldwork for social scientists typically requires intense negotiation for access to data sources. This is especially true where social scientists are seen as genuine "outsiders," with different world-views and different scientific, professional or institutional backgrounds from those in the environments to be studied. The possible resulting clash of cultures increases mistrust between social scientists and other actors. For example, some mental health workers may not see the pertinence of research on coercion, if coercion is an unacknowledged prerequisite to some of their interventions. Yet a social science researcher is not led primarily or at all by therapeutic considerations. Social scientists may wish to observe more than what happens within a program, service, or institution to see how those processes are constituted and altered by community or broader social influences. The latter exceed program delivery concerns to incorporate economic, cultural and political influences and their associated dynamics (McCubbin & Cohen, 1999a; Scott, 1985).
This role is especially difficult in the justice-psychiatry interface. Facing powerful institutions, constrained by organizational rules, negotiating with various professionals and authorities, social researchers dealing with hospitals and courts must constantly justify their activities to them — and to the funders and consumers of the research and the usual scientific peer groups. Social scientists independent from the studied organizations can expect a priori distrust from those whose behaviours will be indirectly or directly brought to light by a research project. We have seen such distrust expressed, even overtly, by claims that the researcher is "biased" or — more perniciously, because it cannot be disproved — that the researcher has a "hidden agenda."
We do not know how widespread such attitudes are, nor to what degree they may have impeded worthy social science research in the key areas under discussion. If the problem were rare, we would not expect the literature to say so, unless a controversy were raised. On the other hand, if social scientists have particular difficulty in entering these areas, would we expect to know about it from the published literature? Obstacles could impede or stop research at various stages:
• some funding programs require that access to data be demonstrated prior to funding approval, yet institutions may not choose to take a project seriously, and hence approve access in principle, until after a project receives funding;
• with the growing emphasis on research "partnerships," "usefulness," "pertinence," "transfer — and now appropriation — of knowledge," much available funding is either decided by the institutions to be observed or by peer review in accordance with criteria reflecting institutional priorities rather than broader perspectives of independent social science inquiry;
• reduced access to data can result in the abortion of a project or severely limiting it compared to its original objectives; 
• social scientists may simply migrate to fields where information is more easily obtained.
Confidentiality Ethics and the Paradox of Patient Rights Protection
A major stumbling block for social scientists who wish to observe what is happening within institutional systems of care and control of vulnerable persons is the need to protect their privacy and the confidentiality of their records. This need creates a paradox precisely because these persons are vulnerable. They tend not to have advocates to represent their interests nor organized advocacy groups strong enough to represent their interests in the face of other stakeholders' claims. This is certainly the case of persons who are severely psychologically distressed and subject to coercion within the mental health and justice systems (McCubbin & Cohen, 1996). "Advocacy" on their behalf often consists in putting forth the claims of their relatives, professionals, and other caregivers. This obvious point is rarely acknowledged in the literature — in our view because of paternalistic attitudes which take for granted that others should speak for mental patients (see Lord & Dufort, 1996).
A critical analysis of the mental health system has to accord great importance to what is not being said and what is not being observed as itself revealing something about the purposes of the system. One such purpose, control of deviant persons, necessitates their powerlessness within that system. Yet powerlessness can go completely unremarked — a sign in itself of the enormous distance of users from the places where decisions are made.  There is also no good reason to believe that the interests of those who do influence or control the mental health system — notably health professionals, pharmaceutical companies and family lobby groups — will converge with those of mental patients, yet caregivers are provided immense authority, in practice if not on paper, to speak on their behalf and to protect their rights. It is often they, and not mental patients themselves, who are the gatekeepers to information about what the institutions are doing to them.
The paradox arises when researchers want to observe what those gatekeepers are doing in their functions of care and control over vulnerable persons. Such observation is of course essential to the accountability of caregivers, treaters, and their institutions to the public with respect not only to therapeutic concerns but also to human and civil rights of patients including, but not restricted to, confidentiality. The role of the researcher in this regard hence conflicts with the specific rights of confidentiality and privacy of the patient. The nature of this conflict has not been appreciated in the psychiatry bioethics literature, in which conundrums involving research, risks and confidentiality have been discussed for more than three decades. That literature has always sought to balance individual safety and privacy with the prospect of therapeutic innovation — rather than with rights in general. Serious abuses of psychiatric patients for research purposes led to formal mechanisms to approve research protocols in order to protect research subjects (Rothman, 1991). These typically involve institutional ethics committees (or "institutional review boards") and, for the smattering of research into civil commitment procedures, the judges who decide the cases.
Ethics Committees as Information Gatekeepers
Today, hospitals and large care institutions in most jurisdictions have ethics committees to deal with ethical issues brought to their attention and to consider requests for access to patients or their records for research purposes. Such committees are usually made up almost entirely of hospital staff, and may also involve one outside bioethicist or lawyer, or a member of the "public." It is questionable whether those with the most day-to-day power over patients should also be the gatekeepers to information about what they are doing to them and hence in effect the only guarantors that their rights are met. Such a circumstance creates not only a conflict of rights but also an obvious conflict of interests which could make any conflict of rights moot. It is hospital representatives, for example, who petition to maintain psychiatric patients as involuntarily committed, or who oppose requests of patients before tribunals or courts to be set free — typically with the aid of assessments from psychiatrists on their staff (Dallaire, Morin & McCubbin, 1999).
In such conditions, requests made by social researchers to institutional ethics committees to obtain access to specific information kept in hospital records may not be simply evaluated on the basis of confidentiality protection procedures planned in the research design or of the methodological necessity of obtaining particular confidential data. These requests are sometimes evaluated on the basis of the researchers' own scientific objectives and intentions — also extended to their partner organizations (see Malavoy, 1998). This was the experience of Morin, Dallaire, McCubbin and Cohen (1999), as documented in their files, during fieldwork on a study of civil commitment procedures in Québec. This project received prestigious funding from a government agency subsequent to a highly competitive peer reviewed selection process (25% success rate) evaluating pertinence, rigour, and detailed procedures to protect confidentiality. The researchers were then refused access to data by several hospital ethics committees, necessitating major revisions in the research objectives and delaying the project. Interestingly, most of these refusals, including some reversals of prior approvals, followed shortly after the directors of psychiatric wards of the region's hospitals decided in a meeting that they should not cooperate with the research.
In most cases grounds for refusing access were not clearly specified to Morin and colleagues, except to state the obvious fact that the data were confidential. If there were ethical problems involving procedures to protect patient confidentiality, they were not explained. One hospital refused access on the grounds that the research would breach the medical profession's "exclusive prerogative of medical practice evaluation." Apart from the possibly questionable grounds for such a prerogative, and its unfortunate impacts where a profession is given state powers, the study was clearly specified to be a description, not an evaluation — of judicial rather than medical practices.
Several hospitals questioned the pertinence of the research, and one verbally expressed concerns regarding the project's methodology as not meeting standards for clinical research (the project followed a partly qualitative sociological methodology, as devised by two experienced researchers with doctorates in sociology). These concerns are not easily reconcilable with the role of an ethics committee, which is not to evaluate the worth of proposed research (already assessed by the panels of experts from the funding agencies) but rather to protect the rights of persons who are vulnerable because of illness, disability, or institutional control (see Corsino, 1996; Fleetwood, Arnold & Baron, 1989; Wilson & Steibelt, 1985).
Where an ethics committee might correctly play a role in assessing the potential scientific value of a research project might be in scrutinizing research which entails risks for research subjects who are unable to give full informed consent on their own behalf. We are not referring here to such circumstances, where an ethical decision requires a weighing of risks and benefits to subjects as individuals and as a class (Freedman, 1987; see Weisstub, 1998). Such concerns rarely arise in research which relies upon documents, observations and interviews (as opposed to research or experimentation involving invasion of the body or psyche). Rather, the main risks to patients in social science research concern protection of privacy, for which detailed methodologies exist.
Throughout the process of negotiating access to data, Morin and colleagues received many indirect indications of the disapproval of hospital information gatekeepers — or of those behind them — of the participation in the research project of a patient rights advocacy organization. This government-funded organization was mandated to provide to users of the mental health system and especially those held involuntarily, advocacy and information about legal rights. The researcher from that organization was well-known to the hospital psychiatrists and administrators as a patient advocate and the organization's coordinator. Concerns about the "hidden agenda" of the project or about conflict of interests given this researcher's participation were expressed verbally but never in writing. That is probably due to the obvious fact that many of those recommending refusal for information access themselves routinely carried out clinical research on the patients they examine and treat, as is common in medical research generally. 
It may also be due to the implicit recognition that if patient rights and advocacy groups are "biased", it is because they emphasize the very values that the institutions, the professions and governments say they aim to enhance and protect: the centrality of the patient. The first principle of advocacy, according to Forster, is that "guidelines for enactment are the interests of the represented person; the advocate is not a neutral, mediating authority" (1998, p. 157). Whether one might object to the objectivity of patient advocates conducting research should be directly addressed if this is the concern. In our view, aspersions of "hidden agenda" with respect to the research interests of patient advocacy organizations (insofar, of course, as they really advocate for patients and are not professionals characterizing helping or treatment practices as "advocacy") is patently absurd — not because these organizations have no agenda, but because their agenda is overt: to represent the interests of patients. Advocates, and researchers who align with them, often have to pay a price for this if it leads to an opposition to the separate interests of institutions and professionals, "especially the risk of incurring resentment and hostility from (mental health) professionals" (Forster, 1998, p. 157).
Similarly, if the issue is that of favouring the interests of patients to the detriment of professionals it should be made explicit so that a morally and politically defensible decision could be made at the public policy level about whose interests should be favoured where they seem to conflict. Certainly the grounds of patient confidentiality are not the place to defend service provider interests.
Judges as Information Gatekeepers
Persons subject to civil commitment procedures are not really "consumers" or "users," in the sense of voluntary recipients of health and social services (McLean, 1995). The public interest with respect to research upon the former persons, and the subsequent necessity to observe documents and procedures generated by the civil commitment process — with careful and vetted safeguards for protection of confidentiality — must be great given the exercise of coercion by the state and deprivation of liberty. While, for example, we may refuse to allow researchers to observe clinical examinations without patient permission, the observation of commitment procedures — without "patient" permission — might be justifiable given the need for accountability of the commitment process to society. This is recognized for criminal hearings, which in most western countries are always open to the public except in exceptional circumstances, and indeed some American states permit live broadcasts of court trials.
Nevertheless, observation of civil commitment procedures in a courtroom, or review of the associated documents, can become almost impossible for independent non-clinical researchers, due to the combination of laws and regulations controlling access to psychiatric reports and testimony, and the discretion that judges can exercise in allowing access to courtrooms, transcripts, and other documents.
The implications of such restrictions were experienced by Morin et al. (1999), in their attempts to gain access to the transcripts of civil commitment hearings. Under the legislation pertinent at the time of data gathering, only two ways existed in Québec to obtain transcripts of court hearings concerning requests for compulsory psychiatric hospitalization or examination: an authorization from each judge who had heard the requests, or a permission from each individual subject to a request. The root of this strict limitation lay in legal provisions stipulating that psychiatric examination reports can only be divulged to the concerned parties (i.e., patients, relatives, or hospitals), otherwise access can be obtained only through a court’s authorization. Because the contents of psychiatric examination reports are orally presented at the hearings, this restriction extended to hearing transcripts as well.
Such restrictions make it practically impossible to construct a reasonably sized representative sample of pertinent hearings spanning several judicial districts, making non-clinical research in this area — meaning research not led by service providers who are also legally designated record keepers of the institutions involved, such as physicians or court professionals — highly problematic. Consequently, a provision requiring the approval of individual judges in order to conduct social research of civil commitment processes is inherently self-defeating in its aim to protect patient privacy: insofar as social science methodology requires either a population of cases, or a sample of cases selected for representativity, it is easier to ensure strict anonymity than, say, when relying upon only a few highly detailed individual case reports, as in clinical research. Even when personal data have been expunged, identifying individuals from case reports is easier when all the cases come not only from specific regions but also specific courts, judges, or hospitals. Thus, providing discretion to authorities at localized sites to approve or deny requests by researchers for data access not only threatens the scientific integrity of research, but also the confidentiality of the human subjects that such discretion presumably aims to protect.
The Dearth of Research on
What Actually Happens During Civil Commitment Hearings
Given the problems that Morin and colleagues had in accessing both hospital and court documents, it will not be surprising that in their province of Québec there were almost no previous studies of civil commitment processes.  Elsewhere, there are only a small handful which include detailed observation or analysis of the process as a whole (involving direct observation of hearings or analysis of transcripts), with the aim of understanding how civil commitment works in practice (a careful review yields Andreassen, 1999; Hiday 1977, 1983; Holstein, 1993; Peay, 1989; Sjöström, 1997; Warren, 1982). This is far less than one would expect considering the over one hundred separate legal jurisdictions in the western world having different civil commitment laws, most of which had been through several reform phases over the last decades (Appelbaum, 1997).
When we also consider the enormous scientific and public interest in the issues surrounding civil commitment (i.e., "dangerousness" of mental patients, the large recent literature on "coercion" of mental patients — or, rather, perceptions of coercion — and, especially, the widespread adoption of "rights-based discourse" [Glendon, 1991]), we must conclude that there is serious under-representation of such research at the interface of mental health and justice. The explanation would not seem to lie in either lack of funding or publication outlets (although these may well be factors). We are left, then, with the suggestion that some institutions operating at the law-psychiatry interface — particularly hospitals and courts — are "padlocking" what goes on inside along with the documentary traces of that activity.
We have demonstrated by the case study of Morin and colleagues (1999) how this can happen, but have found no direct evidence establishing the importance of such padlocking for discouraging research elsewhere (although, as discussed above, this may be expected). However, the "padlocking" argument is buttressed by means, motive and opportunity:
• Means—Every jurisdiction has laws and regulations regulating access to data about what happens to psychiatric patients. These ordinarily provide exemptions for research, leaving blanket discretion to the institions holding the data or to ethics committees within those institutions.  The nature of this discretion is crucial for the feasibility of research. There is some evidence of inconsistency among ethics committee decisions with respect to assessing research protocols (Harries, 1994), reflecting the margin of manoeuvre they have even when clear guidelines exist enabling access for research purposes (Cohen, Dolan & Eastman, 1996).
• Motive—it is normal for professions with statutory powers and authority (judges, psychiatrists, hospital administrators) to wish to avoid the spotlight of independent social science research which may, either by intent or effect, put them in a bad light from some perspectives. It would indeed be surprising if these professions did not use the means available to them to do so. A lack of social science research may be more serious insofar as the professions of law and psychiatry do not provide a check on one another when they meet at the civil commitment interface. However, there is some recent analysis and evidence that in civil commitment processes — including as a crucial element the assessment of "dangerousness" — the legal and psychiatric aims and perspectives converge in a sort of symbiosis and mutual support (Andreassen, 1999; Dallaire, McCubbin & Morin, 1999; Mason & Jennings, 1997).
discussed earlier in this article, the impact of limited access to information
is felt at all stages of the research process — including prior to the
formulation of a research question. The most detrimental effects of barriers
to research data will be in terms of the needed research which was never
initiated because of researchers' perceptions of the forbidding nature
of those barriers.
The Clash of Patient Rights and Professional Prerogatives
If we consider the overall situation depicted in this paper with critical distance – using the observations presented above concerning fieldwork in the mental health-civil justice area as empirical data about data gathering – some pertinent conclusions emerge. First, we observe a paradoxical situation where the enforcement of two kinds of rights are mutually contradictory: in order to enforce rights related to confidentiality (the privacy principle), authorities prevent assessing how other equally important human rights affected by civil commitment measures (liberty, integrity, due process) are impacted. Such a "conflict of rights" has been seen before, as in the "right to treatment" versus the patient’s right to liberty. That conflict occurred mostly because non-patients invoked a "right" of patients to be treated – against their will (McCubbin & Weisstub, 1998).
Behind this ethical paradox lies another, more institutionally concrete, opposition: between patient rights promotion and protection versus claims for professional autonomy. When social science researchers investigating the exercise of power and control meet strong opposition from the public institutions they are investigating — despite the project's approval by scientific peer review and despite the standard adoption of strict measures to preserve confidentiality — then what is at stake is not confidentiality protection but instead professional independence and prerogatives, especially, in the issue of civil commitment, those pertaining to coercive psychiatric practice.  As pointed out by Sjöström (1997) and Scheid-Cook (1990), the overuse of confidentiality rules creates situations where abuses are more likely – or, at least, creates appearances of abuses and, more generally, problems for the credibility of professionals.
This tension illustrates that one cannot expect helping professions to objectively evaluate whether state-sanctioned exercise of power in their practice satisfies the public interest. Allowing those with such power to exercise it without external monitoring is particularly dangerous for society when those upon whom the power is exercised are particularly vulnerable and unable to hold the wielder of the power accountable to them or the public. There is therefore a strong need to conduct independent research in the area of civil commitment, where professionals have a triple role: as caregivers in contractual dyads with patients, as paternalistic protectors of patients from harming themselves, and as social control agents protecting society from persons considered to be dangerous to others due to mental illness. Given the potential for conflict between the roles of care and control (Wells, 1997), and the vulnerability of the persons subjected to care and control in the psychiatric system, it is the responsibility of society — exercised through venues including science, policy and politics, and the media — to ensure that these roles can and are exercised in accordance with the rule of law and the basic values of society. No one profession or institution is capable of exercising this responsibility on behalf of society.
The extent to which the needs, interests, and rights of psychologically distressed persons are actually met by the contemporary mental health system remains a very open question to be explicitly addressed by research (McCubbin & Cohen, 1996). If in fact the system is not geared to meeting the needs of its ostensible clients, we can expect such a system to resist attempts to probe its rationality as a system of care and support for its "clients." Such expectation was borne out in our investigation of civil commitment in Québec, and we argue that elsewhere as well institutional information gatekeepers are likely using what discretion is allowed to them by law to impede independent social science research on civil commitment and coercion in psychiatry. A prima facie argument for that is the surprisingly little published research that has taken a close look at what actually goes on in civil commitment processes — despite the clear importance of these processes for a variety of widely supported government initiatives concerning community care and rights protection and advocacy.
The result may be a virtual padlocking of not only the patients who are committed but also information about the committal of those patients. As a public process it remains unaccountable to the public and cannot be evaluated. Referring to the lack of research on whether the rights of mental patients are actually being protected within the mental health and justice systems, McCubbin and Cohen have put the problem this way:
• Funding Requirements. Research funding organizations must recognize that highly pertinent areas of social research are precisely those where research is difficult to conduct because of obstacles laid by institutional gatekeepers. The peer review process in competitions for funding provides a degree of credibility and support to researchers aiming to get past these gatekeepers. If funding organizations require that researchers demonstrate that permission for access to data has been obtained prior to receiving funding, they will discourage researchers away from areas which are both contentious and where many gatekeepers need to be approached, while making much research feasible only for those within the very institutions that need to be observed.
• Information Gatekeeping. Further to this last point, one must question the utility of having a myriad of information gatekeepers, one for each local instance of a society-wide institution. Hence, permission for access to information for research purposes should be given centrally, e.g., by a government department of health and social services, or by a committee at arms-length from the government and its agencies which can credibly assess scientific, ethics and rights issues. This implies that the approving agency would have the responsibility for evaluating access requests, based on clear guidelines focused on necessity, methodological pertinence and confidentiality protection. Such a policy implies that access to pertinent information would no longer be exclusively controlled by representatives of the legal or medical professions — indeed, psychiatric patients and expatients should have a prominent role in ethics assessment of requests involving mental health services or civil commitment; it is their interests most at stake. Obviously access should not be unreasonably denied when the research project, including its plan for minimizing breaches of privacy and maximizing protection of confidentiality, has already been favourably reviewed by a reputable funding agency and peers.
• Exceptions for Research. Laws regulating mental health services and civil commitment should provide exceptions in confidentiality rules for research which preserves confidentiality according to rigorous guidelines, on the grounds that such research serves both collective and individual interests such as protection of human and civil rights, improvement of services, and better allocation of resources.
Taube & Burkhardt (1997) have suggested an approach which would aim to achieve consistency with the American Psychological Association Ethical Principles (1992) and balance the patient's right to privacy with society's need to know. First, they state that legislation and policy should include explicit provisions providing constrained exemptions from confidentiality provisions for archival research. Second, the rules should indicate that the right to privacy will have to be weighed against the right to know, but with the right to privacy given more weight. This would imply that infringements of privacy should be minimized as much as is feasible and that the need to know must be sufficiently great to allow a minimal abridgment of privacy. Third, consent should be obtained from the individuals whose privacy might be effected, unless the contrary can be justified on the grounds of feasibility and the need to know. Fourth, researcher invasion of privacy should be minimized by, preferably, anonymizing the data prior to supplying it. This will be most feasible for electronic data. Finally, only where researcher access to identifying data can still be justified given the foregoing (with the burden of justification lying with the researcher) should privacy rules be waived.
Taube & Burkhardt do not discuss further means of preserving confidentiality in such a case; we believe that researchers should demonstrate that the methodology limits both their own and others' potential access (accidental or intentional) to identifying information. One such means, followed by Morin et al. (1999), is (1) the denominalization of the information as soon as possible in the data processing stage, and (2) in reporting research results, avoiding unnecessary use of other identifying data such as age, municipality, diagnosis, gender, events, etc. which when put together (as in qualitative research drawing on case studies) could identify the person to others in the person's entourage (as explained in Dallaire & Morin, 1999).
• Patient Rights Protection. Governmental responsibilies for promotion, protection and respect of the rights of psychiatric patients must be translated into concrete measures. This responsibility can no longer be delegated to the helping professions alone (McCubbin & Cohen, 1999b). Both social scientists, because of their interest in contextualized knowledge for its own sake, and users of the mental health system, because it is their rights at stake, have important roles studying how and whether rights are being respected and how the rights of users can be further maintained and advanced. Hence, while the "hidden agendas" of user and advocacy groups might conflict with the interests of those who care and control, from a public policy perspective surely the interests of those groups converge with the principal overt ends of the mental health system.
 The term "mental patient" in critical social science analysis means, or should mean, a person in a dependent patient role with an ascribed mental illness. The reference is to how persons are labeled and "handled," and is not an ontological statement about disease or appropriate "treatment". Similarly, a reference to "mental illness" implies a category in discourse rather than our own characterization of psychological distress or social deviance.
 In this article "institution" is used in the sociological sense as an organized social system with the capacity to maintain survival, reproduce over time, and exert some influence or control over its environment and those within its boundaries - which boundaries do not necessarily consist of walls. "Social science" is understood here to mean the scientific study of social processes - whether from cultural, political, psychological, economic etc., perspectives - as the object of interest.
 For example, in a survey research project by Cohen & Bisson (1997) comparing general practitioners' and psychiatrists' neuroleptic prescription practices, the president of the Québec association of psychiatrists sent a letter to all its members urging them not to participate in the study. This stance was justified on the basis of Dr. Cohen's alleged lack of objectivity and the fact that no physician was on the research team. Analysis of questionnaires received subsequent to that intervention was aborted in order to avoid distortions resulting from it, as described on page 266 of the article. The 36% participation rate in the study led to rejection of the ensuing paper by some public health and psychiatry journals.
 This is demonstrated by the mainstream published history of mental health policy and politics (e.g., Grob, 1991; Rochefort, 1997; Simmons, 1990): it does not affirm that users have little role in setting mental health policy, but once the question is posed one can infer such powerlessness by the fact that there is virtually no discussion of their role. For example, the table of contents in Rochefort's (1997) book has no entry for terms like user, consumer, patient, or survivor, but under interest groups, 21 groups are listed, all of which are professional, institutional, employee, and family groups, aside from a couple of special issue groups concerned with the homeless and civil rights.
 Lehrman and Sharav (1997) provide a detailed report of serious harms being caused to patients today by drug research, funded by pharmaceutical companies, carried out by institutional psychiatrists, and approved by institutional review boards composed mostly of medical researchers. The most serious harms consisted of abrupt drug withdrawal to then compare relapse rates during and after drug administrations. As those authors point out, ethical questions not only arise with respect to the harm to patients who suddenly undergo severe withdrawal and rebound effects, but also with respect to the obvious bias introduced in the research design by sudden rather than gradual withdrawal (see Breggin P. R. & Cohen 1999) which makes the tested drugs look better than they really are. This provides part of the resolution for the puzzle posed by Cohen (1997) as to how claims of improved efficacy of new generations of neuroleptics can be consistent with studies (reviewed by Hegarty, Baldessarini, Tohen, Waternaux, & Oepen, 1994) showing that the relapse rate for schizophrenia has been increasing over the last decade. Lehrman and Sharav (1997) document several instances of quite blatant journal and conference censorship of methodological and analytical reviews and commentaries on these practices.
 The only exceptions to this were a small-scale analysis of legal protection of psychiatric patients (Bergeron, 1981) and our recent study of decisions rendered by Québec's Social Affairs Commission, the administrative tribunal reviewing compulsory psychiatric hospitalizations (Cohen, Thomas, Dallaire, Morin, Fortier & McCubbin, 1998).
 This affects not only researchers, but also the public and policy makers. For example (reported in Canadian Press wire service, 22 Nov. 1999), Dorothy Joudrie, who had been in the Alberta Hospital Edmonton (Canada) for five months, recently spoke to the press to complain about both the physical, mental and emotional abuse of herself by the hospital, and of poor hospital conditions affecting all the patients (missing care programs, no encouragement, no real treatment or therapy; patients baited, humiliated, demoralized, degraded on a daily basis). A spokesman for the Alberta Mental Health Board, responsible for the hospital, refused to discuss the allegations for reasons of "patient confidentiality". While it is certainly true that references by Mrs. Joudrie to her own case does not imply consent to the Board to release further details about her, it is also true that a concern about patient confidentiality has no pertinence to the allegations made by Mrs. Joudrie about the general hospital conditions. Such reliance upon the pretext of confidentiality is extremely unfortunate, given the serious nature of the allegations about widespread abuse and neglect of patients; it effectively removes the hospital and the Board from public accountability and suggests that the real privacy at stake is not the patient's but rather the hospital's.
 We do not object to the necessary autonomy of judges ensuing from their duty to remain independent in order to render impartial decisions, which is a prerequisite to protection and enforcement of patient rights - but the exercise of judicial authority must remain accountable to the public. We also do not object to the autonomy of professionals to contract with clients to provide services desired by those clients - provided, of course, that clients are actually free and able to give informed consent to such contracts.
American Psychological Association. (1992). Ethical principles of psychologists and code of conduct. American Psychologist, 47, 1596-1611.
Andreassen, M. (1999). A comparison between legal and psychiatric statements regarding complaints about commitment: A study carred out in a Danish county 1990-1994. International Journal of Law and Psychiatry, 22, 37-44.
Appelbaum, P. S. (1997). Almost a revolution: An international perspective on the law of involuntary commitment. Journal of the American Academy of Psychiatry and Law, 25, 135-147.
Bergeron, V. (1981). L'attribution d'une protection légale aux malades mentaux [the attribution of legal protections to the mentally ill]. Montreal: Blais.
Boyle, M. (1990). Schizophrenia: A scientific delusion? New York: Routledge.
Breggin, P. R., & Breggin, G. R. (1998). The war against children of color: Psychiatry targets inner city youth. Monroe, ME: Common Courage.
Breggin, P. R., & Cohen, D. (1999). Your drug may be your problem: How and why to stop taking psychiatric drugs. Reading, MA: Perseus.
Cohen, A., Dolan, B., & Eastman, N. (1996). Research on the supervision registers: Inconsistencies in local research ethics committee responses. Journal of Forensic Psychiatry, 7, 413-419.
Cohen, D. (1997). A critique of the use of neuroleptic drugs in psychiatry. In S. Fisher & R. P. Greenberg (Eds.), From placebo to panacea: Putting psychiatric drugs to the test (pp. 173-228). New York: Wiley.
Cohen, D., & Bisson, J. (1997). Médication neuroleptique et risque de dyskinésie tardive: Une enquête auprès de psychiatres et d’omnipraticiens du Québec [neuroleptic medication and the risk of tardive dyskinesia: a survey of psychiatrists and general practitioners in Québec]. Santé mentale au Québec, 22, 263-282.
Cohen, D., Thomas, G., Dallaire, B., Morin, P., Fortier, R., & McCubbin, M. (1998). Savoir, pouvoir et dangerosité civile: Une étude des décisions de révision de cure fermée de la Commission des Affaires Sociales du Québec, 1975-1993 [knowledge, power, and dangerousness: a study of civil commitment review decisions by Québec’s Social Affairs Commission, 1975-1993]. Research report submitted to Québec Social Research Council. Montreal: GRASP.
Corsino, B. V. (1996). Bioethics committees and JCAHO patients' rights standards: A question of balance. Journal of Clinical Ethics, 7, 177-181.
Dallaire, B., McCubbin, M., & Morin, P. (1999). A sociological understanding of "dangerousness" for civil commitment: The symbiosis of the legal and psychiatric systems around the duality of care and control. Submitted.
Dallaire, B., & Morin, P. (1999). Dangerousness and the zone of confluence between the individual and his environment: An analysis of civil commitment hearings in Québec, Canada. GRASP Working Papers Series, 24. Montréal: GRASP.
Dallaire, B., Morin, P. & McCubbin, M. (1999, June). The application of the "dangerousness" criterion for civil commitment in Quebec since the reform of the Civil Code (1994): First empirical results based on systems and problem-centered approaches. Presented to XXIVth International Congress on Law and Mental Health, Toronto.
Fleetwood, J., Arnold, R. M., & Baron, R. J. (1989). Giving answers or raising questions? The problematic role of institutional ethics committees. Journal of Medical Ethics, 15, 137-142.
Forster, R. (1998). Patient advocacy in psychiatry: The Austrian and Dutch models. International Social Work, 41, 155-167.
Freedman, B. (1987). Scientific value and validity as ethical requirements for research: A proposed explication. IRB: A Review of Human Subjects Research, 9(6), 7-10.
Glendon, M. A. (1991). Rights talk: The impoverishment of political discourse. Toronto: Maxwell Macmillan.
Goldbeck, R., Mackenzie, D., & Bennie, P. (1997). Detained patients' knowledge of their legal status and rights. Journal of Forensic Psychiatry, 8, 573-582.
Grob, G. (1991). From asylum to community: Mental health policy in modern America. Princeton: Princeton University Press.
Harries, U. J., Fentem, P. H., Tuxworth, W., & Hoinville, G. W. (1994). Local research ethics committees: Widely differing responses to a national survey protocol. Journal of the Royal College of Physicians of London, 28, 150-154.
Hegarty, J. D., Baldessarini, R. J., Tohen, M., Waternaux, C., & Oepen, G. (1994). One hundred years of schizophrenia: A meta-analysis of the outcome literature. American Journal of Psychiatry, 151, 1409-1416.
Hiday, V. A. (1977). Reformed commitment procedures: An empirical study in the courtroom. Law and Society Review, 11, 651-666.
Hiday, V. A. (1983). Judicial decisions in civil commitment: Facts, attitudes, and psychiatric recommendations. Law & Society Review, 17, 517-529.
Hiday, V. A. (1997). Understanding the connection between mental illness and violence. International Journal of Law and Psychiatry, 20, 399-417.
Holstein, J. A. (1993). Court-ordered insanity: Interpretive practice and involuntary commitment. New York: Aldine de Gruyter.
Lehrman, N. S., & Sharav, V. H. (1997). Ethical problems in psychiatric research. Journal of Mental Health Administration, 24, 227-250.
Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. London: Sage.
Lord, J., & Dufort, F. (Eds.). (1996). Power and oppression in mental health (special issue). Canadian Journal of Community Mental Health, 15(2).
Malavoy, S. (1998). Les scientifiques et la loi: Trop de justice mène-t-il à l’injustice? [scientists and the law: does too much justice lead to injustice?] Interface: La Revue de la Recherche, 19(5), 40-48.
Mason, T., & Jennings, L. (1997). The Mental Health Act and professional hostage taking. Medical Sciences and Law, 37, 58-68.
McCubbin, M., & Cohen, D. (1996). Extremely unbalanced: Interest divergence and power disparities between clients and psychiatry. International Journal of Law and Psychiatry, 19, 1-25.
McCubbin, M., & Cohen, D. (1998a, June). The political economy of mental health: A research agenda. Speech to Radical Caucus of the American Psychiatric Association (APA), APA Annual Conference, Toronto.
McCubbin, M., & Cohen, D. (1998b). Mental health policy from a systemic perspective: Orientations for strategic reform. GRASP Working Papers Series, 18. Montréal: Université de Montréal.
McCubbin, M., & Cohen, D. (1999a). A systemic and value-based approach to reform of the mental health system. Health Care Analysis, 7, 57-77.
McCubbin, M., & Cohen, D. (1999b). The rights of users of the mental health system: The tight knot of power, law and ethics. XXIVth International Congress on Law and Mental Health, Toronto.
McCubbin, M., & Cohen, D. (1998c). Les droits des usagers de services en santé mentale: Le noeud étroit du pouvoir, de la loi et de l'éthique [the rights of mental health service users: the tight knot of power, law and ethics]. Santé mentale au Québec, 23, 212-224.
McCubbin, M., & Dalgard, O. S. (1999). Social determinants of population health: Power, powerlessness and empowerment as determinants of health and well-being. Unpublished.
McCubbin, M., & Weisstub, D. N. (1998). "Meeting the needs of the mentally ill": A case study of the "right to treatment" as legal rights discourse in the U.S.A. GRASP Working Papers Series, 20. Montréal: GRASP.
McLean, A. (1995). Empowerment and the psychiatric consumer/ex-patient movement in the United States: Contradictions, crisis and change. Social Science & Medicine, 40, 1053-1071.
Meyer, J. W., & Scott, W. R. (Eds.). (1992). Organizational environments: Ritual and rationality (Updated Edition). Newbury Park, CA: Sage.
Morin, P., Dallaire, B., McCubbin, M., & Cohen, D. (1999). L'opérationalisation de la notion de dangerosité civile lors des audiences pour ordonnance d'examen clinique psychiatrique et d'hospitalisation psychiatrique obligatoire. [the operationalization of the notion of civil dangerousness in civil commitment and compulsory psychiatric examination hearings]. Research report submitted to Québec Social Research Council (grant #RS-2575-095). Montréal: GRASP & AGIDD-SMQ.
Newnes, C., Holmes, G., & Dunn, C. (1999). This is madness: A critical look at psychiatry and the future of mental health services. Llangarron, UK: PCCS.
Peay, J. (1989). Tribunals on trial: A study of decision-making under the Mental Health Act 1983. Oxford: Clarendon.
Prilleltensky, I. (1999, June). Contesting surplus power-fullness: Towards professional de-powerment in helping relationships. Presented to XXIVth International Congress on Law and Mental Health, Toronto.
Rochefort, D. A. (1997). From poorhouses to homelessness: Policy analysis and mental health care (2nd ed.). London: Auburn.
Rothman, D. J. (1991). Strangers at the bedside: A history of how law and bioethics transformed medical decision making. New York: Basic.
Scheff, T.J. (1999). Being mentally ill: A sociological theory (3rd ed.). New York: de Gruyter.
Scheid-Cook, T. L. (1990). Ritual conformity and organizational control: Loose coupling or professionalization? Journal of Applied Behavioral Science, 26, 183-199.
Scott, W. R. (1985). Systems within systems: The mental health sector. American Behavioral Scientist, 28, 601-618.
Seidman, E. (1978). Justice, values and social science: Unexamined premises. Research in Law and Sociology, 1, 175-200.
Simmons, H. G. (1990). Unbalanced: Mental health policy in Ontario, 1930-1989. Toronto: Wall & Thompson.
Sjöström, S. (1997). Party or patient? Discursive practices relating to coercion in psychiatric and legal settings. Umea, Sweden: Boréa.
Taube, D. O., & Burkhardt, S. (1997). Ethical and legal risks associated with archival research. Ethics & Behavior, 7, 59-67.
Warren, C. A. B. (1982). The court of last resort: Mental illness and the law. Chicago: University of Chicago Press.
Weisstub, D. N. (Ed.). (1998). Research on human subjects: Ethics, law and social policy. New York: Elsevier.
Wells, J. S. G. (1997). Priorities, 'street level bureaucracy' and the community mental health team. Health and Social Care in the Community, 5, 333-342.
Wilson, D. R., & Steibelt, P. (1985). Patients' rights and ethics committee, Douglas Hospital Centre. Canada's Mental Health, 33(3), 24-27.
Affiliations—Michael McCubbin is Associate Researcher, Research Group on Social Aspects of Health and Prevention (GRASP), Université de Montréal. Bernadette Dallaire is Honorary Research Fellow, Institute for Health Research (IHR), University of Salford, U.K. David Cohen is Professor, School of Social Services, Université de Montréal, and Researcher, GRASP. Paul Morin is Coordinator, Montérégie Rights Defense Collective (Québec), and Associate Researcher, GRASP.
Acknowledgments—We would like to thank several anonymous reviewers and Dr. Ursula Harries, IHR, for their comments, and express our appreciation for financial support provided to Michael McCubbin by the Québec Social Research Council (CQRS), by the Canadian Institute for Advanced Research, and by Leverhulme Trust (UK), and to Bernadette Dallaire and Paul Morin by CQRS. This article grew out of the conference presentation: "Dallaire, B., McCubbin, M., Morin, P., & Cohen, D. (1998, Sept.). Impeded research, forbidden reforms? Xth International Conference of the International Association of Health Policy, Perugia, Italy."
Correspondence—Dr. Michael McCubbin, GRASP, Université de Montréal, CP 6128 succ. centre-ville, Montréal H3C 3J7, Canada.