Who Is Identified

The main objects of such requests and the ensuing analyses tend to be large health plans to hospitals, physician groups, individual physicians, and nursing homes. Most of the debate in the past few years has centered on hospitals, especially in the context of the validity and meaningfulness of hospital-specific death rates (Baker, 1992). Generally, arguments in favor of the principle of release of such information on hospitals have carried the day; controversy persists about the reliability, validity, and utility of such information when the underlying data or the sophistication of the analyses can be called into question.

More recently, the debate has turned to release of information on the hospital-based activities of particular physicians—for example, death rates associated with specific surgical procedures. Here the principle of public disclosure also seems to have gained acceptance, again with caveats about the soundness of the analyses and results. Nevertheless, because of the much greater difficulty of ensuring the reliability and validity of such analyses, especially on the level of individual physicians, many observers remain concerned about the possible downside of releasing information on specific clinicians. This criticism is especially pertinent to the extent that this information is a relatively crude indicator of the quality of care in hospitals or of that rendered by individual physicians, especially surgeons.

In the future, attention can be expected to shift to outpatient care and involve the ambulatory, office-based services of health plans and physician groups in primary or specialty care and of individual physicians. In these cases the stance in favor of public disclosure may become more difficult to adopt fully, for three reasons: the problems alluded to above for hospital-based physicians become exponential for office-based physicians; the clear, easy-to-count outcomes, such as deaths, tend to be inappropriate for office-based care because they are so rare; and quality-of-life measures, such as those relating to functional outcomes and physical, social, and emotional well-being, are more significant but also more difficult to assess, aggregate, and report.

Other types of providers and clinicians also must be considered in this framework. These include pharmacies and individual pharmacists; home health agencies and the registered nurses and therapists they employ; and durable medical device companies, such as those that supply oxygen to oxygen-dependent patients and the respiratory therapists they employ. Stretching the public-disclosure debate to these and other parts of the health care delivery environment may seem farfetched; to the extent that their data will appear eventually in databases maintained by HDOs, however, the prospect that someone will want to obtain, analyze, and publicize such data is real. This may illustrate the point raised in Chapter 2 that simple creation of databases may lead to applications quite unanticipated by the original creators.

Finally, some experts foresee the day when HDOs might do analyses by employer or by commercial industry or sector with the aim of clarifying the causes and epidemiology of health-related problems. Cases in point might be the incidence of carpal tunnel syndrome in banks, accidents in the meatpacking or lumber industry, or various types of disorders in the chemical industry. Here the issue is one of informing the public or specific employers in an economic sector about possible threats to the health and well-being of residents of an area or employees in a particular commercial enterprise.

Vulnerability to Harm

The examples above can be characterized by level of aggregation: large aggregations of health care personnel in, for instance, hospitals or HMOs, as contrasted with individual clinicians. The committee believes that, in general, public disclosure can be defended more easily when data involve aggregations or institutions than when they involve individuals. Vulnerability to harm is the complicating factor in this controversy, and some committee members affirm that it should be carefully and thoughtfully taken into account before data on individuals are published.

To an individual, the direct harms are those of loss of reputation, patients, income, employment, and possibly even career.⁶ Hospitals and other large facilities, health plans, and even large groups are less vulnerable to such losses than are individuals. Higher-than-expected death rates for acute myocardial infarction or higher-than-expected caesarean section rates are not likely to drive a hospital out of business unless the public becomes convinced that these rates are representative of care generally and are not being addressed. By contrast, reports of higher-than-expected death rates for pneumonia or higher-than-expected complication rates for cataract replacement surgery could disqualify an individual from participating in managed care contracts and eventually spell ruin for the particular physician.

How one regards harms and gains may depend in part on whether one views public disclosure of evaluative information about costs or quality as a zero-sum game. In a highly competitive market, which may have the characteristics of a zero-sum game, clear winners and losers may emerge in the provider and practitioner communities. Furthermore, in theory this is what one would both expect and desire. Nevertheless, when markets are not highly competitive—for instance, when all hospital occupancy rates are high or when the number of physicians in a locality is small—the information may less directly affect consumer choice, although it may well influence provider behavior by changing consumer perceptions. In this situation, clear winners and losers are neither expected nor likely, but establishing benchmarks that all can strive to attain should, in principle, contribute to better performance across all institutions and practitioners.

Methodological and Technical Issues

Several factors influence the degree of confidence one can have in the precision of publicly disclosed analyses, and this dictates how securely one can interpret and rely on published levels of statistical significance and confidence intervals and generalize from published information. Two factors involve the quality of the underlying data and the analytic effort, as introduced in Chapter 2. Others, discussed below, involve the level of aggregation in published analyses, the appropriateness of generalizing from published results to aspects of care not directly studied, and the difficulty of creating global indexes of quality of care.

In the committee's view, proponents of public disclosure have an obligation to insist that the information to be published meet all customary requirements of reliability, validity, and understandability for the intended use. Such requirements vary, to some degree, according to the numbers of cases or individuals included in the report-that is, according to the level of aggregation, from a single case or physician to dozens or hundreds of cases from multiple hospitals. When HDOs cannot satisfy these technical requirements, they should not publish data in either scientific journals or the public media. The committee was not comfortable with the idea that publication might go forward with explanatory footnotes or caveats, on the grounds that most consumers or users of such information are unlikely to accord the cautions as much importance as they give to the data themselves and may thus be unwittingly led to make erroneous or perhaps even harmful decisions.

This position may not be sustainable in all cases, however. The New York Supreme Court rejected the argument "that the State must protect its citizens from their intellectual shortcomings by keeping from them information beyond their ability to comprehend" (Newsday, Inc. and David Zinman v. New York State Department of Health, et al.) and ruled that physician-specific mortality rate information be made public pursuant to a Freedom of Information Law request. In this particular case it could be argued that the data and analyses met all reasonable expectations of scientific rigor. In the future, however, one cannot assume this will be the case. One solution in problematic circumstances may be for HDOs to disclose information only at a much higher level of aggregation than that at which the original analyses may have been done.⁷

Generalizability is a related methodologic matter with ramifications for the harms and gains noted above. It refers here to the proposition that information on one dimension of health care delivery and performance will in some fashion predict or otherwise relate to other dimensions of performance. For hospitals, for instance, the thought might be that information on adult intensive care can be generalized to adult cardiac care, pediatric intensive care, neonatal intensive care, or even to orthopedics, obstetrics, or ophthalmology. A similar proposition might hold that information on the management of patients with acute upper respiratory infections in the office setting can in some manner predict care of patients with long-standing conditions such as chronic obstructive lung disease, congestive heart failure, or even diabetes mellitus.

Those involved in public disclosure of evaluative information must take care to reflect expert opinion on this matter-inferences about one aspect of care cannot always successfully be drawn from information, whether positive or negative, about another aspect of care. The point is complex because some extrapolation or generalization may be supportable. For example, good or bad ratings for a hospital on death rates for congestive heart failure or acute myocardial infarction might well be generalizable to that hospital's performance on pneumonia or chronic obstructive pulmonary disease (Keeler et al., 1992b), but they might well be completely irrelevant to ratings for asthma in children, hip replacement, or management of high-risk pregnancies. The committee thus believes that HDOs will have a duty to make clear the limits of one's ability to draw conclusions about quality of care beyond the precise conditions and circumstances reported for analyses.

A related problem involves the understandable desire to reduce several separate measures of quality of care into a single, global index intended to represent the performance of an entire hospital, plan, or individual provider. The presumption is that an index measure will be easier for HDOs to report and for the public to understand. Developing index measures is extremely difficult for conceptual reasons, which mainly relate to the difficulty of aggregating measures that come from a variety of sources or represent disparate variables (essentially an "apples and oranges" or even "apples and giraffes" problem); quantitative and statistical problems are also significant. In practice, developing index measures for quality-of-care analyses rarely if ever has been successful,⁸ although at least one committee member believes that the general level of intrahospital correlation is probably underestimated.

How Information Is Publicly Disclosed

Assuming that some of the above issues have been adequately addressed, one arrives at the question of the content and appearance of publicly disclosed information. The structure, level of detail, and other properties of such information will differ by the disclosure media used, by the nature of the information, by the type of provider or practitioner under consideration, and by the level of confidence that can be placed in the numbers, statistics, and inferences to be presented. Some of the more problematic factors in presenting data are noted here. The committee does not take a formal stand on how these matters might be resolved, however, because it believes that those decisions need to be governed by local considerations.

One difficulty in presenting data involves how and in what order HDOs elect to identify or list institutions, clinicians, or other providers. The most obvious choice is to do so alphabetically. This option has the advantage of making it easy to find a given provider and would probably be the likely approach when publicly disclosed information is purely descriptive. It has the related disadvantage, however, of complicating the task of comparison when the issues of interest involve evaluative information.

Other approaches are nonalphabetic. HDOs might, for instance, order providers of interest on a noncontroversial or descriptive variable; for a given region, these might be the number of beds for institutions, the number of free-standing clinics for HMOs, or the number of primary care physicians for PPOs. This method, however, does not have the advantages of alphabetic ordering and still has the disadvantages noted above. A variant is to sort providers on the basis of an essentially descriptive variable, such as charges for a particular service, that has the potential of some evaluative content.

Yet another option is to list providers in order from high to low (or vice versa) on a particularly sensitive evaluative variable, such as death or complication rates. This option may be the least desirable or the most open to misinterpretation, as exemplified by PrimeTime Live's referral to such an arrangement as a "surgical scorecard" (PrimeTime Live, ABC, June 4, 1992).

Some choices in the category of other-than-alphabetic ordering present special problems or considerations. For one, the distinction implied above between descriptive and evaluative information may be incorrect or not always applicable. What for some consumers may be purely informational, noncontroversial, or irrelevant—for instance, numbers of specialists or fees charged for a procedure—may for others be a significant or decisive matter in choosing or leaving providers. Opting to list providers or practitioners by these variables may thus reflect the biases or predispositions of those publicizing the data and may not serve all consumers equally well.

Rankings—for instance, from highest to lowest on some variable—may imply greater differences than are truly warranted, and indeed may be positively misleading. This is especially the case when adjacent ranks differ numerically but the differences have no clinical meaning or statistical significance (stated degree of certainty). The committee believes that those disclosing such information must indicate where no statistical differences exist between ranks.

Mixed approaches—for instance, grouping into thirds, quartiles, quintiles, or essentially equivalent ranks and then ordering alphabetically within the groups—are of course possible. Scores, indexes, or other combinational calculations may be appealing for publication purposes, but they often have technical or methodologic weaknesses that will be difficult to convey to the public.

Other aspects of public disclosure involve whether information is representational or symbolic. Data can be presented quantitatively as numbers, rates, dollars, and so forth. Alternatively, data can be rendered qualitatively—for example, as one to five stars or dollar signs; open, half-open, or closed circles; or other symbolic figures. Combinations of such approaches are possible. Exactly which approaches convey what kinds of information best, with the least implicit bias, is open to question and deserves empirical study.

Another aspect of disclosure involves how analyses are released. Up to the present time, most observers would expect HDOs to release analyses in printed form; in the future, however, electronic outlets, such as CD-ROM diskettes or computer bulletin boards, may come into play. A computer-based approach may speed information to some audiences, which would be a positive outcome. Depending on the extent to which the underlying data accompany analytic results in computer-based media, however, some opportunity may arise for unauthorized analysis of data that could distort the original information. From the viewpoint of HDOs, this would be an undesirable outcome.

A final set of factors concerns the extent of explanatory information and technical footnotes. When public disclosure relies on judgmental or symbolic approaches, more explanatory and definitional material is probably needed than when information is given in straightforward, nonqualitative ways. Some of this information might best be left to a technical report for researchers and other very knowledgeable readers. In any case, the committee agrees that HDOs involved in public disclosure must make available, in clear language, the key elements of their methods, including discussion of possible threats to the internal validity and generalizability of the work that analysts believe they have dealt with adequately.

Minimizing Potential Harms

Up to this point, the committee has taken an extremely strong pro-disclosure stance toward comparative, evaluative data, but it sees some potential for harm in instantaneous public release of comparative or evaluative studies on costs, quality, or other measures of health care delivery. This might be the case, for example, if those doing such work fail to provide information to hospitals, physician groups, or other study targets in advance of release, even to permit them to check the data or develop responses. Disclosure proponents assume that such studies will be done responsibly, and the public has every right to expect that to be the case. To the extent that is true, the generators of the work will be believed and the public interest will have been served.

What is not clear is how well such initiatives will be carried out and what brake or check will exist to ensure high-quality studies. One option is to require HDOs to exercise formal oversight over their work and, insofar as possible, over work done with data they provide. For instance, HDOs might be expected to impose an expert review mechanism on their own analyses before public release, or to require that such peer review be done for analyses performed by others on HDO data, or both.

Alternatively, reliance might be placed on an essentially market-driven set of checks and balances. This approach holds that poor, biased, or otherwise questionable work will eventually be discovered and those carrying it out discredited, because the "marketplace of ideas" has its own discipline to prevent reckless analysis. Studies cost money and are likely to be done only by organizations large enough to have a reputation to protect; poorly done analyses will be criticized and discounted in the press; analysts will fear diminishment of their professional reputations; and fear of lawsuits for defamation can be a powerful dissuader.

The committee did not, in the end, wish to rely solely on marketplace correctives; it believed that a more protective stance was needed.

RECOMMENDATION 3.3 MINIMIZING POTENTIAL HARM

The committee recommends that, to enhance the fairness and minimize the risk of unintended harm from the publication of evaluative studies that identify individual providers, each HDO should adhere to two principles as a standard procedure prior to publication: (1) to make available to and upon request supply to institutions, practitioners, or providers identified in an analysis all data required to perform an independent analysis, and to do so with reasonable time for such analysis prior to public release of the HDO results; and (2) to accompany publication of its own analyses with notice of the existence and availability of responsible challenges to, alternate analyses of, or explanations of the findings.

This set of recommendations reflects what might be regarded as a fairness doctrine. It holds that an important safeguard for providers, especially individual practitioners, dictates that they be allowed to check their own data and comment thereon.

Meeting a fairness principle could take the form simply of giving the subjects of analyses prerelease copies of the publication or at least of the information about them. Such subjects would not be given a veto over whether their data are used, nor would they necessarily be afforded an opportunity to have their data amended in public studies, but their comments would be maintained by the HDO so that interested parties could review them. The committee assumed that HDOs might well choose to append such comments to their own reports and publications. HCFA adopted this approach for its mortality rate releases, for instance, as did Cleveland Health Quality Choice for its first data release.

The committee actually has gone further, however, to advise that HDOs give providers and practitioners (or their representatives) the relevant data and sufficient time to analyze them, should such requests be made. Because such situations will differ among HDOs in the future, the committee did not develop specific guidance on these points. For example, it cautions that HDOs will have to devise ways, in conveying such data to one requestor, to conceal the identities of other providers or practitioners in the data files to be transferred, but the committee believes that HDOs can accomplish this.

In addition, the committee did not reach any consensus on what constitutes sufficient time, believing that this would vary by the nature, size, and scope of the analysis in question, but common sense about the difficulties of data analysis and fairness to all parties might dictate the length of waiting time; a week's, or a year's, delay would doubtless fail both tests, two months might not. The committee also thought that the time permitted for reanalysis might be longer in the earlier years of HDO operations, when all parties are developing procedures and skills in this area, and for approximately the first year of any new or especially complicated analyses.

This discussion began with the premise that potential harms to institutions and individual practitioners need to be prevented or minimized, and the steps recommended above have that intent. The committee believes, however, that a benefit for HDOs of requiring them to make such data available for review and possible reanalysis is that feedback from providers may reveal problems with data quality and study methods that HDOs would want to remedy.

Promoting Wide Applications of Health-related Data

To this point, the chapter has focused on what HDOs might do internally to analyze and publish information on providers and practitioners in their regions. Consistent with the discussion in Chapter 2, however, HDOs might well be expected to do more in the public interest to promote responsible use of health-related data. Specifically, they can serve as a key repository of data to which many other groups should have access.

RECOMMENDATION 3.4 ADVOCACY OF DATA RELEASE: PROMOTING WIDE APPLICATIONS OF HEALTH-RELATED DATA

To foster the presumed benefits of widespread applications of HDO data, the committee recommends that health database organizations should release non-person-identifiable data upon request to other entities once they are in analyzable form. This policy should include release to any organization that meets the following criteria:

• It has a public mission statement indicating that promoting public health or the release of information to the public is a major goal.
• It enforces explicit policies regarding protection of the confidentiality and integrity of data.
• It agrees not to publish, redisclose, or transfer the raw data to any other individual or organization.
• It agrees to disclose analyses in a public forum or publication.

The committee also recommends, as a related matter, that health database organizations make public their own policies governing the release of data.

In referring to "non-person-identifiable data," this recommendation is intended to protect the confidentiality of person-identifiable information in HDO databases. The latter pertains to specific patients or other individuals and might include persons residing in the community who appear in population-based data files but have not received health care services. The distinction is made because individual practitioners or clinicians, who might well appear in the databases as patients or residents of the area, could be identified or identifiable by their professional roles (in line with earlier recommendations in this chapter). Chapter 4 of this report explores issues of privacy and of confidentiality of person-identifiable data in more detail.

The committee debated at length the desirability and propriety of advocating that HDOs make data available to all requestors, rather than constraining the transfer of data as in the above recommendation. It was uncomfortable with flat prohibitions on all transfers of data, but it was equally uncomfortable with the possibility of open-ended or blanket transfers of data to a wide variety of groups who would not be expected to place public dissemination of information high on a list of organizational objectives. To thread a path through this dilemma, therefore, the committee advocated that HDOs make data available to those entities that can demonstrate their clear goal of public disclosure of descriptive or evaluative information and their ability to realize this goal. HDOs should not place prior restraints on which entities might receive such data simply on the grounds that others may conduct analyses or release findings that dispute those of the HDOs themselves.¹⁰

To characterize such entities, therefore, the committee devised the criteria in the above recommendation for two purposes. The first is to underscore its view that databases held by HDOs (or at least those mandated by law or supported by public funds) should be available for science and the public good. The second is to constrain the use of such data purely for private gain, particularly in anticompetitive actions. Examples of this second concern might be employers, insurers, PPOs, hospitals, or other health care delivery organizations in a given region using data for price collusion. Thus, it is expected that those requesting data tapes from an HDO will perform and publish analyses that can be said to serve science and/or the public interest.

The committee recognized that a tension may develop between the understandable desire on the part of HDOs to hold data until they have completed studies they wish to conduct and the need to be responsive to requests from eligible organizations for reliable, valid, and up-to-date data. Responding to such requests might even delay or prevent studies and publications by the HDOs themselves. The committee believes, therefore, that HDOs might consider developing independent units; one group could be responsible for data management and release of data to authorized recipients, and the other could take the lead for the HDOs' own internal analyses and public disclosure activities.

Requiring Recipients to Protect Data Privacy and Confidentiality

The committee debated at some length the advisability and feasibility of recommending that HDOs require recipients of their data to protect the confidentiality of the information. Ultimately, the committee elected simply to observe that such behavior on the part of HDOs might be desirable; it is certainly desirable on the part of data recipients. It concluded, however, that the practical aspects of insisting that HDOs police the actions of their data recipients were too difficult to make this step an integral part of HDO operations.

Certain kinds of database organizations, already in existence for some years, have long experience in designing descriptive, public-use data tapes that are consistent with all the principles of privacy and confidentiality advanced in Chapter 4. Thus, to the extent that any affirmative action is required of HDOs, it should at a minimum include that they create data files according to well-known precepts and methods for public-use tapes and transfer data only through those means. Further steps are examined in Chapter 4. In sum, HDOs can probably not be expected to police the proper and responsible use of their data beyond their own walls. They can, however, cut off from further access to their data any users who have abused these principles, and the committee advises that HDOs be authorized or empowered to do so.

Using Valid Analytic Techniques

This chapter has dealt chiefly with descriptive or evaluative studies and related activities that HDOs might pursue and then release publicly; it has also considered the responsibilities that HDOs might have to make data available to others for private QA/QI programs and for other analytic purposes. Consistent with the discussion of data quality in Chapter 2, however, the committee wishes to emphasize that valid studies will require valid analytic techniques, often multivariable methods.

Such techniques are powerful, complex, and arguable in details among experts, but they are easily understood in principle. Properly used, they provide the best methods for comparing institutional, provider, plan, or clinician performance. This is so because only such approaches can isolate performance from the potentially confounding effects of differences in the prevalence of disease, severity of illness, presence of large numbers of elderly or poor patients, and so forth. Variables that reflect these factors are often termed severity-, risk-, or case-mix-adjustors.

Serious criticisms can be leveled at studies with inadequate or inappropriate adjustors and analyses;¹¹ such studies, and the organizations that conduct them, can readily be discredited. The consequences could be devastating to the entire public-disclosure effort. To help protect against such criticisms, the committee advises that HDOs use only proper statistical techniques—particularly multivariable analyses—for comparisons released to the public.

In addition, the committee holds that HDOs ought to use only severity- and risk-adjustment techniques and programs that are available for review and critique by qualified experts. It also believes that HDOs would do well to require, as a condition of data release, that entities or investigators conducting secondary analysis of these data adopt the same precept.

Privacy Protections for Person-identifiable Data

Chapter 4 deals with the issues of privacy and confidentiality of person-identifiable data in depth. In advocating release of data to providers, researchers, or others, however, the committee recognizes the possibility that information identifying patients or other persons may be inadvertently revealed to those with neither a right nor a need to know. This brief commentary on how privacy might be protected is intended to signal the committee's strong view that it must be protected.

Generally, committee members believe that encrypting, encoding, and aggregating patient data can go a long way toward protecting the identity of patients and other persons in HDO databases. Although no one can ever guarantee zero probability that an individual can be identified through concerted effort and ingenious devices, these methods make it possible to issue assurances about privacy at a reasonable level of confidence. Such assurances are stronger regarding tabular or aggregated information than data in discrete records, but they might still be reasonably strong at the level of tapes and other disaggregated databases.

Some committee members were not convinced that such steps will be sufficient to ensure anonymity and protect the identity of patients. They argued that a less ambitious goal—that of protecting the identity of patients to the extent feasible—is more realistic. It has the further advantage of conveying to the public that absolute protection of the identity of individuals when their information is in a computerized data bank is very difficult, if not impossible.

Consistent with the principles developed in Chapter 4 on access to person-identifiable information for researchers, the committee argues that appropriately qualified, institution-based researchers with approvals from their institutions' Institutional Review Boards can receive data with intact identifiers. The committee suspects that most investigators will not wish to acquire data tapes on analyses that HDOs have already performed. Rather, they will wish to obtain data that permit original analyses on a broad array of topics relating to the effectiveness and appropriateness of health care interventions. The committee does make provision in Chapter 4 for researchers interested in HDO data for these purposes to receive person-identifiable information.

Constrained Staff Capabilities

This chapter assumes that HDOs will have an affirmative responsibility to carry out analyses and public dissemination activities. Clearly, however, what studies they choose to do and how actively they pursue communication of results will be constrained by available resources and dictated by their own perceptions of critical issues. HDOs, for reasons of political disinclination, lack of staff, or other factors, may not wish or not be able to conduct certain analyses that outside requestors, such as consumer groups or newspapers, regard as extremely significant or timely. Such outside requestors might even be able to pay for such work, or at least to underwrite the costs of acquiring the data; HDOs may still find it impossible or unattractive to attempt to respond to such requests. The committee developed no consensus on how these problems might be addressed, beyond the points made earlier that HDO charters ought to mandate consumer-oriented studies.¹²

Obligations to Correct Analyses or Retract Information

A complex issue that proponents of HDOs need to consider is what incentives are required to ensure the accuracy of their analyses and subsequent publications. The market for the work may act as one corrective, as discussed earlier. In theory, if it becomes known that an HDO has published information demonstrated to be false, wrongly interpreted, or inappropriate, then political, economic, and social support for the HDO will falter. In practice, the committee was not convinced a market-oriented solution like this would have much effect on an HDO's later performance—there is essentially no evidence on the point—and it was even less persuaded that harms to individuals could be prevented or redressed in this manner.

An alternative, more activist approach may be to require that HDOs publish retractions in the same way and through the same media that published the original erroneous material. Some remedies for injury to individuals or providers might be available through civil litigation when false information has done material harm to their reputations, incomes, or careers, although bringing libel actions that demonstrate malicious intent or foreknowledge of the falsity of the information may be extremely difficult. A further drawback to this strategy is the same feeling that inhibits individuals whose privacy has been breached from bringing lawsuits—the disinclination to make public again what was painful, defamatory, or otherwise harmful when it first was publicized.

A related question is what to do when data are corrected long after they have been processed, used in analyses, or transferred to other users. For example, should HDOs notify users or recipients of their data when something has been augmented, corrected, or changed? Ought they go further and insist that the users alert the public or others to whom they have disclosed study results or transferred data of such matters? The committee did not develop considered opinions on these questions, but it did believe that individual HDOs should be prepared to devise policies to address these issues in anticipation of the day when the questions will arise in their own operations.