Online Issues

Journal Vol 49 (4)

The Institutional Official and Postapproval Monitoring: The View from 10,000 Feet

Robert P. Lowman

Robert P. Lowman, PhD, is Associate Vice Chancellor for Research at the University of North Carolina at Chapel Hill.

Address correspondence and reprint requests to Dr. Robert P. Lowman, Associate Vice Chancellor for Research, University of North Carolina at Chapel Hill, CB 4100, Bynum Hall, Chapel Hill, NC 27599-4100 or email lowman@unc.edu.

Abstract

The role of the institutional official (IO) in ensuring a high-quality program of laboratory animal care and use has received relatively little attention in the literature compared to that of the institutional animal care and use committee (IACUC) or attending veterinarian. Yet the IO is critical to institutional effectiveness. Often not a scientist, especially at colleges and universities, the IO is an executive who bears ultimate responsibility for the care and use of laboratory animals. An IO operates largely through delegation to the IACUC and the institutional veterinarian, yet must maintain a clear view of the program from the executive level—the "view from 10,000 feet." To do so, the IO must bring four critical leadership qualities to the position: (1) vision, (2) a commitment to quality and integrity, (3) strong planning and resource development, and (4) accountability to the laboratory animal program. The IO's ability to act on these qualities and to evaluate the effectiveness of the resulting program requires reliance on a strong program of postapproval monitoring (PAM). Summary data from PAM, based on metrics carefully defined by the IO in consultation with the IACUC and the institutional veterinarian, provide the feedback necessary to guide executive decisions and allow the IO to place the needs of the program of laboratory animal care and use successfully into the larger context of the mission of the institution.

Key Words: institutional official; leadership; management; postapproval monitoring

Background

Every institution that sponsors research with laboratory animals is required to have a designated institutional official (IO¹). At most institutions, the IO is an individual who holds an executive position, often a vice president or a vice chancellor, or who reports directly to an executive at that level. According to the PHS Policy on Humane Care and Use of Laboratory Animals (NIH 2002b), the IO is "An individual who signs, and has the authority to sign, the institution's Assurance, making a commitment on behalf of the institution that the requirements of this Policy will be met." In plain English, when it comes to the humane care and use of laboratory animals at any institution, the buck stops with the IO.

Not every IO may appreciate the authority and responsibility that go with the position, and even the Guide for the Care and Use of Laboratory Animals (NRC 1996) is of little help on this point. Its single reference to the IO specifies only that s/he must appoint an institutional animal care and use committee (IACUC¹). With limited information readily available, and because the typical IO is an executive with many other responsibilities, the IO may see his or her role as involving little more than signing assurance forms and annual reports and meeting periodically with those who seem to be in charge of the laboratory animal programs at their institution. The Office of Laboratory Animal Welfare (OLAW) reinforces this position in its own online tutorial (NIH 2000) by stressing that "The Institutional Official relies on the IACUC to oversee the program, to develop plans to correct program deficiencies, to address concerns that may arise regarding the institution's use of animals, and to make recommendations with regard to the program." One could conclude that the IO should play a rather passive role, signing off on the work of the IACUC, but otherwise having little to do with the institution's program of laboratory animal care and use.

Nothing could be further from the truth. The IO is the individual ultimately responsible for compliance with the PHS Policy. As the Institutional Animal Care and Use Committee Guidebook (NIH 2002a, 12) makes clear, the IACUC "makes recommendation [sic] to the IO regarding any aspect of the institution's animal program." When it comes to making programmatic decisions, especially those with implications for the policies of the institution, authority rests with the IO. His or her signature on the Assurance of Compliance is not a formality; it is a legal commitment by the institution to obey the rules, and in signing the document the IO takes responsibility for honoring that commitment.

Reality intrudes all too often at this point, especially at colleges and universities. The IO at an academic institution is typically a vice president, vice chancellor, or vice provost for research, or perhaps an assistant or associate vice president, vice chancellor, or vice provost. While most such individuals are hired into their administrative positions from the faculty, they may come from any discipline. During the nearly 27 years I have reported to a vice provost or vice chancellor for research at two different universities, I have served three chemists, one biochemist, one physicist, one psychologist, and one physiologist. Only two of those individuals had personal experience in the use of laboratory animals in research, but all of them were at least scientists who understood the value of laboratory animals as scientific tools. Many IOs come from disciplines far removed from laboratory research of any kind, including the humanities and fine arts. The result, as assessed by Greene and colleagues (2007), is that IOs "seldom have the requisite knowledge or experience to fulfill their charges effectively."

Symptoms of Ineffective Leadership

A lack of knowledge or experience in operating a program for the humane care and use of laboratory animals often results in ineffective leadership. Two major symptoms are predictable and easy to describe. The first is an overreliance on subordinates, such as the director of laboratory animal care, campus veterinarian, or IACUC staff director. The IO may reason that these subordinates are professionals who know what they are doing, whereas the IO's background by comparison includes little knowledge about laboratory animals. If the IO is unfamiliar with the laws and regulations governing the use of laboratory animals, s/he may be uncomfortable second-guessing or questioning the expertise of those who are technically trained as scientists or veterinarians, and may believe that those with the expertise are likely to make the best decisions. The result is an uncritical acceptance of whatever professionals in the field say or do.

The second symptom involves the IO playing a passive rather than an active role. In this scenario, the IO meets with relevant program officials infrequently. Recommendations from the director of laboratory animal care or the IACUC may languish for weeks or months on the desk of the IO, who finally makes decisions only when failure to do so would result in a crisis for the program. When the IO does make decisions, they may appear to be arbitrary or inconsistent with existing policies or procedures, made without reference to or study of the supporting materials prepared by subordinates for the IO's consideration.

Of course, the IO has every right to expect and demand expertise from the director of laboratory animal care, the campus veterinarian, and the director of the IACUC. All should be highly skilled professionals with relevant advanced degrees and the knowledge and experience to maintain and operate a high-quality program of laboratory animal care and use. One might argue that accepting their decisions uncritically is a reasonable management technique for a supervisor who knows little or nothing about laboratory animal care and use. But overreliance on the experts and passive management are both dangerous ways to operate and are likely to lead to serious difficulties for an institution over the long run.

IO Leadership in the Context of the Institution

Programs of laboratory animal care and use operate in the context of the institution that sponsors them. They depend on adequate resources and facilities and the high-level support of the administration when it comes to enforcing exemplary standards of care and use. As the Institutional Administrator's Manual for Laboratory Animal Care and Use (NIH 1988) points out, IOs can "influence institutional priorities" and "assure that sufficient monetary and personnel resources are allocated." Research institutions are inherently political systems. It takes more than technical or scientific expertise to maintain and operate a high-quality program of laboratory animal care and use, and it is up to the IO both to provide leadership and direction and to insist on accountability where laboratory animals are concerned. Compliance starts at the top, and thrives only in an institutional environment where doing things the proper way is valued.

Without leadership, direction, and accountability, a program of laboratory animal care and use will drift. Employee morale will suffer, because employees are quick to understand when their programs are not institutional priorities. Standards may slip, because the institution sends the message that there are no consequences for noncompliance. Programs can fall behind in the quality of facilities and equipment, because low-priority programs do not attract new resources. Programs can fail to provide high-quality customer service, because animal care professionals may be more concerned with making things appear to run smoothly than with meeting the needs of scientists and research staff. Research can come to be seen as a disruption of the daily routine of animal care and veterinary staff rather than the reason for the program's existence in the first place. Alternatively, an animal care program may be inappropriately influenced by a research investigator who resists limitations and operates without appropriate standards.

In all these situations, the quality of laboratory animal care and use can fall into a downward spiral based on neglect from the one person at the institution most responsible for compliance—the IO. The IO's active management of the program is not optional. An IO must accept the challenge of leadership regardless of her or his disciplinary background or prior knowledge or experience with laboratory animal care and use. But the IO faces a common executive dilemma: how does an executive manage a program about which the executive is not a substantive expert?

The Challenge of Leadership

An IO faces an executive challenge that is not all that different from that faced by any business executive. A chief executive officer (CEO) who came up through the ranks of marketing may know little or nothing about manufacturing or finance. An expert in finance may know little or nothing about new product development or engineering. An engineer may know little or nothing about human resource development or sales. Companies are successful not because every new CEO masters the intricacies of every aspect of the business: they are successful because the CEO is an effective leader.

Similarly, it is the role of the IO to provide effective leadership to the institutional program of laboratory animal care and use. It is not the IO's job to provide the expertise of a veterinarian or a research scientist. The view of the IO is from 10,000 feet: a successful IO leaves the details to others and devotes energy to those aspects of a successful program that come uniquely from effective leadership.

In this article I stress four aspects of effective leadership that mark a successful IO: (1) vision, (2) commitment to quality and integrity, (3) planning and resource development, and (4) accountability.

Vision

The American Heritage Dictionary of the English Language (2000) defines vision as "unusual competence in discernment or perception; intelligent foresight." Applied to the oversight of a laboratory animal care and use program, vision calls for a clear understanding of the role and purpose of the program and how it fits into the overall mission of the organization. For example, if the purpose of the program is to provide and maintain a valuable scientific resource for the use of faculty or staff, vision might include knowing the kinds of research most likely to be undertaken, the numbers and species of animals most likely to be needed, and the expectations of faculty or staff members for the performance of the program. A simple strategy for gaining that kind of information might include regular visits with research leaders to discuss their future needs for laboratory animals and to solicit their feedback about how well the institution's program is meeting their needs. It might also include the regular examination of a set of metrics designed to keep the IO informed about trends in the program, such as changes in the animal census from month to month, a list of the most commonly used species and strains of animals, and an updated list of the largest animal users or breeders among the faculty or staff. It would also include regular meetings with the chair of the IACUC and periodic meetings with the entire IACUC to enable an assessment of changes in research direction at the institution and the likely impact of such changes on management of the laboratory animal care program.

At my own institution, the IO and I meet every other month with the IACUC chair and staff director and the director of laboratory animal medicine to discuss priorities, future needs, and ways of communicating and collaborating more effectively. We also meet every other month with the associate deans for research from across campus, and the needs of the laboratory animal program are often on the agenda. An IO does not have to be a scientist to understand and use this kind of information to advantage in anticipating the needs of the program. "Intelligent foresight" is a lot easier when you have the data, and almost impossible when you do not.

Commitment to Quality and Integrity

As mentioned above, quality and integrity start at the top of any organization. Hitt (1990, 1) observed, "Effective leadership is a consequence of ethical conduct, and ethical conduct is a consequence of effective leadership. Ethics and leadership function as both cause and effect." When executives stress a personal commitment to the highest standards of performance, that commitment will be reflected throughout the organization.

But talk is cheap. Commitment to integrity must be backed up with action. Ignoring violations, discounting whistleblowers, and failing to take effective action when problems are identified negate any talk about quality and integrity. An IO can claim a commitment to quality and integrity, but the message must be backed up with consistent action. A few examples that signal the importance of doing things the right way include (1) establishment of an anonymous hotline—with effective follow-up—that encourages reports of potential compliance problems to someone of authority outside the laboratory animal care and use program, (2) an active program of reporting all violations to regulatory and accrediting oversight authorities, and (3) insistence on the equitable application of rules to all animal users, regardless of their seniority, prominence, or influence in the organization.

Several years ago, when my institution came under attack from an animal rights organization, we held a press conference the same day our attackers went public with their concerns. In our press conference, the university's chief academic officer, the vice chancellor for research and economic development, and the dean of the school of medicine were available to reporters and concluded the press conference by leading members of the press on a tour of one of our laboratory animal facilities. The message of the press conference was that the institution was committed to the highest standards, would investigate each and every allegation, would make whatever corrections were necessary, and had nothing to hide. The message was also intended for our own faculty, staff, and students. The university was making a public statement at the highest level that quality and integrity were of paramount importance.

Internal policies and practices must be consistent with public statements. The lowest-paid animal husbandry technician on the staff has to be convinced that speaking up for quality or reporting a potential violation is a smart career move. Jennings (2006, 10) has observed that "Most companies and their ethics programs fall short of taking the steps necessary to create a culture in which employees come forward with concerns and feel comfortable making ethical choices and in which ethics is paramount in decision making." While top executives may forget this fact, rank-and-file employees feel powerless on a daily basis and usually believe that no one will take their word against that of a nationally famous research scientist who brings millions of dollars of research money into the institution every year. A culture of compliance exists when employees at all levels believe that the institution is committed to high standards of care and that their supervisors will back them up when they insist on doing things the right way and reporting those who do not.

Planning and Resource Development

Resource allocation is almost always a zero-sum game at any institution, but an IO's active leadership of the laboratory animal care program produces both an understanding of the needs of the program and an appropriate investment of resources. Robert (1988, 4) has written that "Most managers are so engrossed in operational activity that they have not developed the skill of thinking strategically." Strategic thinking results from a clear understanding of the mission and purpose of the laboratory animal care and use program and an analysis of what it will take to accomplish them.

Putting strategic thinking to work in the organization requires the IO to recognize situations that call for his or her involvement. For example, unless the IO becomes an active participant in the planning process, research directors or deans will hire new faculty or staff members without regard to whether there are adequate laboratory animal facilities at their institution. New laboratory buildings will be constructed without including adequate space for laboratory animals unless the IO insists on having those needs considered. Institutions may subsidize their per diem rates for the care of laboratory animals out of institutional funds or research overhead, but without the active involvement of the IO at the highest executive levels of the institution, institutional support can erode in favor of other projects or other needs, based on whose advocates are speaking the loudest.

To respond to these issues, my university formally adopted a policy that requires the approval of the IO for every construction or renovation project that may affect the demand for laboratory animal space, including projects that would build additional laboratory space for faculty members who are likely to need animals in their research. Not every project has to include animal space, but every project has to show that sufficient animal space has been identified somewhere on campus to meet anticipated needs for new animal facilities created by the new or renovated research space.

Ironically, an IO who is not an animal user may actually be more effective in obtaining resources for the laboratory animal program. A robust finding in the psychological literature describes the role of what is called knowledge bias, or "the presumption that a communicator has a biased view of the issue" (Perloff 2008, 227). Speakers are perceived as being more credible, and are hence more effective at persuading others, when they violate knowledge bias, or argue against their own interests (Eagly et al. 1978). Thus while a research vice president might logically be expected to argue in favor of increased resources for a program that provides an important research tool, an IO who is not an animal user is likely to be perceived as more credible, because it will be clear that her or his own department or own research or scholarship will not benefit from the allocation of resources.

Accountability

Vince Lombardi, the famous coach of the Green Bay Packers, is reported to have said, "If you aren't keeping score, it's only practice." His words speak directly to the issue of accountability. An IO does not need to be a scientist or an animal user to be able to help her or his direct reports set challenging but attainable goals for themselves and the programs they direct and to hold them accountable for their performances. But accountability is realistic only if the goals can be measured accurately. Thus goal setting must include agreements about the metrics of performance that will be used to measure success and will help everyone understand whether the program is meeting the standards set for it.

Accountability must also go far beyond setting goals and measuring their success at the end of the year. A good IO must be actively engaged with the directors of the laboratory animal care program to help plot midcourse corrections as events unfold, provide guidance and consultation at major decision points, and provide feedback gleaned from meetings with users and other leaders of the institution. A good IO should assume the roles of a coach, a sounding board, and a source of information about the policies and actions of the top leadership of the institution.

It is critically important for the IO to provide feedback at regular intervals to the professionals who are managing the program of laboratory animal care on a daily basis and to be seen as an ally and protector. It is at this level that the political skills of the IO will be tested. The institution has unlimited aspirations and limited resources. Whether adequate resources will be allocated to laboratory animal care depends on how strong a case the IO can make for the importance of the program. That, in turn, depends on the high-quality performance of the program directors, accurate and impressive metrics of performance, insightful knowledge of the needs of animal users, and the IO's leadership skills in presenting those data to higher institutional authorities.

Leadership and Postapproval Monitoring

Leadership as described above sets the stage for an effective program of postapproval monitoring (PAM¹) by creating and promoting a culture of compliance at the institution. Each of the four components of leadership contributes directly.

Vision encourages investigators to undertake worthwhile scientific projects that make wise use of limited resources. It frees scientists to think creatively and take calculated scientific risks that have the greatest potential to advance knowledge. Vision also creates an understanding of how the program of laboratory animal care and use fits into the mission and priorities of the institution, and helps everyone involved in the program realize that they are part of a larger whole. This, in turn, allows employees to take pride in their work and increases morale.

Commitment to quality and integrity leads to a demand for training and upgrading of skills. When employees know that an institution is serious about creating and maintaining quality programs, most respond by voluntarily looking for ways to do their jobs better. Just as, for example, there is pride in being associated with a winning athletic team, being associated with a research program recognized as the best is a source of pride for everyone concerned. Employees want to work for employers with a reputation for quality, and given the opportunity will do whatever they can to protect and enhance that reputation. A commitment to quality also motivates employees to refuse to accept situations or activities that can damage the reputation of their programs. It encourages everyone to meet high standards and refuse to accept unacceptable work.

Appropriate planning and resource development ensures that everyone associated with a program has the tools necessary to do things the right way. Inadequate resources quickly lead to lapses in quality as people are forced to compromise to make up for the shortcomings. Shortages force people to cut corners and lower standards, and thus militate against a strong culture of compliance.

Accountability provides the feedback necessary to improve both individual and program performance. It allows everyone associated with the program to see how they measure up to expectations, both their own and those of the institution. In programs where accountability has been lacking, employees may initially resist the measurement of performance, because they assume they will fall short and do not know what the consequences of failing might be. This is especially true in situations where standards of performance have not been clearly articulated. Once realistic performance standards are made clear and employees are convinced that measurements are accurate and reliable, they are likely to accept accountability systems with little difficulty. Over time, as their own performance improves with training, proper supervision, and experience, they may even welcome measurements of performance as continued proof of excellence. In a mature system, accountability becomes the yardstick by which quality is measured.

Postapproval Monitoring as Accountability

Monitoring the performance of staff members as they conduct IACUC-approved research is nothing more than a means of holding everyone accountable for their care and use of laboratory animals. Yet many institutions do not have effective PAM systems. They put their energies into the review and approval of protocols, without checking afterward to see if the protocol has been followed. Unless the IO takes proactive steps to institute PAM, s/he will be left in the dark when it comes to accountability. It would be somewhat analogous to betting on the outcome of a game that no one watches and for which no one keeps score.

It is essential for the IO to know that everyone on the husbandry, veterinary, and scientific staff has received appropriate training in the care and use of laboratory animals. It is also important to know that every scientific protocol has been carefully reviewed by a thoughtful committee of experts, that appropriate policies and standard operating procedures are in place, and that facilities are well maintained. But even the knowledge of all these things does not allow the IO to state with confidence that the program of laboratory animal care and use is fully in compliance with all regulatory standards. In the absence of a systematic PAM program, the IO cannot accurately evaluate the performance of direct reports or know with any certainty that the IACUC is maintaining high ethical and scientific standards in the research it approves.

A thoughtful PAM program is not designed to catch and sanction people for violating the rules. The overwhelming majority of research personnel do not tolerate inappropriate behavior in the laboratory and are committed to the ethical and humane care and use of laboratory animals because they know that healthy animals, humanely treated, result in better science. PAM is accountability, pure and simple. In the best case, it allows the IO and those who report to the IO to "catch people doing something right." In other words, it provides the qualitative and quantitative evidence that everything is happening as it should.

At the same time, PAM provides an opportunity for coaching and consulting to promote best practices. Given the very nature of research, not every protocol works out as intended. Not every procedure goes smoothly. Not every technique results in the quality of data that the investigator hoped to collect. With regular PAM observations, institutional personnel have the chance to speak with research personnel about problems as they occur and to make suggestions of alternative procedures that are known to have worked in other laboratories. Where research personnel are having difficulty with particular procedures, PAM can support a prompt assessment of the situation and identification of the need for either additional training or modifications in procedures.

Again, the IO works at 10,000 feet, empowering others to design good PAM systems and motivating others to implement PAM successfully. The IO benefits from PAM by receiving regular performance reports in the form of metrics and qualitative observations that allow the IO to discuss with direct reports any changes in policy or procedure that might be necessary to ensure the highest quality of animal care and use.

Problems Occur Routinely, but No Problem Is Routine

Based on my many years of personal observation—more than 15 of them as an IO—problems in laboratory care and use fall into two broad categories. The first category covers failures of husbandry or veterinary care. An automated watering system malfunctions. A cage door is improperly secured. The heating system goes down. A technician fails to administer medication ordered by the veterinarian. These kinds of problems are those of operational quality control. They are not specific to one protocol. They often result from unobserved equipment malfunction or errors on the part of institutional husbandry or veterinary staff. They regularly occur at a low level with random frequency in spite of the best efforts of the IO and managerial personnel to create and sustain a high-quality program of laboratory animal care and use. Problems of this nature vary in severity from "no harm, no foul" to situations that have a major impact on the animals' health or welfare. PAM is generally not going to identify these problems because they have to do with the daily routine of operating a laboratory animal facility and are unrelated to the approval of a protocol.

The second kind of problem is research-driven, and it is in this area that PAM proves its worth. A group of animals dies after the administration of an experimental drug although no mortality had been predicted. Postsurgical infections point to a failure of aseptic technique. Litters of knockout mice are developmentally delayed and fail to thrive when weaned at the usual age. Live animals are found in the dead animal cooler. A cage of animals is found left in a procedure room with no food or water. An investigator with a mouse breeding colony allows cages to become overcrowded. An investigator changes the dose of an experimental medication without IACUC approval because the approved dose is producing severe side effects. These are usually protocol-specific rather than issues of operational quality control. They are almost always errors or lapses of judgment by research personnel. In a well-managed program with strong PAM, these problems will be observed by compliance personnel or self-reported by investigators who realize they have a problem.

Both of these kinds of problems occur regularly, even in the best programs. IOs who believe that the quality of program at their institutions is so high that such problems do not occur are out of touch with reality. Consider the following armchair analysis of the frequency with which problems occur. Assume there are 1800 faculty members, staff, and students at a major research university who are actively involved in research with laboratory animals. Further assume that these personnel are well trained, committed to compliance, and very careful in their work. Even so, they are human, and humans make mistakes. If each person makes only one mistake every 5 years that results in a violation of policies or procedures that might affect animal health or welfare, there will be an average of about 360 such events every year, or nearly one a day, 7 days a week—and that does not count facility or equipment failures unrelated to human error.

Most problems (or potential problems) will never come to the attention of the IO, and rightfully so. A technician discovers a malfunction in an automatic watering system and makes a correction before any animals are harmed. A postdoc notices that a pharmaceutical is outdated and discards the product without using it. A research technician observes that weaned knockout mice seem less developed than expected and provides them with supportive care until they are able to thrive on their own. A technician having difficulty performing a gavage procedure seeks the assistance of a more experienced colleague. Each of these events could cause serious harm to one or more laboratory animals, but the training and experience of personnel allow them to recognize the problem and deal with it appropriately before any harm occurs. So it is with a good program of PAM.

The best leaders instill in others the desire to solve problems, learn why those problems occurred, and implement policies or practices to keep them from happening again. The best IOs are good leaders who encourage their direct reports to put in place effective PAM systems that empower personnel across the board to recognize and solve problems before they affect the animals' health or welfare. The ideal is for compliance officials who administer PAM programs to be seen as resources, consultants, teachers, and trusted colleagues who can be counted on to help solve problems and protect animals.

Of course, an investigator will occasionally break the rules, whether by accident or design, so there are instances when a compliance official, in the course of normal PAM activities, has to report a violation for investigation and possible sanctions. It is not the role of the IO to investigate allegations of wrongdoing or second-guess the work of the IACUC and compliance personnel as the investigation continues. Still, at my own institution, the IO receives every preliminary report of possible violations as one of the metrics of accountability. This also allows the IO to inform a dean or department head, if necessary, that an investigation is ongoing, and to keep upper-level management informed of any allegations that might come to the attention of the press. If the IACUC determines that a serious violation occurred, the IO is responsible for notifying the appropriate regulatory agencies and extramural sponsors. The IO will also make the final decision about any sanctions to be imposed on the individual found guilty of the violation, normally on the recommendation of the IACUC.

No IO wants to learn of large numbers of violations that s/he must report to regulatory agencies and sponsors. And sponsors probably dislike receiving such notices even more than the IO dislikes sending them. Yet, although it may seem counterintuitive, the number of problems reported to supervisors, principal investigators, or the IACUC is probably positively correlated with a more active culture of compliance. When personnel do not report problems, it does not mean problems are not occurring (as illustrated in the armchair analysis above); what it may mean is that people see no benefit in reporting them. That suggests a lack of faith in the system, or a belief that the system places no value on the judgment and credibility of staff members. According to ReVelle (2002, 35), "We cannot expect the best effort from people who don't feel trusted and respected." As previously discussed, rank-and-file employees will not generally report problems unless they believe their reports will be perceived as credible and taken seriously. Employees at all levels will not report problems unless they believe the institution may take action to solve them. And whistleblowers will not speak up unless they believe they will be protected from retribution for having reported a problem in the first place.

Particularly noteworthy may be self-reports of possible violations. The circumstances in which investigators or employees are most likely to make self-reports include the following: (1) they believe they have an obligation to report problems, (2) they perceive the institution as fair, (3) they see the institution as being more interested in correcting problems and keeping them from happening again than in punishing those who violate the rules, or (4) they believe the problem will probably be discovered and reported by someone else anyway. An institution with an effective PAM program under the effective leadership of the IO should meet all four conditions.

PAM Decreases Regulatory Burden in the Long Run

Strong PAM programs help IOs measure success, both their own and that of their laboratory animal programs. In the long run, PAM ought to decrease regulatory burden, because it becomes part of a culture of compliance that maintains program continuity and improves performance. There are several ways that PAM makes the job of the IO easier and contributes to overall program effectiveness.

First, PAM allows regular analyses of commonly observed problems and mistakes. Detection by PAM of recurrences of the same problem over a short time frame suggests a problem in policy, procedures, or training best addressed at the system level. An IO who sees such a pattern should follow up with the IACUC to begin the process of modifying policies or practices accordingly. Lack of discernible patterns in violations and other problems identified suggests more routine problems of quality control—still serious, but best dealt with through more consistent management practices while continuing to stress the importance of accuracy and attention to detail.

As manufacturing facilities learned long ago, quality control is best when processes are designed to either decrease the probability or increase the detection rate of errors. For example, an isolation facility for infectious disease research with a two-door entry vestibule represents a potential barrier protecting those outside the facility only if the latches on the doors at both ends of the vestibule are set so that one door can be open at a time. IOs should challenge their direct reports to find ways of reengineering existing processes to make errors more difficult or detection more sure, and should be ready to identify resources that can assist in making needed changes when creative ideas are presented.

Second, PAM provides feedback about the success or failure of changes in policy or practice and becomes one of the metrics that allows an executive to have confidence in the quality of the system. Several years ago at my own university, we identified a problem of overcrowding in mouse cages due to a breeding policy that was not uniformly enforced. When the university attempted to address the problem, investigators complained that the policy was overly restrictive and that some cages categorized as overcrowded under the policy represented no threat to the health and welfare of the animals. The IACUC conducted a careful evaluation of the situation and developed a new policy. While less restrictive in some respects, it was also more complicated to administer, and the IACUC was concerned that personnel would not be able to implement it without making a lot of mistakes. After extensive training and massive publicity, the new policy was put into effect. Over a period of 6 months after implementation, as part of the university's ongoing program of PAM, every breeding cage was checked repeatedly for overcrowding under the new policy—more than 100,000 observations. The monitoring showed that fewer than 1 percent of all cages were in violation of the new policy. Thus in spite of a new policy that was more complex than the one it replaced, PAM demonstrated that the new policy was a success, and the university has not experienced any serious problems of overcrowded cages since.

Third, PAM provides a high level of system continuity. While PAM is clearly time-consuming and represents a significant expense, especially in a large program, it is much less burdensome than recovering from a system failure or the consequences of a serious noncompliance that brings regulatory agencies to your doorstep. We can learn something in this regard from the nuclear power industry. Safety is paramount in nuclear power plants, so every variable conceivably related to normal operation or safety is monitored. But system failures are extremely uncommon, so the question is how to keep employees vigilant when the probability of detecting anything out of the ordinary is extremely low. The answer is not to rely on human operators to notice when something is not right but on bells and whistles and bright lights: if any critical variable deviates beyond its normal tolerances, alarms sound and lights flash. For laboratory animal care and use programs, PAM represents the alarms and flashing lights. The ideal is for PAM to produce nothing worth noticing, month after month. But rather than count on husbandry, veterinary, or research staff to notice when something goes wrong, we learn to trust the metrics of a high-quality system of PAM.

Finally, in this day and age when some members of the public do not recognize the value of research with laboratory animals, strong PAM programs allow IOs to speak with confidence about the quality of their programs and the value of the research conducted at their institutions. Ideally, members of the public should trust our institutions "to do what is right, just, and fair even when they are not compelled to do so by the legal framework" (Crane and Matten 2007, 50). At my university, we stress to investigators that using laboratory animals in research is a privilege, not a right. To maintain that privilege, investigators must be committed to excellence and must demonstrate that commitment on a daily basis in the conduct of their research. PAM gives us the assurance that our investigators, who are the teachers and role models for the next generation of scientists, are living up to those high standards. With that knowledge, we can speak to the public with confidence about the research at our institution. It is an obligation of scientific leaders to share the knowledge they have gained from their research, and our ability to speak with confidence contributes to public understanding and acceptance of scientific methods—methods that lead to advances in knowledge and improvements in our health and welfare. That, in turn, completes the circle where this article began the discussion of leadership—the vision to understand the role and purpose of the program of laboratory animal care and use in the mission of your organization.

Conclusions and Future Directions

By the very definition of the term, an IO is an institutional executive who bears overall responsibility for the humane care and use of laboratory animals at the institution. The IO works by delegating authority to the IACUC and the institutional veterinarian, but that delegation does not absolve the IO from ultimate responsibility for the success or failure of the program. A high-performing IO—whether or not s/he has an educational background in a discipline related to the care and use of laboratory animals—uses the tools of any good executive to ensure high quality. Those include (1) vision, (2) commitment to quality and integrity, (3) planning and resource development, and (4) accountability. In this article I have attempted to show that these four essential functions of the IO cannot be delegated or ignored.

PAM is most closely related to accountability and is an essential tool in any effort to measure performance against standards. An actively engaged IO will want to see summary PAM data from the IACUC on a regular basis, perhaps monthly or quarterly, as a means of assessing whether the laboratory animal care and use program is on course, is dealing with problems quickly and effectively, and has policies in place that define and regulate the program appropriately. The IO complements the IACUC's efforts by looking at the big picture to reconcile the content of the PAM reports with the IACUC minutes, working with the IACUC when questions arise. For the IO, a PAM program is also a useful tool in evaluating the performance of subordinates to whom authority for operating the program has been delegated.

PAM is also relevant to the other three executive tools listed above. When the IO expresses interest in PAM and the data it produces, s/he shows a commitment to quality and integrity. This, in turn, sends the message to everyone involved in the program that these attributes are important, and that message leads directly to higher-quality performance. As previously noted, concern about quality and integrity can act as both a cause and an effect of good leadership. Similarly, PAM program data allow the IO to determine whether resource allocations are adequate to sustain the laboratory animal care and use program and whether the institutional vision for the program is being met. Like an effective CEO who keeps abreast of corporate performance, the IO doesn't need to drill down into the details but does need PAM data and reports to independently assess the functioning of the lab animal program and the IACUC. Summary reports of PAM programs are therefore essential to any IO who is committed to the highest standards of performance.

As research universities and for-profit and nonprofit research organizations grow in size, programs of laboratory animal care and use will grow as well. While in this article I take the position that an IO—even one with training in a discipline far removed from laboratory animal care and use—can be effective as an executive through the use of four essential leadership tools, it is reasonable to ask whether this will always be true. The bar continues to be raised when defining what it means to operate a program of laboratory animal care and use successfully. Laboratory animal veterinary medicine takes on added complexity as genetic research produces laboratory animals that cannot survive without specialized care. Members of the public correctly ask whether our treatment of animals is humane and can be justified by the importance of the research we conduct.

It is not difficult to envision a time, perhaps not all that distant in the future, when some might argue that an IO cannot do her or his job adequately without scientific training in a discipline that uses laboratory animals in research. When that day comes, I will argue in favor of the present system. The IO, as an executive, must place the institution's program of laboratory animal care and use in the larger perspective and context of the institution. Someone selected to be IO specifically because of scientific training may be more likely to argue from within the narrow interests of the discipline. An IO hired independently of a background in laboratory animal research is more likely to perceive the job as one of leadership for the benefit of the institution as a whole, even if the IO happens to have a relevant scientific background. I am not arguing that laboratory animal research is too important to be left to scientists. Instead, I am arguing that laboratory animal research is so important that it needs to receive executive attention at the highest levels of the institution from an IO with vision and integrity, regardless of whether that executive happens to be trained in a discipline that employs laboratory animals in research.

Abbreviations used in this article: IACUC, institutional animal care and use committee; IO, institutional official; PAM, postapproval monitoring

References

American Heritage Dictionary of the English Language, 4^th ed. 2000. Boston: Houghton Mifflin.

Crane A, Matten D. 2007. Business Ethics: Managing Corporate Citizenship and Sustainability in the Age of Globalization, 2^nd ed. Oxford: Oxford University Press.

Eagly AH, Wood W, Chaiken S. 1978. Causal inferences about communicators and their effect on opinion change. J Pers Soc Psychol 36: 424-435.

Greene M, Pitts ME, James ML. 2007. Training strategies for IACUC members and the institutional official. ILAR J 48:131-142.

Hitt WD. 1990. Ethics and Leadership: Putting Theory into Practice. Columbus: Battelle Press.

Jennings MM. 2006. The Seven Signs of Ethical Collapse: How to Spot Moral Meltdowns in Companies . . . Before It's Too Late. New York: St. Martin's Press.

NIH [National Institutes of Health, Editorial Committee of Institutional Administrators and Laboratory Animal Specialists for the Henry M. Jackson Foundation Advancement of Military Medicine, Uniformed Services University of the Health Sciences]. 1988. Institutional Administrator's Manual for Laboratory Animal Care and Use. NIH Publication No. 88-2959. Bethesda: Office for Protection from Research Risks, National Institutes of Health.

NIH [National Institutes of Health, Office of Laboratory Animal Welfare, Department of Health and Human Services]. 2000. PHS Policy on Humane Care and Use of Laboratory Animals Tutorial. Available online (http://grants.nih.gov/grants/olaw/tutorial/index.htm), accessed May 8, 2008.

NIH [National Institutes of Health, Office of Laboratory Animal Welfare, Department of Health and Human Services]. 2002a. Institutional Animal Care and Use Committee Guidebook. Bethesda: NIH Office of Laboratory Animal Welfare.

NIH [National Institutes of Health]. 2002b. Public Health Service Policy on Humane Care and Use of Laboratory Animals. Bethesda: NIH Office of Laboratory Animal Welfare.

NRC [National Research Council]. 1996. Guide for the Care and Use of Laboratory Animals. Washington: National Academy Press.

Perloff RM. 2008. The Dynamics of Persuasion: Communication and Attitudes in the 21^st Century, 3^rd ed. New York: Lawrence Erlbaum Associates.

ReVelle JB, ed. 2002. Manufacturing Handbook of Best Practices: An Innovation, Productivity, and Quality Focus. Boca Raton FL: St. Lurie Press.

Robert M. 1988. The Strategist CEO. New York: Quantum Books.