Online Issues

Journal Vol 48 (1)

Verification of IACUC Approval and the Just-in-time PHS Grant Process*

Michael D. Mann and Ernest D. Prentice

*Reprinted from ILAR e-Journal.

Michael D. Mann, Ph.D. is a Professor of Cellular and Integrative Physiology and the Executive Chair of the University of Nebraska Medical Center IACUC. Ernest D. Prentice, Ph.D. is a Professor of Genetics, Cell Biology, and Anatomy and the Associate Vice Chancellor for Academic Affairs. Both authors are at the University of Nebraska Medical Center, Omaha, Nebraska.

Address correspondence to M.D. Mann, Cellular and Integrative Physiology, UNMC, Omaha, NE 68198-5850 (Email: mmann@unmc.edu).

Abstract

Public Health Service Policy requires institutional animal care and use committees (IACUCs) to review and approve the animal care and use aspects of grant proposals submitted for funding. Initially, submission of verification of approval was required with submission of the proposal; however, in 2002, a “just-in-time” (JIT) process was begun to allow submission when it appeared that the proposal would be in the fundable range. After 3 yr, no evaluation of JIT and its impact on the IACUC has been published. An online survey designed by the authors to determine this impact is described. Invitations to participate in the survey were sent to 1002 IACUC chairs and administrators, and responses were received from 168. Surprisingly, 30% of respondents indicated that they did not review grant proposals at all. The majority of those respondents reported that they rely on principal investigator assurance that approved IACUC protocols match grant proposals. In the authors' experience, this procedure is not safe 23% of the time, and several respondents described the same experience. Most respondents indicated that they use JIT whenever possible for all investigators. Of those who reviewed grant proposals, most indicated that each review consumes 20 to 29 min, and most respondents indicated that the granting agency does not allow enough time for a JIT verification to complete a de novo submission and review of a protocol. Instead, the protocol must have been submitted and reviewed prior to the JIT request. Respondents indicated that there is considerable uncertainty about what the granting agency would do when an IACUC requires changes in a protocol during the JIT process.

Key Words: grant proposals; IACUC; just in time; PHS Policy; protocols

Introduction

The 1986 Public Health Service (PHS¹) Policy on Humane Care and Use of Laboratory Animals (PHS 1986) requires that “Applications or proposals (competing and noncompeting) covered by this Policy from institutions which have an approved Assurance on file with OLAW shall include verification of approval (including the date of the most recent approval) by the IACUC of those components related to the care and use of animals. . . . If verification of IACUC approval is submitted subsequent to the submission of the application or proposal, the verification shall state the modifications, if any required by the IACUC.” For competing applications or proposals, verification of institutional animal care and use committee (IACUC¹) approval may be filed at a time not to exceed 60 days after the proposal application deadline (PHS 1986). This original policy did not give institutions much flexibility in requiring IACUC review and verification as to whether a project was likely to be funded.

To increase flexibility and reduce the burden on applicants and IACUCs, the 1986 policy was amended in 2002 to allow verification of IACUC approval to be submitted upon request from the PHS—so-called just-in-time (JIT¹) verification. The PHS issues the request when it determines that a grant proposal is likely to be funded. Institutions are not required to use JIT; they may still choose to follow the original requirements (NIH 2002).

The 2002 amendment included the following principles: (1) “Under no circumstances may an IACUC be pressured to approve a protocol or be overruled on its decision to withhold approval”; (2) “. . . in no way is peer review intended to supersede or serve as a replacement for IACUC approval. . . . The existing PHS Policy requirement that modifications required by the IACUC be submitted to the NIH with the verification of IACUC approval remains in effect”; and “The NIH understands its responsibility to ensure that institutions are given adequate notice to allow for timely IACUC review prior to award” (NIH 2002).

The JIT procedure has now been in place for nearly 4 yr, time enough for institutions, IACUCs, investigators, and the PHS to have thoroughly used the procedure and formed opinions. Actually, JIT has been in place for institutional review board reviews since 2000; however, there still exists only one evaluative article (Kelly and Johnson 2005), and it is addressed specifically to the Veterans Administration JIT process. We have not found any published evaluations of the JIT process for IACUC approvals of PHS applications.

In an effort to assist in this evaluation, we constructed the survey that is presented as the Appendix at the end of this article. The document was designed to collect information about how IACUCs approach verifying approval of the animal use in PHS applications and how the IACUC has been affected by the JIT process. No questions were included about work load reduction for principal investigators (PIs¹) or for the IACUC because this topic was not the primary interest of the study. Although the original 60-day grace period and the new JIT verification may reduce work loads of both the PI and the IACUC, it may create a situation in which the IACUC requires changes to a protocol that would result in a change in the grant proposal after the proposal has been peer reviewed. As we pointed out recently (Mann and Prentice 2004), it is not clear what the granting agency will do in this event. There may also be pressures exerted on the IACUC to approve the animal aspects of grant proposals that IACUC members might not otherwise approve. We endeavored to obtain this type of information in the survey.

Methods

Participants

We invited IACUC chairs and administrators from a contact list supplied by the Office of Laboratory Animal Welfare (OLAW¹) to complete an online survey consisting of 24 questions. The invitation list consisted of 1002 individuals representing institutions with approved, current Assurances. Initial invitations were sent in January 2005, but by July only 95 responses had been received. A follow-up letter was sent in August.

The study design and the primary and follow-up invitation letters were approved by the University of Nebraska Medical Center Institution Review Board. The approved design did not preserve identification of the responders unless they chose to give an email address for future contacts (see below).

Survey

The online survey consisted of two parts. The first part requested information about how the institution provided verification to the PHS of IACUC approval of the components of grant proposals related to animal care and use. It also included questions about the use of JIT by the institution, its experience with the process, and any difficulties it may have encountered with the process. The second part of the survey requested information about the institution's IACUC, its administrator, and its chair. Responses to the survey were numerically coded and stored in an Access database. No identifiers of the respondent were requested or saved unless the respondent chose to participate in follow-up questions should they arise. In the latter case, the respondent's email address was saved with the record.

Most respondents completed the online survey, but a few elected to print and mail a hard copy. Data from those hard copy responses were entered manually into the online survey vehicle so that all data could be analyzed together. The printed and online surveys were the same except that in the online survey, questions #3 through #17 were skipped if the respondent indicated that the institution did not review grant proposals in any way. Such an answer made those questions irrelevant.

Other Data

Email Comments from Participants

The investigator's email address was included both in the letter of invitation and several times during the online survey. Some respondents chose to initiate an email contact and exchange other information. Relevant points from such email exchanges are included in this manuscript.

In-house Data

During the collection of these data, a running tally was kept of the comparisons of grant proposals with IACUC protocols at our institution. Information was recorded relating to the match or mismatch between these documents, the nature of the mismatch if any, and the resolution that was achieved. In most but not all cases, IACUC protocols were amended to match the grant proposals.

Analysis

Responses to survey questions, including complete text of freeform answers, were automatically stored in an Access database. All answers were loaded into an Excel spreadsheet, and totals and percentages were computed.

Results

Response Rate and Adequacy of Sample

Of the 1002 IACUC chairs or administrators invited to participate in the survey, only 168 responded. The percentage of responses (17%) was small, and it is questionable whether this limited sample is adequate to represent the population. Nevertheless, a number of individuals responded by email that they had no federal funding in their institutions, and the questions in our survey were therefore irrelevant to them. It is impossible to know how many invitations were ignored for this reason, so the actual percentage response was undoubtedly greater than 17%.

Respondent Characteristics

In question #23, we asked for the position/title of the responder. Of those who answered this question, approximately 71% were IACUC chairs, 20% were administrators, and the rest (9%) were attending veterinarians, institutional officials, or other personnel. To glean some information about the sizes of the institutions the respondents represented, we asked for the number of IACUCs (question #19) and administrators (question #20) the institution used. The vast majority (94%) of respondents came from institutions with only a single IACUC, whereas 3% came from institutions with two, 2% from institutions with three, and 2% from institutions with more than three IACUCs. Most respondents' institutions had one IACUC administrator (90%), but some had two (5%), three (2%), and four or more (2%). Most institutions had part-time IACUC chairs (95%), but 5% had full-time chairs. Only 12% of these chairs were compensated (question #22); most (88%) were not.

Grant Proposal Reviews

We asked how the institution arrived at the verification of approval by the IACUC of the grant proposal (question #1). Only two (1%) respondents indicated that their institutions review grant proposals directly, that they do not use protocols. Most (62%) indicated that they review protocols and compare them with grant proposals. A few (6%) responded that they use both of these procedures for different grant proposals. Surprisingly, 30% responded that they do not review or compare grant proposals at all.

It was of interest to learn who compared grants (answers to question #2a are summarized in Table 1). Most respondents indicated that the IACUC or a subcommittee of the IACUC makes the comparison (58%); the IACUC chair and administrator assume this role 11% and 17% of the time, and the veterinarian only 8% of the time.

Grant Comparison Criteria

We asked which part of the grant proposal reviewers looked at in the review process (question #3). Respondents, who were asked to use the RO1 PHS application as a model, could choose from the Abstract; Section a, “Specific Aims”; Section d, “Research Design and Methods”; and Section f, “Vertebrate Animals.” They could select as many of these section headings as appropriate, and they could also add a different section if they wished. Of those who answered this question, 57% look at the Abstract, 49% look at “Specific Aims,” 78% look at “Design and Methods,” and 97% look at the “Vertebrate Animals” section. Two respondents added that they look at the budget—one at personnel, and one at the proposal cover-to-cover.

Next we asked what criteria were used in the comparison of grant proposals with IACUC protocols (question #4). Of those who responded, 33% indicated that the protocol and proposal must be identical in all details of animal usage including specific aims, procedures, and number of animals. Nine percent responded that they allow the number of animals to be different, 18% that they allow the specific aims to be different, and 13% that they allow animal numbers and specific aims to be different. Nineteen percent of reviewers reported that they allow wide latitude; the proposals and protocols must be only roughly the same. Of those who indicated that they use some other criterion, four respondents pointed out either that a proposal may be associated with several protocols or that a protocol may be associated with several grants. In the latter case, protocols may include projects to be covered in future research proposals. One respondent wrote, “Proposals are treated different than protocols. We review proposals and grant Conditional Approval only based on a full review of the final Protocol since the protocol tends to be slightly different than the proposal. Since the protocols are more detailed, we only grant full approval to protocols.” We report this response for completeness, but do not completely understand it.

Time Commitment

To gain some information about the work load generated by the review process, we asked how long it took to complete the average review (question #5). Of those who responded, 40% indicated that an average comparison requires 20 to 29 min, 32% indicated 30 or more min, 24% indicated 10 to 19 min, and only 5% indicated less than 10 min (Table 2). Unfortunately, we did not ask for the total number of comparisons completed in a unit of time, so a real estimate of time commitment is impossible. If other IACUCs complete a number of protocols similar to that for our IACUC, a preliminary estimate is possible. Over the period from September 2004 to September 2005, we performed 40 comparisons. That total would require between 200 min/yr and 1200 min/yr. Although 20 hr/yr is not exorbitant, this time does not include interaction with the PI, reading a revised protocol, interacting with local grants administrators and NIH grant personnel, or other activities that might be generated when there is a failure to match. It seems likely that institutions with larger research programs could perform many more comparisons, requiring much more time.

We designed and emailed a follow-up survey to individuals who provided their email addresses when they agreed to participate. A total of 53 such surveys were sent, and 26 responses were received. Of the 26, 15 were received from institutions that compare protocols with grant proposals. Of this latter group, eight estimated the amount of time they spend doing such comparisons. The average was 238 hr/yr (n = 8), with a range from 20 to 600. This average is considerably larger than our estimated values, making the values in Table 2 underestimates.

Matching

We sought information about how often the protocol was found to match the grant proposal, and the responses to this question (#6) are summarized in Table 3. Perhaps surprisingly, 20% of the respondents indicated that a match (by criteria, as discussed previously) is found in less than 50% of the comparisons. The values in Table 3 are based on estimates made by the respondents; none indicated that the value given was based on actual data. In our in-house sample, we tracked the results of 40 comparisons over a period of 1 yr. Only 23 (58%) of the proposals matched with protocols in the first comparison (i.e., before the protocol was altered).

Question #7 asked about the most common nature of the mismatch if one occurred. Of those who responded to this question, 26% indicated that the animal numbers do not match, 31% that some procedures in the proposal are different from those in the protocol, and 29% that some procedures in the proposal are missing from the protocol. Respondents were encouraged to add their own responses, which included the following: “Training of animal caretakers was deemed insufficient,” “Mismatch is a combination of both different procedures AND lack of proper description,” “It is often the Vertebrate Animal section that doesn't match because it was cut and pasted from another grant application without careful review,” and “Some variation in Specific Aims, objectives etc.” One respondent added, “We have few mismatches and they run the gamut, from different method of euthanasia to different species to different numbers, etc.” Several respondents mentioned that the protocols are often larger in scope than the grant proposals. Presumably, this factor would not be a reason to withhold approval.

Use of JIT

In response to our question (#8) regarding the use of JIT for agencies that allow it, 12% of the respondents indicated that their institutions do not use it at all, whereas 71% use JIT whenever possible. The remainder (18%) reported that they use JIT sometimes but not all of the time when granting agencies allow it. Of this group, two responded that they decided to use JIT on a case-by-case basis, five that they use it for only some agencies that allow it, and four that they leave the decision to use JIT to the PI or to the department/division. One respondent indicated that JIT would be used whenever possible, but there had not been an occasion to do so.

When asked whether there are certain investigators who are excluded from using JIT (question #9), 88% responded that they allow all investigators to use JIT, and 1% that they exclude an investigator with previous noncompliance history. Eleven percent indicated that they exclude some investigators, but the reason given had no connection with investigator qualifications or history.

Review Prior to JIT Request

To obtain information about the timing of protocol review with respect to the time of JIT request, we posed question #10. Twenty percent of respondents indicated that a protocol has always been reviewed and approved prior to the request by the granting agency for JIT verification, and 74% indicated that this is often but not always the case. Only 6% said that this is never the case.

Time Allowed for JIT Verification

Thirty-six percent of respondents indicated that they are typically given 4 or more wk to provide the JIT verification, and an equal number responded that they are given 2 to 4 wk (question #11). However, 24% indicated the typical allowance is 1 to 2 wk, and 5%, less than 1 wk. In response to the question regarding the shortest time allowed for verification (#12), 27% responded with more than 3 wk, 39% with 8 to 21 days, 31% with 1 to 7 days, and 2% with less than 1 day.

Pressure to Approve

Of the respondents, 14% indicated that they have experienced some pressure to provide JIT verification of approval, whereas 86% responded that they have not (Question #13). This pressure was perceived as being exerted by the PI in 60% of cases, by the Institutional Official (IO) or other institutional official in 8% of cases, by the granting agency in 23% of cases, and by some other person in 9% of cases. The other person was identified in one case as a collaborative PI from another institution, in one case as the IACUC itself, and in one case as an outside sponsor.

Changes to Grant Proposals

Only 10 respondents indicated that the IACUC had informed a granting agency that a grant proposal must be changed to conform to an IACUC protocol (Question #17). Estimates of the number of times this situation had occurred ranged from one to 10. The description of the JIT process does not include the potential consequences if the IACUC is unable to accept all of the procedures in the grant proposal. In that circumstance, it is appropriate for the IACUC to require changes. We also asked respondents to describe what they think the NIH would do if the IACUC required a change in the grant proposal. Thirty-two percent of respondents indicated that they believe for small changes, the study section administrator may approve them without consulting the study section, whereas 13% believe that all changes can be approved by the administrator. Nineteen percent responded that they think “significant” changes would require reconsideration by the study section at a special meeting and 20% that such changes would have to await the next regular meeting of the study section. Some other process was selected by 15% of respondents, who indicated the following: The study section administrator should be called directly (1 person); the study section administrator does not have this authority and study sections do not normally reconsider proposals under such circumstances (1 person); small changes could be approved administratively, but larger changes could be evaluated by the primary reviewer from the study section (3 persons), reviewed, and approved by the project officer; or they (respondents) had no idea what would happen (3 persons).

Discussion

Grant Proposal Review/Proposal-Protocol Comparison

As mentioned above, the PHS Policy (1986) is clear in requiring the IACUC to review PHS grant proposal submissions. The last two directors of OLAW have made it clear at IACUC national meetings that the regulations specify, and OLAW expects, IACUCs to review and approve the animal care and use aspects of grant proposals. Verification of this approval must be communicated to the NIH in response to a JIT request. According to OLAW, the grant proposals may be reviewed directly or may simply be compared with local IACUC research protocols. The importance of the consistency of protocols and proposals is underscored in the ARENA/OLAW Institutional Animal Care and Use Guidebook: “. . . it is imperative that the protocol that the IACUC approves is consistent with the information submitted to the PHS. Institutions should devise a mechanism to verify that consistency” (ARENA/OLAW 2002). Given this directive, we were surprised to learn that 30% of our respondents indicated that they neither review the grant proposals nor compare them with IACUC protocols. In a sample of 15 IACUC and grant administrators, Klemfuss (2002) reported that half of their institutions do not review grant proposals. This sample is admittedly small, but the 20% decrease over 3 yr (the author's results compared with ours) could indicate some improvement in compliance.

In our sample, of those who provided a reason for performing grant proposal reviews or comparisons, 37% responded that they did not know they were supposed to do it, 44% that they rely on assurance by the PI that there is a match, 6% that it is too much work, and 14% that there is some other reason. Other reasons included that the institution has not yet established a procedure for this, it is too much work and they do not have the manpower (2 persons), reviewing the proposal implies to PIs that they are not trustworthy (2 persons), the review of grants is not an IACUC responsibility (2 persons), and we do not have grant proposals (2 persons).

Superficially, it would seem that relying on PI assurance might be a sound policy; however, in our institution, PIs often use existing protocols to support new grant proposals. Our experience is that PIs sometimes have only superficial recollection of protocol content. In our in-house data set, 23% of protocols were not consistent with grant proposals at initial review. Based on these data, giving verification of approval would be inappropriate in nearly one fourth of grant proposals if PI assurance were used as the sole criterion. Looking back at Table 3, it is clear that some IACUCs had an even greater percentage of mismatches. In our in-house data set, the mismatch was due to procedures missing from the protocol in 67% of the cases and due to procedures different from those in the protocol in 33% of the cases. In two cases, the mismatch was eliminated by altering the grant proposal; and in the remainder, the IACUC protocol was altered. As indicated in the Results section, our experience is probably not different from that of other IACUCs. Klemfuss (2002) reported that only 10% of the proposals at the University of California at San Diego (UCSD¹) submitted prior to JIT were marked as “pending” because there was no “sanctioned” protocol, because they were missing critical information, or because there were important discrepancies between proposal and protocol. Another 10% of the grant proposals contained “problems that can be resolved with the investigator without protocol revision.” Some of the examples given of such resolvable problems (e.g., conflicting descriptions of animal procedures) would clearly require alteration of the protocol application in our institution. In any case, it appears that accepting PI assurance is not a reliable procedure.

Review Criteria

Most respondents who reported that they review/compare grant proposals indicated that they look at the “Vertebrate Animals” section. OLAW has instructed study sections to require sufficient information in this section for this type of review (Nelson Garnett, former Director of OLAW, personal communication, 2003); however, sufficient information has never been included in our experience. We therefore believe that the use of an IACUC protocol is necessary because the proposal never contains sufficient information for the committee to do its job. It is clear from the responses that most IACUCs look at many parts of the grant proposal to find the information required (see also Klemfuss 2002). Respondents indicated that they look at several parts of the grant application, in some cases including the budget. The approval criteria also vary, ranging from a requirement that the protocol and proposal be virtually identical to allowing a relatively wide latitude in variance, as long as they are “roughly similar.”

In our experience, some latitude is often appropriate. Protocols often include experiments well beyond those covered in a single grant proposal. In fact, it is not unusual for a single protocol to cover a number of different grant proposals. It is less common for a single grant proposal to be associated with more than one IACUC protocol. Comments by some respondents indicated a similar experience.

Time Commitment

One reason given by respondents for not reviewing proposals or comparing them with protocols is that it requires too much time. Unfortunately, we did not request an estimate of the total time spent in this process. However, if our estimates are close and the load of comparisons is close to what we do, then it is unlikely that more than 20 hr are spent per year. This investment of time does not seem excessive for an IACUC (the most frequent reviewer in our survey), especially if the load is distributed across the entire committee. However, if the time estimates in our follow-up survey are typical, IACUCs may indeed spend considerable time in this process.

Use of JIT

Not surprisingly, 67% of the respondents indicated that their institutions use JIT for all of the proposals that qualify for its use. Because the use of JIT is not required by granting agencies, it is also not surprising that 12% of respondents said they never use it. Klemfuss (2002) calculated that strict use of JIT by UCSD would reduce the number of protocols reviewed by as much as 60% or 180 per year. However, we believe that this calculation was based on incorrect assumptions. In our own institution, the minimum time between submission of a protocol application for full-committee review and receipt of approval is 3 wk, assuming the application does not require modification to receive approval. This interim allows sufficient time for the application to be copied, distributed to the committee, reviewed by the committee in advance of the meeting, and then discussed and approved at the meeting. Twenty-nine percent of respondents indicated that the average time allowed by the granting agency for the IACUC to supply verification of approval is 2 wk or less, and 65% indicated 4 wk or less. Furthermore, 31% reported that the shortest allowed time is 7 days or less, with the shortest being less than 24 hr (2%). Given these numbers, a favorable response to a JIT request within these time frames can be problematic if there is no IACUC protocol in place at the time of the request for verification. In our follow-up survey, we requested the average time required for full-committee review and for designated-member review. The average for full-committee review was 29 days (n=23), with a range from 7 to 90 days. The average for designated-member review was 21 days (n=16), with a range from 3 to 60 days. Some responses were given as ranges, in which case we recorded the shortest time provided. Clearly, these averages are much too long for the review process to be completed in the time allowed by the granting agency for JIT verification. We also asked whether the time allowed by the granting agency for this process was enough or too little. Of the respondents, 60% (12 of 20) indicated that it is enough, whereas 40% (8 of 20) reported that it is too little.

Thus, it is not surprising that 20% of respondents indicated that there is always an approved protocol in place at the time of the JIT request, 74% that it is often in place, and only 6% that it is never the case. Recognizing this time constraint, we advise investigators at our institution not to wait for receipt of the JIT request in submitting a protocol application. In our follow-up survey, several respondents indicated that one of the advantages of JIT is that PIs are not required to complete the IACUC protocol with the grant proposal but instead, can complete that task at a more leisurely pace. That advantage results in more “concise” and “precise” submissions; however, to achieve the benefit, the protocol application procedure must be started before the JIT request is received from the granting agency.

We received some emails from respondents who wished to add additional information to responses they had provided in the survey. One respondent wrote, “My opinion from the very beginning was that JIT would not make any difference in our IACUC workload or scheduling. This has proved to be true.” She continued by describing a situation much like our own in which most protocols are in place before JIT requests are issued. Another respondent wrote, “From my experience the NIH does not give the researcher enough time to get their protocols in for review and approval . . . As a result people are not adhering to the JIT and sending in their protocol for review anyway because they know they never had enough time. It is very frustrating from the administration end.”

On one occasion, an investigator at our institution was informed by email on a Thursday that IACUC verification for a large multicenter grant proposal was due by the following Friday at 5 p.m. Fortunately, the PI had already submitted and received approval for an IACUC protocol application. Fortunately too, the coinvestigators in other institutions had previously approved protocols, and the appropriate administrators were in their offices and could quickly send copies of those protocols and the approval letters. We were able to supply verification of IACUC approval on time. It is not clear what would have happened had we failed to supply the verification by 5 p.m. (see below). On another occasion, the IACUC was unable to supply verification of approval because there was a mismatch between the proposal and protocol. We were unable to contact the PI to request a modification of the protocol, and the grant was lost. Respondents gave no report of similar experiences; however, it is clear that in some cases the NIH may have failed to meet its “responsibility to ensure that institutions are given adequate notice to allow for timely IACUC review prior to award” (NIH 2002).

One of the reviewers of this manuscript suggested that the appropriate trigger for IACUC review is the “Priority Score Mailer,” which is sent to the PI once the study panel review is completed, rather than the “JIT Mailer.” Certainly the timing of this notification more closely fits the IACUC review process. However, the Priority Score Mailer is sent only to the PI; the sponsored programs office does not get a copy. It is not the PI who must submit the verification; rather, it is the institution or the IACUC. The PI's response might be to submit a copy of the IACUC's notice of approval of the protocol. This action clearly does not fulfill the NIH requirement. In addition, the NIH has recently announced that it will no longer mail the Priority Score Mailer. Instead, the scores will be posted in the eRA Commons (NIH 2005). To make use of this information, it would be necessary for the sponsored programs office to monitor the eRA Commons continually for all pending proposals. Clearly, this procedure is not workable, especially for large institutions.

According to the initial announcement of JIT, “Under no circumstances may an IACUC be pressured to approve a protocol or be overruled on its decision to withhold approval . . .” (NIH 2002). We have experienced some pressure both from PIs who are awaiting final grant approval and from the sponsored programs office, perhaps in response to the pressure from PIs. The perceived pressure is mild and exerted in the form of repeated inquiries about the status of the review process. Having experienced this pressure, we were surprised that only 14% of respondents indicated a perception of any pressure to approve proposals. The pressure reported was most often perceived as emanating from the PI. Given that a significant portion of PIs' research enterprises depends on grants, it is not surprising that PIs are anxious about receiving IACUC verification quickly. Perhaps the majority of responders actually experienced this sort of pressure but considered it a normal, expected event in the granting process.

Changes to Grant Proposals

We were unable to find any documentation of the procedure the PHS would follow should an IACUC require changes in a grant proposal. According to the initial announcement of JIT, “The existing PHS Policy requirement that modifications required by the IACUC be submitted to the NIH with the verification of IACUC approval remains in effect, and it remains the responsibility of institutions to communicate any IACUC-imposed changes to NIH staff” (NIH 2002). The announcement fails to identify what the NIH will do when it receives this information. Most respondents indicated that they think small changes could be approved by the NIH administrator, and a few indicated that any change could be so approved. Thirty-three percent indicated that significant changes would require review by the study section at a special meeting or at the next regular meeting.

Only 10 respondents indicated that they have informed the granting agency of a necessity to revise the grant proposal. Some of these respondents indicated that they have done so multiple times, the greatest estimate being 10 times. We were unable to design a question to ask the respondents what happened. One of our investigators discussed a change in the proposal with the program director for his particular grant. He quoted the director as saying, “. . . it was acceptable to make the change. He indicated that the IACUC protocol should reflect what we intend to do, not what was described in the grant.” We interpret this statement to mean that it is not necessary for the grant proposal and the protocol to match because it is the protocol that governs what actually will be done. If our interpretation is correct, the message is indeed mixed, when taken together with the written and stated policy.

Conclusions

On the face of it, JIT would appear to be a highly desirable improvement in the grant submission process. It certainly would reduce the number of protocols that an IACUC would have to review and approve; however, this reduction would apply only if submission, review, and approval of protocols could be delayed until after the request for JIT verification. As indicated above, it is not a good idea to wait for the request for verification to begin this process, at least at our institution. Perhaps other institutions have a review process that is much speedier, or perhaps they speed up their normal process for use with JIT. We have done this in our institution, but there is a limit to how much the review time can be reduced and the IACUC still comply with regulations and do a thorough review. This process probably reduces the time-savings that has have been projected for use of JIT. In our follow-up survey, only a few respondents indicated that they shorten the review time, but several indicated that they are more likely to use designated-member review rather than the usual full-committee review. No one admitted to a less rigorous review in these cases.

A related problem with JIT arises due to the lack of standardization across the various institutes at the NIH as well as across the different kinds of grants offered. One letter (from NIH/OER/OEP, dated July 14, 2003, and titled “Request for Just-in-time Information”), which was sent to investigators requesting JIT verification, states that the verification should be “submitted within 2 weeks of receipt of this notice.” However, we (and others who completed the survey) have received requests with as little as 24 hr to supply the required information. In part, these deadlines reflect differing procedures of different institutes or study sections, which apparently set their own procedures. Clearly, there would be some benefit for the NIH to standardize the procedures across all of its institutes and study sections. A point of needed standardization concerns the actual recipient of the JIT request. In some cases, the request is sent only to the PI, and in others the sponsored programs group also receives the request. If the PI receives the only request and delays in informing sponsored programs of the request, the time for IACUC review can be greatly shortened. All JIT requests should be sent to the institution, regardless of whether the PI receives a copy.

It would also be beneficial to publish specific procedures for handling IACUC-required modifications of grant proposals. The statement of grants administrators, that PIs simply must conduct their research in accordance with their approved IACUC protocols, seems contrary to the spirit of the requirement for grant proposal review and approval by the IACUC, whether by direct review or comparison with a protocol. If abiding by the IACUC protocol were truly the sole requirement, then the whole process of grant review by the IACUC would be without purpose.

Finally, it would be reasonable to survey individuals as to whether the purpose of JIT verification of IACUC approval—namely the reduction in the workload of the IACUC—has been realized. Unfortunately, we did not pose this obvious question in the initial survey, but we did include it in the follow-up survey. Of those who responded, 59% (10 of 17) indicated that they have experienced no reduction. Of those who indicated a reduction in work load, the amount of reduction varied from “little” to 40%. Of those who responded, 74% (14 of 19) reported that no problems have resulted from JIT. The problems noted by the others were delays in submission of IACUC protocols resulting in delayed payment of grants, an increase in urgency felt by PIs, and the need for increased communication with the NIH. We also asked about benefits experienced. Approximately 30% (5 of 16) of responders indicated that there is no benefit. Those benefits mentioned by the other responders include reduced PI anxiety, reduced time pressure, better protocols, a smoother process of grant submission, and fewer protocols to review. Thus, perhaps there are real benefits of the process. However, reduced workload for the IACUC does not appear to be a universal benefit.

Abbreviations used in this article: ARENA, Applied Research Ethics National Association; Assurance, PHS Animal Welfare Assurance; IACUC, Institutional Animal Care and Use Committee; JIT, just in time; NIH, National Institutes of Health; OLAW, Office of Laboratory Animal Welfare; PHS, Public Health Service; PI, principal investigator; UCSD, University of California at San Diego.

References

ARENA/OLAW [Applied Research Ethics National Association/Office of Laboratory Animal Welfare]. 2002. Institutional Animal Care and Use Committee Guidebook. 2nd ed.

Kelly PA, Johnson ML. 2005. Just-in-time IRB review: Capitalizing on scientific merit review to improve human subjects research compliance. IRB 27:6-10.

Klemfuss H. 2002. Matching protocols to grant proposals. Lab Anim 31:36-39.

Mann MD, Prentice ED. 2004. Should IACUCs review scientific merit of animal research projects? Lab Anim 33:26-31.

NIH [National Institutes of Health]. 2002. Laboratory Animal Welfare: Change in PHS Policy on Humane Care and Use of Laboratory Animals. Notice Number NOT-OD-02-064. Available online (http://grants.nih.gov/grants/guide/notice-files/NOT-OD-02-064.html).

NIH [National Institutes of Health] 2005. NIH Announces Plans to Eliminate Mailing of Paper Notifications: Summary Statements & Peer Review Outcome Letters. Notice Number NOT-OD-05-075. Available online (http://grants.nih.gov/grants/guide/notice-files/NOT-OD-05-075.html).

PHS [Public Health Service]. 1986. Policy on Humane Care and Use of Laboratory Animals, Section IV, D, 2.