Follow @jonochshorn

Example of human subject research (smog chamber breathing) at the Stanford Research Institute in 1956 (photo by Jack Carrick).

Cornell University is organizing an enormous research project involving human subjects that appears to violate its own guidelines—based on the so-called Common Rule promulgated by the U.S. Department of Health and Human Services (HSS)—as well as principles embedded in the Nuremberg Code of 1947 and the Belmont Report of 1979.

Specifically, Cornell is bringing together thousands of students, faculty, and staff during a global pandemic and subjecting them to in-person instruction where the risk of contracting and spreading the SARS-CoV-2 virus are high. Cornell acknowledges the risk and likelihood of serious consequences: President Pollack wrote, in her June 30, 2020 announcement of Cornell’s reactivation plans that “there is simply no way to completely eliminate risk, whether we are in-person or online; even under the best-case projections, some people will become infected with the SARS-CoV-2 virus, and some will develop the severe form of the COVID-19 disease.” The Nuremberg Code directly prohibits such activity, stating: “No experiment should be conducted where there is a priori reason to believe that death or disabling injury will occur.”

Cornell’s own Institutional Review Board for Human Participant Research (IRB) would, in principle, evaluate such “experiments” with human subjects, but the Board has come to the conclusion that no such evaluation is required. Why? Because they claim that Cornell’s proposal to expose its human subjects to these known risks is not “research.” In an email to me on July 30, 2020, the IRB justified this conclusion as follows:

The regulatory definition of “human subjects research” has two components: the “human subjects” piece and the “research” piece.  While Cornell will be obtaining information and biospecimens from living individuals through intervention and interaction (so, meeting that “human subjects” piece of the definition), the activities are not being performed for the purposes of research. Here is the “research” definition from 45 CFR 46 (the “Common Rule”): 

• 46.102(l) Research means a systematic investigation, including research development, testing, and evaluation, designed to develop or contribute to generalizable knowledge.

Cornell’s surveillance activities are not designed to develop or contribute to generalizable knowledge, but rather, to track the transmission and prevalence of SARS-CoV-2 and COVID-19 in the Cornell community, and develop practices to help curb that transmission. In addition, there is a public health surveillance activities carve-out from the Common Rule (i.e., deemed not to be research):

• 46.102(l)(2) Public health surveillance activities, including the collection and testing of information or biospecimens, conducted, supported, requested, ordered, required, or authorized by a public health authority. Such activities are limited to those necessary to allow a public health authority to identify, monitor, assess, or investigate potential public health signals, onsets of disease outbreaks, or conditions of public health importance (including trends, signals, risk factors, patterns in diseases, or increases in injuries from using consumer products). Such activities include those associated with providing timely situational awareness and priority setting during the course of an event or crisis that threatens public health (including natural or man-made disasters).

The HHS Office for Human Research Protections also published guidance on COVID-19 this past spring, which might interest you.

I hope this helps clarify our determination that Cornell’s current proposed surveillance activities do not meet the definition of “human subjects research.”

In other words, the IRB argues (1) that Cornell’s activities are not “research” because they are not “designed to develop or contribute to generalizable knowledge,” and (2) that public health surveillance activities are, in any case, exempted from review. Both of these arguments, however, can be challenged. On the question of whether Cornell’s proposed activities constitute research, compare the HSS definition of research to Cornell’s own research statement, embedded in its reactivation plan, in which Cornell describes behavioral surveillance and surveying strategies that are designed to “provide valuable insight on subsequent guidance that could be provided to students to reduce transmission”:

Behavioral surveillance is an important tool for monitoring compliance with these directives. A standard survey instrument will be developed to observe adherence in public places (only) on campus, such as classrooms, libraries or dining facilities. In addition, Cornell will monitor infections on a daily basis. In collaboration with TCHD, Cornell Health will receive identified information of positive students and will determine activities or practices associated with becoming COVID-19 positive. Using this information, the university will be able to correlate clusters of infections in individuals sharing residences, classrooms or other activities. This may allow for the identification of places or behaviors associated with an increase in risk of transmission and provide valuable insight on subsequent guidance that could be provided to students to reduce transmission.

In an email reply to the IRB on July 30, 2020, I argued that Cornell’s intended surveillance and survey activities do constitute “research” and that the HSS public health exemption does not apply to these research activities:

First, it seems to me that Cornell’s surveillance and surveying activities are designed to develop or contribute to generalizable knowledge. In their reactivation plan document, Cornell draws upon the results of research conducted elsewhere (for example, Cornell writes that “experiences at other institutions of higher education indicate that there will likely be individuals positive for COVID-19 among those returning to campus” — p.4). In the same way, it can be safely assumed that the results of Cornell’s surveillance and surveying will not merely be used internally but will also be disseminated. This is, by definition, a contribution to generalizable knowledge.

Second, the public health exemption does not seem to apply to Cornell’s research since Cornell’s activities are not “conducted, supported, requested, ordered, required, or authorized by a public health authority.” Cornell is collaborating primarily with Cayuga Medical Center, which is not a public health authority. It is true that Cornell is coordinating some of its reactivation planning with the Tompkins County Health Department (and TCHD is a public health authority), but this relationship is not connected to Cornell’s research on human subjects: TCHD will provide counseling to students who test positive (p.5), TCHD is conducting its own independent “syndromic surveillance system among outpatient providers in Tompkins County” (p.13), and, per state law, TCHD will conduct contact tracing. Similarly, Cornell has consulted with the NYS Department of Health Wadsworth in order to use the Animal Health Diagnostic Center (AHDC) at the College of Veterinary Medicine to expand PCR testing capabilities in Ithaca (p.3). But such ad hoc contact with public health authorities is entirely separate from the independent research—the behavioral surveillance and surveying of their human subjects—that Cornell is undertaking. And these independent research activities are not “conducted, supported, requested, ordered, required, or authorized” by TCHD or by any other public health authority.

The OHRP Guidance exception for “actions taken for public health or clinical purposes” does not seem to apply to the research activities Cornell is undertaking on its human subjects. The example given, of a hospital implementing “mandatory clinical screening procedures related to COVID-19 for all people who come to that institution” is clearly not research, as its purpose is simply to identify patients carrying the virus. Cornell’s surveillance and surveying, on the other hand, is undertaken as research, to enable “identification of places or behaviors associated with an increase in risk of transmission and [to] provide valuable insight on subsequent guidance that could be provided to students to reduce transmission.”

Third, the OHRP Guidance document, in response to this global health emergency, states: “Given the current circumstances, the research community is encouraged to prioritize public health and safety.” Yet Cornell is not reacting to a health emergency, as it was when shutting down the campus last spring. In the present case, Cornell is actually creating the public health emergency for which it has organized its research program with human subjects. The campus was shut down, and Cornell has decided to open it back up knowing that infection, disease, and—possibly—death will inevitably follow. This was a choice Cornell made, not one forced upon it. Other universities have opted not to engage in in-person instruction.

But this should not be a debate about whether Cornell made the right decision, or whether other options would have an even worse outcome. The only question here is whether Cornell’s research on human subjects should be evaluated by the IRB. I believe that the facts and definitions cited above show that Cornell’s planned activities constitute research and that this research involves human subjects.

On July 31, 2020, the IRB sent me the following reply:

The IRB staff have discussed this together with the IRB Chair, and using our collective knowledge of U.S. human subjects research regulations and the compliance community’s accepted interpretation of them, we have determined that Cornell’s surveillance activities are not research, as defined by the Common Rule. As I noted in an earlier email, if any human subjects research is proposed in conjunction with or using data from these surveillance activities, those research projects will require an IRB review.

In other words, the IRB is willing to let Cornell expose its human subjects to a highly contagious and dangerous virus, knowing that—in President Pollack’s words— “some people will become infected with the SARS-CoV-2 virus, and some will develop the severe form of the COVID-19 disease”; is willing to let Cornell implement behavioral surveillance and surveying tools that, in Cornell’s own words, are intended to “provide valuable insight on subsequent guidance that could be provided to students to reduce transmission”; but does not consider any of this to be worthy of IRB review unless the data gathered from these activities are used in what the IRB determines is human subject research, i.e., after Cornell’s human subjects have already been exposed to a virus that is likely to cause severe disease or even death in some of the human subjects. To repeat, the Nuremberg Code states: “No experiment should be conducted where there is a priori reason to believe that death or disabling injury will occur.”

There is certainly some potential ambiguity between what constitutes research and what constitutes ordinary practice, and Cornell’s behavioral surveillance and surveying may well fall into this gray zone. However, it is precisely this ambiguity that makes IRB review essential. As argued in the Belmont Report cited above: “The general rule is that if there is any element of research in an activity, that activity should undergo review for the protection of human subjects.”

There are many problems with Cornell’s reactivation plan. Consider the following (page numbers refer to the PDF linked above):

1. Probably the most serious problem with Cornell’s plan is its stated intention to conduct research on its own human subjects, in apparent violation of protocols established by the Institutional Review Board for Human Participant Research (IRB): see the IRB’s “decision tree.”

Cornell outlines its human-subject research goals in the plan itself, describing how behavioral surveillance and the use of survey instruments will enable researchers to determine which activities or practices—that Cornell itself has put into place—are associated with viral infection with SARS-CoV-2. Knowing that its practices will have a high probability of spreading infection (Cornell President Pollack wrote, in her June 30, 2020 announcement of Cornell’s reactivation plans: “There is simply no way to completely eliminate risk, whether we are in-person or online; even under the best-case projections, some people will become infected with the SARS-CoV-2 virus, and some will develop the severe form of the COVID-19 disease”), Cornell seeks to gather statistically valid information that can only be obtained by placing its human subjects in jeopardy (p.12–13):

Behavioral surveillance is an important tool for monitoring compliance with these directives. A standard survey instrument will be developed to observe adherence in public places (only) on campus, such as classrooms, libraries or dining facilities. In addition, Cornell will monitor infections on a daily basis. In collaboration with TCHD, Cornell Health will receive identified information of positive students and will determine activities or practices associated with becoming COVID-19 positive. Using this information, the university will be able to correlate clusters of infections in individuals sharing residences, classrooms or other activities. This may allow for the identification of places or behaviors associated with an increase in risk of transmission and provide valuable insight on subsequent guidance that could be provided to students to reduce transmission.

[Updated July 29, 2020: Not surprisingly, Cornell’s IRB has determined that “the public health surveillance work currently proposed by Cornell does not meet the regulatory definition of human subjects research, and, therefore, does not require IRB review or approval.” (Email to me from Senior IRB and COI Administrator, dated July 29, 2020). Added July 31, 2020: see my detailed blog post on Cornell’s human subject research.]

2. In describing local medical capacity (p. 3) Cornell states that “Cayuga Medical Center has 212 beds and its three-phase surge plan would increase capacity for a total of 318 critical care and ICU beds and 50 ventilators.” Yet, it is well known that effective treatment of COVID-19 is not only constrained by the number of beds or ventilators, but also by the number of doctors and nursing staff. Cornell does not say how many such trained personnel are available locally.

3. Cornell suggests that “faculty who wish to meet with advisees in person may do so, provided they are able to maintain strict distancing, wear masks and keep their office doors open (for air flow and to increase compliance with health precautions).” (p.6) While modern building codes have eliminated the requirement for fire-rated corridors and self-closing fire-rated doors in many cases, there are still instances where such advice (to “keep office doors open”) would violate fire-safety regulations.

4. Cornell references the ASHRAE guidelines for restarting operations—dealing primarily with HVAC and related systems—but both ignores and contradicts their recommendations in at least one of Cornell’s linked (and restricted¹) documents, stating: “Restore all set-points and scheduled modified during NY-pause.” In other words, the advice is to return to “business as usual” by restoring default set-points, rather than disabling certain default conditions to provide more fresh air circulation. The ASHRAE start-up checklist mentions things like flushing out buildings, and so on, which are not mentioned at all in Cornell’s version.

Where Cornell does repeat some of the AHRAE guidelines that contradict this advice to return to default settings—for example, to disable demand control ventilation (p.8)—there is no mention of the fact that providing sufficient outdoor air in some buildings may not even be possible, since existing HVAC equipment may be incapable of dealing with the extra loads caused by the use of more effective filters (MERV-13 or better) and high humidity associated with the conditioning of outside air (at least in summer; this problem could be reversed in the winter). Where are the actual metrics on a building-by-building, or even a space-by-space basis? Faculty, students, and staff have no idea about the number of air changes per hour (of fresh air) that will be provided in academic buildings and classrooms.

5. Cornell states that “vulnerable” faculty are being “urged” to avoid in-person instruction (p.9): “According to the World Health Organization, people of all ages can be infected by COVID-19. However, individuals over 65, as well as individuals with preexisting medical conditions (such as asthma, diabetes, and heart disease) or compromised immune systems are considered to be at higher risk. While the university cannot compel individuals in higher-risk categories to avoid in-person instruction or other work, the university will urge these individuals to do so.” This is, simply, not true. Faculty (myself included) have never been “urged” to avoid in-person instruction, neither at the university, college, nor departmental level. I’ve received scores of emails from provosts, presidents, deans, and department chairs, none of which suggested that I should not teach in-person because I’m 68 years old. In addition, the statement ignores the obvious fact that even younger faculty who become infected face serious risks to their health, and also risk spreading infection to other household members who may well be more vulnerable than they are. The truth is that the various administrative missives I’ve received often contain an implicit request that more faculty volunteer to teach in-person, since the university’s financial model depends upon having enough in-person instruction so that they can promise students who return to campus that they will have at least one in-person teaching experience.

6. Cornell claims to have booked 1,200 local hotel rooms “for various blocks of time” to provide for isolation or quarantine if needed. But what does this actually mean? Over the course of the summer and the fall semester, there are approximately 100 or more days (let’s say 100) which means that Cornell could have booked a mere 12 hotel rooms per day and met its goal of reserving 1,200 hotel rooms “for various blocks of time.” So how many rooms are actually available on any given day? Is the number 12 rooms per day, or 24 rooms per day, or something else? Cornell doesn’t say.

7. Cornell, in one of its bullet points (p.16) talks about how often its buildings will be cleaned, referring to “frequent cleaning and/or disinfection of high-touch surfaces …” This “frequent” cleaning turns out to be just once a day, except that “high-touch items such as light switches, door handles and push bars, elevator call buttons and handrails … will be sanitized twice per day” (p. 10). One has to wonder about the seriousness of Cornell’s reactivation plan if they can state that wiping down a door handle twice a day is part of a viable strategy for controlling this virus.

Actually, my own experience with Cornell and hygiene, in a related context, may be of interest. See Appendix below.

Appendix
I raised the issue of poor sanitation practices in Cornell’s fitness centers more than a decade ago, in September, 2009. After more than two years of follow-up complaints, Cornell’s solution was, and remains, highly problematic. There are still no sinks for hand-washing in Teagle or Helen Newman Hall fitness centers, and no mandatory wipe-down protocols for equipment. The paper towel dispensers and Simple Green All-Purpose Cleaner that were provided are patently inadequate, especially since this particular cleaner is useless against transmission of the virus causing COVID-19. Here is my initial email complaint from 2009:

On 9/8/09, 7:29 AM, “Jonathan Ochshorn” <jo24@cornell.edu> wrote:
TO: fitness@cornell.edu; Thomas W. Bruce <vpcommunications@cornell.edu>; Gannett Health Center <gannett-mailbox@cornell.edu>
 
The fitness centers at Cornell (I only refer to Helen Newman and Teagle) should take more account of hygiene, especially given the incidence of flu this season. Gannett has written: “HELP STOP THE SPREAD: Watch for seasonal flu vaccine late this month or early October. Currently, flu prevention consists of getting adequate rest and nutrition, and practicing excellent hand hygiene.” (www.gannett.cornell.edu). <http://www.gannett.cornell.edu/>
 
There is no access to water for washing hands in the two fitness centers refer to, and no non-water hand-washing options are made available.
 
There is no systematic process for cleaning the equipment after each use. The current policy of intermittent cleaning is useless. It is essential that the equipment be cleaned after each use. This cannot be voluntary or discretionary.
 
At the point where uniforms and towels are issued, “fresh” towels are placed on the same counter as “soiled” towels. Personnel do not seem to have a systematic protocol in place for the sanitary transfer of soiled uniforms so that they do not contaminate the fresh ones. They also seem to put themselves at risk.
 
If Cornell is serious about reducing the spread of disease, please don’t just issue emails: take productive and effective action.
 
Thanks.
     
Jonathan Ochshorn
Associate Professor

After two years of repeated follow-up emails, Cornell, in early 2012, finally agreed to set up a committee of in-house people to figure something out, which ultimately resulted in the paper towel dispensers and Simple Green Cleaner. I responded to this “solution” later in 2012:

From: Jonathan Ochshorn <jo24@cornell.edu>
Date: Thursday, November 29, 2012 at 12:17 PM
To: Beth McKinney <bm20@cornell.edu>
Cc: Janis I Talbot <jit1@cornell.edu>, Tisha Lea Tipping <tlt29@cornell.edu>
Subject: Re: Fitness Follow Up
 
Beth,
 
Some follow-up on the issue of sanitation practices at the Cornell fitness centers:
 
Providing cleaning materials for fitness center users is an improvement. However, by leaving it as a voluntary activity, the transmission of infections — which affects the entire Cornell community — is not addressed. This is because very few users seem to be using the available cleaning material. It really needs to be made mandatory, and an educational component needs to be built into Wellness (or other) memberships.
 
The actual materials provided are not satisfactory. The paper towels fall apart when used; the blue cloth towels work much better, but I was told yesterday that they are not available for users, only for staff. And no cleaning materials have been provided for the floor mats.
 
Finally, the poor sanitary practices at the issue rooms (at least in Teagle) persist: clean towels and gym uniforms are placed on the same counter surfaces with soiled uniforms and towels. While the issue rooms may not be within your jurisdiction, it seems to me that sanitation practices should be coordinated centrally — someone needs to take the initiative to provide a consistently safe exercising and fitness environment.
 
Thanks again for your interest in these issues.
 
Jonathan Ochshorn

In the end, Cornell just said no: supplying paper towels and Simple Green Cleaner was all they were prepared to offer:

From: Beth McKinney <bm20@cornell.edu>
Date: Friday, November 30, 2012 at 2:35 PM
To: Jonathan Ochshorn <jo24@cornell.edu>
Cc: Janis I Talbot <jit1@cornell.edu>, Tisha Lea Tipping <tlt29@cornell.edu>, Kerry Howell <kk253@cornell.edu>, “Frank A. Cantone” <fac2@cornell.edu>, Mary J Adams-Kucik <mja12@cornell.edu>, Mark Joseph Bilyk <mjb17@cornell.edu>
Subject: RE: Fitness Follow Up
 
Dear Jon,
 
I am happy to reply to your concerns on behalf of Cornell Fitness Centers, Gannett Health Services and Environmental Health and Safety.
 
We appreciate your conscientious efforts to help the fitness centers minimize the spread of infections. The team that was created to address these concerns consisted of representatives from the Cornell Fitness Centers, Gannett Health Services, CU Wellness Program, Recreational Services, and Environmental Health and Safety. After several meetings, these subject matter experts agreed the most effective way to mitigate the problem was to provide what you currently see in the fitness centers.
 
At this time, we remain firm that the current procedure has improved the situation and is working well for our members. There is nothing more that our group plans to change at this time.
 
I will, however, let Mark Bilyk, who oversees the issue rooms, know about your concerns regarding the issue room procedures. 
 
Best,
Beth

Note:

¹ Following the link on p.7 to “please visit Cornell’s Environment, Health and Safety website,” we are taken here.  I went to the “Facilities Start-up Checklist” since it seemed to correspond to ASHRAE’s start-up advice, but the site required a Cornell netID. Yet, where ASHRAE’s start-up advice was comprehensive and nuanced, the Cornell start-up advice was generic and incomplete.

I’ve posted two new papers for your reading pleasure. “Utility’s Evil Twin: The Function of Venustas and the Fear of Reality” was first published in the Cornell Journal of Architecture: 11 (March 2020). “Sullivan’s Eagle: Form and Function Artistically Considered” was accepted for the Proceedings of the 108th Annual Meeting of the ACSA but that San Diego conference, originally scheduled for March 2020, was postponed/canceled because of the coronavirus.