On September 15, 2006, Patricia Sachs Catapano, Esq., an associate general counsel of Columbia University in the City of New York and in-house lawyer for its Medical Center, wrote a surprising letter to Doctors Marc Dickstein and Mark Heath, staff anesthesiologists who had doggedly criticized the university’s handling of a long-standing controversy concerning a 1999-2001 research study known as Protocol #9256. Writing “on behalf of Dean Lee Goldman and the University,” Catapano informed Dickstein and Heath that “Dr. Goldman is willing to undertake a complete chart review of each patient in the Study”.(1)
After years of denials and obfuscations, determining whether patients had actually been harmed was a triumph of sorts, albeit a belated one. By 2006, many of Heath’s and Dickstein’s concerns about the study had largely been confirmed. Columbia had acknowledged, and the federal government’s Office for Human Research Protections (OHRP) had accepted, that the Medical Center had failed to inform patients of either the study’s true purpose or its inherent risks. Dr. Elliott Bennett-Guerrero, the study’s architect and principal investigator, had been sanctioned by Columbia and removed from his supervisory position and tenure track.(2)
Yet, while Columbia had previously investigated claims of patient harm in 2002, conflict continued because Dickstein and Heath maintained that Columbia’s investigations were incomplete or otherwise contrived to reach a no-harm conclusion. While a complete chart review was clearly a step in the right direction to determine patient harm, Catapano’s letter remained worrisome.
As Dickstein and Heath wrote Goldman, “We would like to underscore that the questions that need to be addressed are ones of science, research ethics and institutional policy; while these questions may be informed by legal counsel they need to be answered by physicians and scientists.”
In the story of Columbia’s three investigations into Protocol #9256, the facts suggest that had these inquiries been scientific studies rather than administrative inquiries, the parties responsible would likely have been found guilty of scientific misconduct arising from their deliberate omission of relevant data, disregard of undesirable results, misleading use of statistical methods, and framing of conclusions inconsistent with objective findings. Labeling this “administrative misconduct” somewhat understates the ethical transgressions.
Overall, the story suggests that Columbia has been covering up a guilty secret, acting more like a malfeasant corporate enterprise than a center of higher learning. Viewed through a larger lens, Columbia’s behavior also illustrates the limitations of self-policing. Despite extensive federal regulations and the university’s impeccably high standards of scientific conduct, faced with potential legal liabilities, Columbia compromised its principles, dragged its feet investigating charges of whistleblower retaliation(3), and held fast to its denials of patient harm for nearly a decade.
Background: The Study in Brief
Titled “Effect of Different Intravenous Fluids on Thromboelastography [TEG] During Cardiac Surgery,” the study began in December 1999 and ended fourteen months later. It was described as safe: its title was innocuous, its formal description to the Institutional Review Board (IRB) that had to approve it was benign, and the risks disclosed on the mandatory patient consent form would raise no concerns in anyone’s mind. Given the stated minimal risks, the study team had even applied for a regulatory waiver to bypass the required patient informed-consent process, misleadingly stating that the only identified risk, the withdrawal of a small amount of blood for TEG analysis, “will not increase your chances of needing a blood transfusion.”In reality, Protocol #9256 was a continuation of its creator’s work: Dr. Bennett-Guerrero, then Columbia’s chief of cardiothoracic anesthesia, had for some years been investigating the clinical differences between various fluids used for “blood expansion” after injury or during surgery. He was also a past consultant and paid speaker for BioTime Inc. – developer of a newly approved fluid, Hextend, one of the four fluids he proposed to study at Columbia, and a paid speaker for Abbott Laboratories – Hextend’s distributor and the large pharmaceutical firm that had indirectly sponsored the Columbia study with an unrestricted grant.(4) In contrast to the headlined TEG measurements, it would become clear in time that Bennett-Guerrero was primarily interested in the differences in clinical outcomes – bleeding, required returns to the O.R., and kidney complications – associated with each fluid. In order to make definitive comparisons, the protocol required each of the more than two-hundred patients randomly assigned to one of four study sub-groups to receive only one fluid during their bypass or heart valve replacement surgery, a departure from standard hospital procedure that called for a combination of fluids.
While Bennett-Guerrero and associates would later argue in formal recorded interviews conducted by Columbia(5) that the known clinical outcome differences between the fluids were indeterminate, Bennett-Guerrero’s prior research as well as other studies had found that patients who received one of the study fluids, Hespan, at dosages that exceeded 20 milliliters per kilogram of body mass over a 24-hour period tended to bleed excessively. While not conclusively established, Hespan’s side effects were sufficiently worrying that a “routine safe practice” had evolved to limit high-volume use of Hespan, especially in cardiac surgery. In Columbia’s operating rooms, as in most other hospitals, the simple rule of thumb was never to administer more than 1.5 liters of Hespan to patients.
Nevertheless, by design Protocol #9256 placed no limits on the amount of any of the study fluids that could be given, including Hespan. In some cases, over three times the normal safe limit was administered, potentially a massive overdose. A number of patients in the Hespan group developed complications, and two died. There were no deaths in the study’s other sub-groups.
While the focus on clinical outcomes and the risk of some bleeding complications were added to the Columbia IRB document just before the study began, it appears that the IRB was misled about the study’s risks arising from the history of bleeding complications associated with Hespan. As a result, no changes in study purpose and risk were ever reflected in the consent form or communicated to the two hundred patients who agreed to participate. Patients were given to understand that despite receiving no benefit beyond a contribution to medical knowledge, participating in the study would present virtually no additional danger.
In June and July 2001, Bennett-Guerrero summarized the study’s results in a series of hospital presentations, one of which was recorded on video. He described the clinical outcomes his research was investigating and the study’s results. In particular, he highlighted that the patients in the Hespan group had bled more, required more transfused blood products, and were considerably more likely to have to return to the operating room for additional surgery. Bennett-Guerrero’s results were also included by BioTime and Abbott Laboratories in documents prepared for and presented at the June 2002 FDA Blood Products Advisory Committee meeting that discussed the bleeding risks associated with Hespan and Hextend. Bennett-Guerrero’s results were specifically used to support the contention that the recently approved Hextend was sufficiently safer than its older competitor to not require the additional safety warnings that the FDA was contemplating adding to the Hespan label.(6)
As one of the operating room anesthesiologists participating in the study, Marc Dickstein had a specific interest in and direct contact with the details of Protocol #9256 as it affected his patients. In his 2002 interview, he recalled that he had concerns that were only partly assuaged by Bennett-Guererro. However, Dickstein’s faith in the IRB process encouraged him to suppress his concerns.
Not long after, however, Dickstein changed his mind when patients who should not have bled after their surgery had to return to the O.R.: “You can’t be conclusive about the implication of a single patient bleeding. But there were several of those patients. And everyone’s concern at that time was heightened that there was a possibility this was related to the study.” By November 2000, eleven months into the study, Dickstein and Heath had become sufficiently concerned to approach Paul Papagni, administrative head of Columbia’s IRB and responsible for the safety and ethics of the Medical Center’s research programs. Surprisingly, Papagni did not immediately put the study on hold, nor did he launch a thorough investigation into Dickstein’s and Heath’s concerns about patient harm. His report, delivered October 10, 2001, eight months after the study ended, was critical of the IRB process and of Dr. Bennett-Guerrero for failing to conduct the research in an ethically proper manner, but it made no comments about patient harm despite Bennett-Guerrero’s conclusions that patients receiving Hespan had experienced greater bleeding and other adverse clinical outcomes.
The Second Investigation
Frustrated, in December 2001 Dickstein and Heath wrote a detailed letter to then-Dean Gerald Fischbach to reiterate and more completely document their concerns about research oversight and patient protection, especially regarding patient harm arising from Protocol #9256. Several months later, in February 2002, Dean Fischbach informed Bennett-Guerrero and Anesthesiology Department Chair Margaret Wood that the results of the research were not to be published.(7) Formal publication was largely moot, however, since Bennett-Guerrero’s results had already been publicly previewed in abstract form. Furthermore, the lack of a ban on private circulation of the findings meant that they could be presented before the FDA Blood Products Committee meeting in June, where they remain available on the FDA Web site to this day.
In May, Fischbach launched a second committee to investigate the Bennett-Guerrero study. In addition to interviewing all the key players and other involved parties in depth, the 2002 committee appointed subcommittees to review the clinical impact of the study and its statistical results. However, the committee determined that neither Dickstein and Heath’s additional charges of retaliation arising from their whistleblowing nor the commercial conflict of interest issues arising from the study’s funding and Bennett-Guerrero’s relationships with BioTime and Abbott were within their scope, referring them to Dean Fischbach for follow-up. The retaliation issue has never been resolved, nor has there been any official investigation into the possible commercial influences on the design and timing of the research or its relationship to Hextend’s commercial prospects.
While the committee was strongly critical of many procedural and ethical issues, according to a summary to OHRP dated September 30, 2002, they found neither “evidence of harm to any particular patient that could be attributed to the study treatment” nor “evidence of harm to the Study population as a whole.”(8) The foundation for these conclusions, the data on which they were based, and the statistical methods used would be at the center of the controversy for the next seven years.
The Third Investigation: The 2007 Committee
After the medical records review described by Catapano’s September 2006 letter, analysis of the results was led by Bruce Levin, professor of biostatistics at Columbia and an expert on statistics and the law. Levin collaborated with Professor Howard Andrews.
Levin found that applying the committee’s own criteria for harm to the newly collected data led to results that “do not accord” with the committee’s official conclusion that “the conditions necessary for a finding of patient harm had not been met.”(9) Levin’s finding supports the long-standing results presented by Bennett-Guerrero in 2001 as well as Dickstein’s and Heath’s reanalysis, both of which concluded that patients in the Hespan group, in particular, had suffered adverse clinical outcomes and thus had been harmed.
Inexplicably, although the 2007 committee’s written report also noted a number of important and statistically significant adverse clinical outcomes between the treatment groups, and Levin’s report and subsequent additions at the committee’s request merely added to the proof case, on March 31, 2009, Senior Vice Dean Steven Shea wrote to the OHRP that despite these findings the “conditions necessary for a finding of patient harm had not been met” because “differences in clinical outcomes are expected in trials, and that a finding of such differences does not in itself mean that patients were harmed by participation in the trial.”(10) This conclusion explicitly contradicts Levin’s determination that applying the statistical evidence to the committee’s a priori criteria did not support a no-harm conclusion. On June 8, Kristina Borror, Ph.D., OHRP’s director of the Division of Compliance Oversight, using a separate set of criteria, stated that “some study subjects appear to have suffered actual harm as a result of their randomization,” thereby also disagreeing with Columbia’s official position.(11)
While research studies often do create differences in clinical outcomes, as Dean Shea correctly argues, the risks patients take in research programs are generally offset by the possibility of benefit. On balance, no arm of any trial is supposed to be knowingly more dangerous than another. Under federal regulations, it is the obligation of the research team to ensure that patients make benefit vs. risk trade-offs with complete knowledge of the known facts. In this case, however, there were no potential gains (since the existing hospital fluid regimen was already well-proven and considered safe), and there was a one in four chance of being exposed to large quantities of Hespan, a fluid that had a long history of bleeding complications. If there was any uncertainty – and it was a key one – it was whether Hextend manifested similar risks. As it would turn out, Levin’s findings cast considerable doubt on Hextend’s safety(12), and recent research would indicate that Hextend use is associated with bleeding complications at dosages as small as one liter.(13)
Why Was the Study Designed the Way It Was?
None of Columbia’s three investigations examined the possible influence of commercial factors, which are an obvious consideration. In 1998, equity research firm Asensio & Company had put out a “strong sell” recommendation for BioTime Inc., the makers of Hextend, arguing that its similarity to Hespan would cause it to sell at a substantial discount to the more established product.(14)
BioTime was well aware of this weakness, however. In a 1997 press release, Victoria Bellport, then the company’s chief financial officer and executive vice president, clarified the known problems with Hespan and the company’s hopes for Hextend:
“The reports of bleeding problems with Hespan have been carefully addressed by BioTime scientists and their medical collaborators…. Hextend is currently in Phase III double-blinded clinical trials at Mt. Sinai Medical Center in New York, and at the Duke University Medical Center. In these trials, Hextend’s ability to replace substantial quantities of blood volume … is being compared to Hespan … [in] amounts varying from one-half to five liters of test solution. Based on the extensive pre-clinical experimentation that showed the superiority of Hextend, it is our firm belief that if similar results are shown in human clinical trials, Hextend sales will eclipse those of the present hetastarch solutions.” [Emphasis added.](15)
The clinical trials to which Bellport refers are those in which Elliott Bennett-Guerrero participated. These studies produced an important set of results that both underscored the bleeding complications associated with Hespan and the relative superiority of Hextend. The new fluid’s benefits were later touted in BioTime’s 2008 annual SEC filing.(16)
As Protocol #9256 was being designed in 1999, two years after Bellport’s press release, it seems likely that as an expert in this field and a paid speaker and consultant for BioTime, Bennett-Guerrero would have been aware of the company’s hopes for its new product as well as the past complications associated with Hespan, its primary competitor.
Judging from the Protocol’s design as well as from various statements about its purpose, the study was hoped to definitively establish the clinical differences between the fluids, especially at high volumes. With regard to risks, one might generously believe that the consent form’s omission of the dangers associated with large amounts of Hespan was an oversight, but the Protocol team’s prior request for a consent form minimum-risk waiver was clearly deliberate. In light of Bennett-Guerrero’s earlier research findings, omitting the risks appears irresponsible as well as counter to federal regulations. On the other hand, in order to manipulatively gain IRB approval and to induce patients to participate in a potentially dangerous study that offered them no medical benefits, minimizing the apparent risk appears to have been a necessary expedient.
An additional argument supporting the commercial motivations behind Protocol #9256 was Abbott Laboratories’ unrestricted grant that indirectly funded it. While there appears to be no overt tie between the grant and the study, Abbott’s role as Hextend’s distributor raises the likelihood that the company’s motives were tied to its hope that Bennett-Guerrero’s research would improve Hextend’s attractiveness relative to Hespan.
Support for this view comes from Abbott’s and BioTime’s inclusion of the Bennett-Guerrero results in their submission to the June 14, 2002, FDA Blood Products Advisory Committee.(17) Subsequent to that meeting, additional warnings were added to the Hespan label, but were not required for Hextend.
Why Has Columbia Expended So Much Effort to Avoid the Conclusion of Patient Harm?
The deep involvement of Ms. Catapano and Jane Booth, Columbia’s general counsel, throughout this decade-long controversy suggests that the university’s legal position was of ongoing concern to Columbia’s top administration. It is not difficult to see that if harm were established, and especially since the only deaths occurred in the Hespan group, that there was a possibility of litigation as well as damage to the university’s reputation and to the reputations of those members of the Medical Center staff who had misrepresented the study’s risks, biased its three investigations and misrepresented known facts about patient harm to OHRP in violation of federal regulations.
One possible explanation for Columbia’s behavior is that it was hoped that if the inquiry could be dragged out long enough the controversy would eventually disappear. In fact, in 2003 OHRP accepted both the university’s admissions of procedural and ethical violations and its proposed administrative remedies, putting the matter to rest until Dickstein and Heath raised the issue of patient harm again in 2006 after the arrival of Dean Goldman.
Despite Columbia’s admissions of administrative failure, however, its investigations were never led by independent parties, and were not, until 2007, based upon a review of all the pertinent clinical outcome data. Even putting all such questionable actions to one side, the most visible indication of systematic, self-serving bias was the administration’s drawing of bottom-line conclusions that were inconsistent with the evidence, a clear breach of scientific ethics.
The 2002 Committee and Ms. Catapano refer to examining the harm to the “study population as a whole.” Applying the criterion of harm done to the entire population guaranteed a no-harm conclusion. It was the statistical equivalent of “cooking the books.”
A similar misrepresentation arises from the question of whether bleeding could be attributed to the study for any individual patient. While such causal determinations often confound many individual medical liability cases, such those related to Merck’s Vioxx, an argument of individual harm was never made here, nor is one necessary. By Columbia’s own written admissions, patients in the Hespan group did suffer a number of adverse clinical outcomes at rates significantly greater than other groups, and it is precisely these outcomes that the 2007 committee had established as its criteria for patient harm. The conclusion of harm should therefore have been automatic, yet Dean Shea wrote to OHRP that the conditions for harm had not been met.
Beyond this, to determine whether patients were harmed as a result of the study, we need go no further than Kristina Borror’s simple argument: Since the study’s risks were not disclosed, patients could not choose whether they wished to accept them. Consequently, any avoidable adverse outcomes that affected groups of patients at higher than the rates of Columbia’s normal clinical practice were the result of the study’s design. This is fundamentally different from cases in which patients voluntarily accept risks in the hope of some otherwise unobtainable medical benefit.
Conclusions and Implications
Protocol #9256 is not just a story about a wrong-headed clinical trial, ethical failures or even about harm to patients, as serious as that might be. For nearly a decade, Columbia has manifested a form of “institutional denial” by engaging in a widespread program of obscuring the reasons for the study, the harm actually done, the results of internal analyses and the behind-the-scenes influences that were doubtless necessary to induce sophisticated academics and administrators to ignore both the facts and the unambiguous conclusions of some of their peers.(18)
While we might cynically expect such behavior from profit-seeking corporations, that it should come from a premier institution of high learning seems a more serious transgression. Universities are afforded wide discretion in their governance, particularly their right of self-examination. In return, they are expected to hold fast to the highest standards to ensure that their members’ conclusions are objectively based upon the evidence. In research, failure to abide by these standards is known as scientific misconduct, a dangerous poison that affects the progress of science world-wide and the trust that the public has in the advice and guidance scientists provide to society. Columbia University, of course, unambiguously promotes and supports the highest standards of research integrity.
In light of the known facts, this study should never have been performed as conceived, irrespective of its potential gains in knowledge. However, even though a safer dose-escalation study design might have been approved by the IRB and seen as safe by patients, it might not have been completed in time for the commercially important FDA meeting in June 2002.
With respect to the oversight of clinical trials and the value of external watchdogs in general, a number of weaknesses clearly need correcting. The OHRP is the primary organization that oversees organizations’ compliance to the federal government’s regulations regarding the treatment of human subjects. While its regulations and guideline documents are largely comprehensive, they do not provide sufficient guidance to redress compliance failures. In addition, my research strongly suggests the importance of outside perspectives in the anticipation, prevention, diagnosis and correction of failure.(19) Beyond this, internally voiced ethical and safety concerns, such as those in this case, need an independent, outside ear. At Columbia, alarms were raised in sufficient time for swift intervention to have prevented further harm. Unfortunately, no action was taken for seven months, by which time the study had ended.
In spite of the societal benefits of external review, however, we cannot expect institutions like Columbia to voluntarily accept greater outside scrutiny. Organizations, like individuals, are programmed to keep secrets, especially those known as “dark secrets” that are deeply embarrassing or potentially threatening to achieving their immediate goals. While all organizations – like people – occasionally make mistakes, their instinct is to cover them up and punish those who would reveal the inconvenient truths. Although cover-ups often bring harm to the organization as well to its patients, customers, employees and the larger society, institutions nevertheless impose codes of silence and punish their whistleblowers. Unfortunately, in the mental calculus of some leaders, taking the risks of a cover-up appears worth it if such a gamble provides a chance of escaping loss completely, or even coming out a little ahead. That seems to be the case here. Perhaps the biggest remaining question, therefore, is whether Columbia’s trustees will let those responsible get away with it.
See also Jeanne Lenzer and Shannon Brwonlee’s article, “Government Orders Colombia to Tell Patients True Nature of Drug Study: Officials Say Research May Have Caused Harm to People Who Had Heart Surgery.“
(1). Catapano, Patricia Sachs, assistant general counsel, Columbia University. Letter to: Mark Heath and Marc Dickstein, 15 September 2006. Copy furnished by Mark Heath.
(2). Bennett-Guerrero sued Columbia, but eventually settled in a confidential agreement. He is now at Duke University. Elliott Bennett-Guerrero v. The Trustees of Columbia University, available at: http://www.scribd.com/doc/20643265/Columbia-Drug-Study-Elliott-Bennett-Guerrero-v-Columbia-University, accessed 10/12/2009.
(3). According to Mark Heath, Dickstein and Heath first discussed the matter of retaliation with Dean Gerald Fischbach on July 30, 2001, and they refer to it in their second letter to him regarding Protocol #9256 dated December 20, 2001. Despite the involvement of lawyers on both sides, however, the matter has dragged out uninvestigated and unresolved since that time. Most recently, despite a verbal commitment by Columbia’s external “independent” counsel to complete the matter by the end of January 2010, no visible progress has been made.
(4). Abbott Laboratories, “Issues in Fluid Volume Replacement,” Anesthesiology News and Cardiology Special Addition, undated.
(5). These recorded interviews are available as part of the documentation submitted re: Elliott Bennett-Guerrero v. The Trustees of Columbia University in the City of New York and Gerald Fischbach, Index: 101260/03, January 23, 2003, Supreme Court of the State of New York, County of New York.
(6). See Blood Products Advisory Committee, Department of Health and Human Services, Food and Drug Administration, Center For Biologics Evaluation and Research, transcript of meeting of 14 June 2002, page 173, http://www.fda.gov/OHRMS/DOCKETS/ac/02/transcripts/3867T2.rtf, accessed 11/21/2009 and Wangelin, James (Regulatory Affairs Associate Director, Abbott Laboratories). Letter to Dr. Linda Smallwood, FDA, Center for Biologics Evaluation and Research, 24 May 2002. http://www.fda.gov/OHRMS/DOCKETS/AC/02/briefing/Hextend.pdf, accessed on 15 September 2009.
(7). Consistent with regulations, journals require that patient consent be obtained, and the university’s failure to obtain such consent would have to be disclosed in the article if the journal editors agreed to waive this requirement.
(8). Corwin, Steven (senior vice president and chief medical director, New York Presbyterial Hospital) and Morris, Thomas (vice president for heath sciences, Columbia University). Letter to Michael Carome (director, Division of Compliance Oversight, OHRP), September 30, 2002, OHRP records, Rockville, Maryland, released under Freedom of Information Act.
(9). Levin, Bruce. Letter to Dean Lee Goldman, 19 May 2009, OHRP records, Rockville, Maryland, released under Freedom of Information Act.
(10). Shea, Steven. Letter to Kristina Borror (director, Division of Compliance Oversight, Office For Human Research Protections, US Department of Health & Human Services), 31 March 2009. OHRP records, Rockville, Maryland, released under Freedom of Information Act.
(11). “Subjects in this study had the type of volume expander they received determined by randomization for the research purpose of comparing the clinical outcomes of the use of four different fluids, not by their physicians exercising clinical judgment to determine which fluid was best for them. Subjects in two of those groups experienced a statistically significant higher rate of identified negative clinical outcomes, including bleeding events (requiring use of transfusions) and decreases in renal function. Beyond that, there was the trend toward increased need for re-operation among those groups of subjects. The statistical analysis supports the hypothesis that the explanation for why some clinical outcomes for patients in two groups were worse than in the other two groups is attributable to which fluid was used. Accordingly, it appears that the consequences of randomization in the research study influenced the chances that subjects would suffer harm. Because of the recognized inadequacy of the informed consent process in this study, subjects were not informed of an identified risk of the study, and some study subjects appear to have suffered actual harm as a result of their randomization.” Borror, Kristina (director, Division of Compliance Oversight, Office For Human Research Protections, US Department of Health & Human Services). Letter to: Steven Shea, M.D. (senior vice dean, Columbia University Medical Center), June 8, 2009. OHRP records, Rockville, Maryland, released under Freedom of Information Act.
(12). Levin, Bruce. Letter to Dean Lee Goldman, 19 May 2009, op cit.
(13). Heckt-Dolnik, Marketa et al, “Hetastarch increases the risk of bleeding complication in patients after off-pump bypass surgery: A randomized clinical trial,” Journal of Thoracic Cardiovascular Surgery, 2009; 138:703-11.
(14). Asensio, Manuel, “Strong Sell Recommendation, BioTime Inc. (OTC Symbol: BTIM)”, 29 April 1998. http://www.asensio.com/Data/Pdfs/389.pdf, accessed 11/15/2009.
(15). Bellport, Victoria, “Reply to the erroneous comments made about the company in Alan Abelson’s column ‘Up & Down Wall Street’ on page four of the April 28, 1997 edition of Barron’s,” http://www.thefreelibrary.com/BioTime+replies+to+Barron’s-a019356473, accessed 1/14/2010.
(16). “An important goal of the Hextend development program was to produce a product that can be used in multi-liter volumes. The safety related secondary endpoints targeted in the U.S. clinical study included those involving coagulation. We believe that the low incidence of adverse events related to blood clotting in the Hextend patients demonstrates that Hextend may be safely used in amounts exceeding 1.5 liters. An average of 1.6 liters of Hextend was used in the Phase III clinical trials, with an average of two liters for patients who received transfused blood products.” [Emphasis added.] BioTime, Inc., SEC filing 10Q, 1998, p.12, http://www.sec.gov/Archives/edgar/data/876343/000087634309000009/form10k2008.htm, accessed 14 January 2010.
(17). On June 14, 2002, the FDA’s Blood Products Advisory Committee discussed changes in labeling for hetastarch products, particularly whether Hespan and Hextend should continue to have the same warnings and contraindications. During this meeting, Dr. Tong J. Gan of Duke University Medical Center and Elliott Bennett-Guerrero’s co-author of the article describing the Duke-Mt. Sinai Hextend clinical trial, summarized the results of Bennett-Guerrero’s Columbia research as supportive of the superiority of Hextend over Hespan.
(18). It is particularly telling that much of the investigation has been kept secret from members of the Columbia community, including members of its Institutional Review Board. See Dodson-Lucas, Sybil et al. Letter to George C. Gesparis (assistant vice president and senior assistant dean for research ethics and executive director, Institutional Review Board, Columbia University and Columbia University Medical Center), 15 May 2009, OHRP records, Rockville, Maryland, released under Freedom of Information Act.
(19). Gerstein, Marc S. and Schein, Edgar H., “Dark Secrets: Face-work, Organizational Culture and Disaster Prevention,” paper delivered at Workshop on Forecasting, Warning and Preventive Policy, Department of War Studies, King’s College London, September 18-19, 2009. Available from the author.
See also Jeanne Lenzer and Shannon Brwonlee’s article, “Government Orders Colombia to Tell Patients True Nature of Drug Study: Officials Say Research May Have Caused Harm to People Who Had Heart Surgery.”