Monday, March 06, 2006

Artificial Blood (Polyheme) Trials : King and Kipnis

Ethicists Nancy M. P. King, Ken Kipnis, and Robert Nelson were kind enough to give us permission to re-print their letter originally posted on the Medical College of Wisconsin listserve:

Date: 06-14-04 22:37

We - Nancy M.P. King and Ken Kipnis - have been doing research on the ethics of waived-consent clinical studies now permitted under 21 CFR Part 50. Such a study is underway: a well-publicized multi-center trial, sponsored by Northfield Laboratories and testing a hemoglobin-based oxygen carrier (PolyHeme) on trauma victims. One of us - NK - had noticed a possibly serious problem that seemed to have been missed, and Robert Orr, a physician evaluating the PolyHeme protocol at the University of Vermont, contacted KK about the same issue. This posting is a request for comments and information that might throw light on the concerns we set out below.

The Northfield protocol provides that trial subjects - trauma patients in hemorrhagic shock who are being treated by EMTs - randomly receive either saline solution or PolyHeme. Enrollment occurs in the field under the waived-consent exception, before arrival at the emergency department. The research subject/trauma patient is likely to be in a medically exigent circumstance and incapable of consent, either because of the injuries or because of the gravity of the situation and the complexity of the consent process. Apart from slowing blood loss and replacing fluids, hemorrhagic shock is not satisfactorily treatable in the field.

Once at the hospital, efforts are to be made to secure consent for continued participation either from the patient/subject or a legally authorized representative. However if formal withdrawal from the study does not occur, participation continues by default during a 12-hour period in the hospital. Patients/subjects in the control group receive standard treatment: saline and blood as needed. However patients/subjects in the experimental group continue to receive PolyHeme instead of blood for oxygen delivery: up to six units of PolyHeme for up to 12 hours, at which point their participation in the trial ends.

The study can be divided into two phases. The first (PolyHeme vs. saline) occurs in the field. The second phase (PolyHeme vs. blood) occurs for up to 12 hours after hospital admission.

We believe it is a serious problem that the special circumstances that justify waiver of consent in the field cease to obtain shortly after admission to the hospital. After they arrive, the patients/research subjects can be typed and transfused, complying with a standard of practice not available to EMTs. Provided that certain other conditions are satisfied, we accept that the use of an investigational product like PolyHeme, instead of saline, can be justified when blood is unavailable (as in an ambulance) and the medically exigent patient/subject cannot consent. But this same investigational procedure becomes deeply problematic if the optimal treatment - blood transfusion - is available but being experimentally withheld without consent.

We believe that, at a minimum, it is obligatory to separate the field trial (PolyHeme vs. saline) and the clinical trial (PolyHeme vs. blood). We think it is an ongoing mistake to be piggy-backing the latter onto the former, with its much weaker consent requirements.

Consider that it is inevitable that some hospitalized patients/research subjects on PolyHeme will die during the critical 12-hour interval when blood is available but being withheld. We expect that plaintiffs' attorneys will scrutinize these deaths in efforts to ground claims of liability. Putting the point most dramatically, these people will have died while being denied an available, standard medical treatment (blood transfusions) following unconsented-to enrollment in a research study. Despite the good results obtained in earlier trials, the use of PolyHeme is still an investigational procedure that can only be substituted for established practices with consent (except under circumstances that plainly do not obtain in the hospital setting).

We are aware of the risks associated with the use of allogeneic blood and appreciate that the availability of a safer oxygen carrier will be a medical advancement. But even though blood is less than ideal, it doesn't follow that it is "unsatisfactory" under the Federal Regs. If it did follow, the waiver could be applied whenever an investigator was sufficiently impressed by an experimental treatment that promised superiority to a less-than-perfect standard treatment. Hemorrhagic shock can be treated in the hospital using blood, though this standard intervention may not be as safe and effective as PolyHeme may someday be definitively shown to be. Of course there is a need for further clinical research comparing PolyHeme and blood, but only with the consent of the research subjects.

We have puzzled over the 12-hour clinical phase of the trial. Emergency departments participating in the Northfield study will typically receive patients/subjects less than one hour post-trauma. But the study is designed to mimic a 12-hour period without access to blood. Unlike remote areas and ships (which do not seem to be participating in this study), 12-hour delays are not common problems in the communities where the studies will be done. So why include this troubling feature? We believe it reflects the circumstances of combat-wounded soldiers when evacuation to field hospitals is impossible. The military could plainly benefit from this new technology, once it is approved. If military applications are the reason for the clinical phase of the study, than the additional risks imposed upon hospitalized civilian trauma victims are intended to benefit, neither the patients/subjects nor those subsequently injured in their communities but rather, soldiers fighting overseas. While both of us endorse the obligation to provide the highest-quality care to injured American troops, we think that duty cannot justify departures from ethical principles governing research on non-consenting civilian human subjects.

These reservations have already been communicated to Northfield but there has been no reply. Both of us are intent on keeping our minds open and that is why we post this message now. We welcome comments and suggestions that can throw light on the concerns we have raised above.

Thank you.

Kenneth Kipnis, Ph.D. (University of Hawaii)
Nancy M.P. King, J.D. (University of North Carolina)

A special issue on the Northfield trial will be forthcoming in AJOB with two pieces by King, Kipnis, and Nelson; they also published an article in the March issue of IRB, on problems that have emerged involving oversight of the PolyHeme study: "Trials and Errors: Barriers to Oversight of Research Conducted un ther Emergency Research Consent Waiver."


Kevin T. Keith said...

This is an excellent piece of detail-oriented ethics, and a provocative problem.

I'd be interested in knowing if they have grounds for their suspicion of a military link, or whether that is just a guess at the application of the data. The suspicion seems reasonable and should be pursued (Question: Who funded the trial?), but I would like to hear more about what suggested it to them.

Anonymous said...

NFLD has funded the trial, which a look at their financial statements (available at will confirm. THey have received minimal government funding. There is military interest, and Brooks Army Hospital in San Antonio is one of 32 participants in the trial.

If one reads the available research and diccussion on trauma, rather than just "puzzling" about it from a lay perspective, there are significant medical reasons for continuing the trial in the hospital. There is a growing body of research that transfusions of RBCs after trauma may cause increases in Multiple Organ Failure (MOF), which is the leading cause of death if the trauma patient survives the first 72 hours after trauma. There is also evidence that PolyHeme may decrease MOF. So while Ms. King and her colleagues have "puzzled", they would be better served to read the medical literature on this topic for their answers.

And while Ms. King and her colleagues agree that there is reason fo test PolyHeme against blood, they state this should be informed consent. (I note that Ms. King is not a fan of informed consent, either:
However, the issues in comparing blood to PH after trauma to examine the impact on MOF raises the same issue with respect to consent that are present in the trauma may not be possible to gain consent from a trauma patient, and it is trauma patients at risk for MOF and the focus of the comparison of blood to PH.

it may also be noted that in the Northfield trial, once the patient is in the hospital, if the patient is not able to give consent, significant attempts are made to contact the family to gain consent. The protocol requires frequent, periodic calls to family during the first twelve hours after trauma, continuing at less frequent intervals over the remaining hospitalization.

Anonymous said...

One thing to post anonymously, another to fail to disclose a financial or professional stake in an issue. "Significant attempts..." If it were my child or parent, I wonder if I'd have a warm feeling about it. The question about this bizarre experiment was well-framed. You impugn the posters for their homework, but I think the financial stakes are too high to ignore, and there are serious questions about this product going back to 2004.