Data Suppression

I'm curious to know what researchers think about data suppression and whether they consider it a human subjects issue, academic ethics issue, or something else.

Howard Brody writes about recent new evidence of pharmaceutical companies practicing data suppression for financial gain. But if pharma can do it, surely an individual investigator could.

Without a doubt, data suppression is difficult to detect from the outside. Big pharma collects data from all of their research sites and analyze, clean, and process data without transparency. To manipulate data is frankly too easy and financially tempting.

As you might imagine, I see this as both a human subjects protections issue and an academic/business ethics issue. Let us not fool ourselves, the University is a business center as much as an educational one.

But before we begin to catch data suppression, we as a research community must agree it is a priority for science and the health care industry, and then we can address who should be responsible for investigation and enforcement.

FDA & suicide

I haven't decided whether it's good or bad (I lean to the latter) that the FDA recently announced "findings" in studies involving antiepileptic drugs an increased report of suicidal ideation.

An FDA analysis of suicidality reports from placebo-controlled studies of 11 antiepileptic drugs shows that patients taking these drugs have about twice the risk of suicidal thoughts and behaviors (0.43 percent), compared with patients receiving placebo (0.22 percent). This risk corresponds to an estimated 2.1 per 1,000 more patients in the drug treatment groups who experienced suicidality than in the placebo groups.

Keep in mind, these are not big numbers. But when it comes to risk ratios, for researchers, it's the comparison that matters (0.43 vs. 0.22). It translates to twice the risk!

Scary! Fear! Boogeyman!

It would not be surprising--given FDA is inherently political despite their best intentions--that soon research studies involving any antiepileptics will require procedures for addressing situations in which subjects exhibit suicidal thoughts. Again, I'm not convinced this is bad necessarily, but it is yet another issue researchers will need to be aware of when preparing their research applications and another item to be added to the consent form.

As far as suicide procedures, this is what IRBs like to see:

- A qualified individual should evaluate all subjects whose responses indicate that they may be suicidal.

- Potentially suicidal subjects should be referred for appropriate clinical care. Please outline your procedures for ensuring that all potentially suicidal subjects are provided with referrals for appropriate clinical care.

- If a subject is acutely suicidal, they should be transported to the emergency room by trained personnel and a trained member of the study staff should remain with the subject until they have been transferred.

And then, in the consent form, languages something like this:

In the event that you tell the research staff that you are thinking about killing yourself or you answer “yes” to a question about having thoughts about suicide, the investigator will ask you more questions about the thoughts. Depending on how intense your thoughts are or how much you feel like hurting yourself, the research staff may provide you with referrals for treatment, work with you to contact your personal physician, trusted family member, or therapist to discuss your thoughts of harming yourself; or work with you on a plan that may include getting you to a hospital for safety.

Don't hate me me for this. If and when the time comes, complain to the FDA.

Shades of Gray

The New England Journal of Medicine recently published an article of an OHRP determination and apparent backtrack with a mea culpa.

The article is slim on details, however, it appears a research protocol was being followed involving hand-washing using different sanitizing techniques and placement location in the insertion of central venous catheters to determine whether a protective effect exists based on the technique and/or placement strategy used.

Yet the authors suggest this study could have, and should have, been expedited under Category 4 and Category 5.

Expedited review can occur only when specific criteria are met. Category 4 states:

Collection of data through noninvasive procedures (not involving general anesthesia or sedation) routinely employed in clinical practice, excluding procedures involving x-rays or microwaves. Where medical devices are employed, they must be cleared/approved for marketing. (Studies intended to evaluate the safety and effectiveness of the medical device are not generally eligible for expedited review, including studies of cleared medical devices for new indications.)

Examples: (a) physical sensors that are applied either to the surface of the body or at a distance and do not involve input of significant amounts of energy into the subject or an invasion of the subject=s privacy; (b) weighing or testing sensory acuity; (c) magnetic resonance imaging; (d) electrocardiography, electroencephalography, thermography, detection of naturally occurring radioactivity, electroretinography, ultrasound, diagnostic infrared imaging, doppler blood flow, and echocardiography; (e) moderate exercise, muscular strength testing, body composition assessment, and flexibility testing where appropriate given the age, weight, and health of the individual.

Meanwhile, Category 5 reads:

Research involving materials (data, documents, records, or specimens) that have been collected, or will be collected solely for nonresearch purposes (such as medical treatment or diagnosis). (NOTE: Some research in this category may be exempt from the HHS regulations for the protection of human subjects. 45 CFR 46.101(b)(4). This listing refers only to research that is not exempt.)

Yet the protocol in question involving the placement of a catheter, appears to disqualify this study from Category 4 as a catheter placement is inherently invasive. In addition, the protocol also appears to vary placement of the catheter for comparison purposes in the hopes of acquiring generalizable conclusions (e.g., it is safer/better minimize the placement of the catheter at the femoral site).

45 CFR 46.102 defines research as:

Research means a systematic investigation, including research development, testing and evaluation, designed to develop or contribute to generalizable knowledge. Activities which meet this definition constitute research for purposes of this policy, whether or not they are conducted or supported under a program which is considered research for other purposes. For example, some demonstration and service programs may include research activities.

My sense is a reasonable person would concur this is research.

Therefore, if this is research, invasive, and data in part is collected for research purposes, it seems clear such a study would not qualify for expedited review.

How can so many people, including the federal body regulating IRBs come to so may different conclusions? One, someone at OHRP does not know what he is doing. Two, Johns Hopkins IRB/staff do not know what they're doing. Three, I don't know what I'm doing. Or four, in a world of "black and white" regulations, we live in an IRB world of frustrating grays.

Change is in the air

We have lost another senior staff person this week.

Such is the nature of IRBs around the country, actually.

Turnover comes is waves and it appears unavoidable. One person leaves and others follow. The effect is significant change--some good and some bad. With the changes in staff, institutional knowledge is unquestionably lost. But also with change comes new perspectives and new energy.

My hope is that the new energy can somehow make up for the significant loss when a co-worker and friend sets sail to another opportunity.

IRB Tip #2: HIPAA

Old school medical researchers (and social-behavioral, for that matter) find HIPAA the bane of their collective existences. Codified into federal law in 1996, HIPAA established national standards for health care insurance portability and patient privacy rights. From the researcher point of view, she could no longer simply open a medical record and review it for research purposes without written permission from the patient-subject or a waiver by a privacy board / IRB.

Patient advocates rejoiced. And despite being on the research side at the time, I celebrated too. How dare I, you ask? What self-respecting researcher likes HIPAA?

Frankly, I'm a big believer of individual rights. And this is a clear-as-day issue of individual rights.

Unfortunately, then states' rights ruled the day when it came to patient privacy. These rights were not consistent from state to state and depending on where you lived, someone could take a peak in your medical record without your permission. HIPAA ensures patient-subject rights to some degree. IRBs are often the entities which review requests to open medical records in a research setting.

Regardless of your feelings about HIPAA, I bet we'll agree on something HIPAA related: our HIPAA Research Application (HRA) is terrible. It is perhaps the worst form we have next to the Internal Adverse Event report form. More on that another time. Back to the HRA.

First the 4 page application is backwards. The first determination investigators must make is, "When will I look at the medical record?" If the answer is to enroll subjects, without first obtaining their permission, you will want to ask for a waiver of HIPAA and informed consent for research purposes. Respond to Section B (on Page 2 of the application), Section II.B (on page 1), Section I.A, I.B, I.C (also on page 1). Asking for a waiver of HIPAA will also lead you to complete section B.1-B.4 as appropriate.

If you will be reviewing the medical record only after obtaining signed informed consent, complete Section I.A, I.B, I.C (on page 1) and Section A (on page 2) by writing, "I assure the Board I will use the UCLA HIPAA authorization form and IRB approved informed consent form."

If at any time you add research information to the medical record, more often than not you must get HIPAA authorization and obtained signed informed consent to do so. This means complete Section I.A, I.B, I.C (on page 1) and Section A (on page 2) by writing, "I assure the Board I will use the UCLA HIPAA authorization form and IRB approved informed consent form."

10,000 emails

That's how many I have in my email inbox after one year in this office. Keep in mind, I delete items from my inbox regularly.

So now it's come to haunt me and I'm sorting and cleaning my inbox.

This is what I do in my "spare" time. Sad, I know. But over the last 3 days, I'm down to a mere 8,800. It's a party.

IRB Tip #1

One of the reasons for starting this blog was to help investigators and research assistants and coordinators in an unofficial way. Don't get me wrong, I think I'm a pretty helpful guy in real life, but there are limits to what is proper in an official setting.

However, this blog gives me a good opportunity to expose the "secrets" of IRB functionality. And they really aren't secrets at all. But my sense is that IRBs generally do a poor job of educating, and researchers do a poor job of passing on information from staff to staff due to high turnover. Let's be honest, some (many?) PIs know little to nothing about an IRB submission and they can't teach their staffs what they don't know.

Although the idea here is for most tips to be specific to the IRB I know best, other IRBs follow similar policies. These policies, after all, have basis in federal regulations. Still, it's remarkable even to me--a "professional"--how varying IRBs work from institution from institution. Nevertheless, I will try to provide helpful tips to readers of this blog as best as I am able. Contact your own IRB for specific consideration. And feel free to let me know if their handling/processing is different from what I've described.

And with that, my first IRB Tip of GTMR:

IRB Tip #1
When your IRB takes forever to review and approve a new study, and you just want "administrative approval," you can get it pretty easily but only if you haven't submitted the full IRB application already.

Federal regulations (45 CFR 46.118) permit approval of a study prior to review by the IRB when "lacking definitive plans for involvement of human subjects."

This is a particularly good option when you can't access your grant money until IRB approval is obtained or if you're applying for a Certificate of Confidentiality from the NIH. Sometimes the NIH accepts it, other times they don't. But when it takes 3 months to get a CoC, it's worth trying to apply early, isn't it?

Frankly, I think many investigators should request administrative approval under 45 CFR 46.118, especially if your IRB is notorious for being slow or "picky."

Here's how you do it: In a simple letter, explain that you've been awarded X grant and wish to prepare the IRB application but cannot until you get IRB approval to release those funds. Provide your assurance that you will not initiate research involving human subjects until you obtain prospective IRB approval. State the title of the grant, the title of the study, and include a copy of the grant application (the protocol is sufficient). Complete the first 2 pages of the IRB application. Turn it in. That's it. Really. IRB approval is on its way. You can thank me another time.