Fast forward to post marketing studies

A few months ago, the FDA happily reported that most post marketing studies are conducted/completed within reasonable time frames.  Move ahead to today when the Government Accountability Office made public its results on the FDA's history with using data from post-marketing studies to actively review a product's effectiveness over time.  So what did the GAO's study show: a kind of nothing for the FDA.

Apparently the FDA has not ever pulled a drug from the market because of a lack of effectiveness information provided by the manufacturer.  The article says that even when the sponsor provides the post-marketing data up 10 years late, the FDA takes no action to punish the sponsor for the delay in data reporting.  Even worse, the FDA has also never removed a drug's marketing approval when real data that prolongs the life of terminal patients does not exist.  Some time ago the FDA instituted a policy where new drugs could be approved based on lab based data, not actual patient survival times.  This was implemented to speed the drug development process and get more novel therapies to market for life-threatening conditions.  Where did this go wrong?

The idea is good, I agree that for some life-threatening illnesses the standard timeline for new treatments is simply too long and involved to be of benefit to patients.  Along the same lines manufacturers are less likely to pour millions of dollars into a development program that may or may not work without some help from the government authorities.  The problem, however, lies with the FDA's enforcement of its own policies.  There is no excuse for granting permission to market a drug and letting the data sit unattended for years.  The FDA should be following up with developers on a regular basis for status checks on post-marketing studies or even encouraging more comparativeness trials to be conducted to better identify the best treatments for patients.  To me, the evaluation of drug absolutely should not end at the approval of a marketing application (and while I'm sure the FDA agrees with this philosophy, you wouldn't necessarily know this looking at the report published by the GAO).

How can this be rectified?  Well for starters the FDA needs to put some teeth in its requests for post-marketing data.  If a sponsor does not get data to the FDA in a timely manner, hit them with a fine, a warning letter, or even better a temporary marketing suspension.  You could say that the FDA would be depriving patients of needed medications, but it would really be on the sponsor to ensure that the required post-marketing safety/effectiveness data is reported expeditiously so the approval remains valid.  As the top pharmaceutical regulatory authority in the United States, you would think the FDA would have a better handle on all types of data being developed/presented and not have the sort of almost lackluster enforcement of post-marketing trials actually seems to exist.

I'll post the actual GAO report when it becomes available, but until then here's a brief article on the findings

Time is of the essence

How long should it take for a coordinator to complete a study visit?  A loaded question for sure, because the answer depends on the number of tasks and procedures to be completed. For example a screening visit may easily take 2+ hours because of all the initial paperwork, while a follow vaccine visit may be only 20-30 minutes because it only may require a blood draw and brief physical exam.  A significant issue among managers and coordinators is how to maximize the coordinator's time so that down time is maximized to ensure that as many participants can be seen as possible during the day.

For some coordinators, study visits will be conducted one step at a time with no overlap.  This is not the right way to take advantage of the way a clinical trial visit is set up.  Coordinators should be aware of the procedures required at each visit and set up the visit to allow for the most of use a minimal amount of time.  Clinical research study visits often have significant down time built into the protocols.  For example subjects are often required to be resting for at least 5-10 minutes prior to taking vital signs.  Instead of just leaving the patient by his/herself, a coordinator should be copying a consent, reviewing medical records, preparing a lab kit, or entering demographic info in an ECG machine.  This will allow for productive use of that 5-10 minutes  because as small a period of time as it is, a coordinator's day is busy and 5-10 minutes can go a long way.  Here's a few more examples of utilizing down time appropriately:

-Reviewing a new participant's intake form while consent is being read, so any questions/concerns can be identified ahead of time
-Filling out source headers while a subject is completing questionnaires
-Taking temperature/scheduling an appointment while the subject is applying pressure to the site of a blood draw
-Presenting the participant's information/demographics to an investigator (prior to a physical) while questionnaires are being completed or a urine sample is being given

Proactive utilization of time of a critical aspect of a coordinator's job and learning how to do this is part of the maturation of an inexperienced coordinator into an efficient, vital member of a clinical research team.

Information please.

I don't think enough information is provided to sites regarding the proper collection and processing of laboratory samples in clinical trials.  For those of you not familiar with the central laboratory process, let me give a brief overview: 

1) site receives bulk supplies, lab kits, lab manual, etc at start up

2) site is supposed to review lab manual for guidance on sample collection/processing

3) material is reviewed with site by CRA at site initiation

4) patient visits start

Unfortunately, that is the extent of most study specific lab education.  For some of you, this may seem sufficient because you have processed thousands of labs and it is like second nature, but can you tell me why samples must be stored at -20 degrees Celsius for 4 hours, then moving to -70 degrees C for long term storage?  Or why you shouldn't leave a tourniquet on too long when drawing blood?   Or how about why you shouldn't let urine samples and blood samples sit unprocessed for longer than an hour? If you can answer those questions, good for you, but I would wager that the majority of commercial sites only follow the rules without an understanding of why they are written as they are.

I believe that sponsors would receive much higher quality lab data (which is crucial, I might add, to almost every study in the form of either safety or pharmacokinetic data) if sites were completely educated on not only how to process lab speciments, but also why.  Knowing that leaving samples unprocessed for too long can give erroneously low glucose values for a metabolic panel or even bacteria growth in a urine sample, might encourage a site to adhere more closely to lab manual guidelines (It's the real world, and we all are guilty of letting samples sit for 5 hours in the lab before processing).  The same goes for processing PK samples.  If sponsors provided reasoning behind the lab manual's guidelines, sites would be much more likely to follow those rules instead of occasionally bending them (and potentially compromising data) because it doesn't seem to make any difference to a coordinator who processes tons of lab samples day in and day out.

The major drawback to this that I could see is a lot of repetition in education of proper processing, but just like GCP I think that too much education is far better than too little.

Learning from others

I took a little time today to review some more older warning letters sent to clinical investigators following FDA inspections.  Here's one I came up with that I thought could be worth discussing.

There are numerous violations identified by the FDA inspector in this warning letter, but the two observations I wanted to talk about were:

1) A discrepancy between sources and CRFs.  This goes to show that you cannot take FDA inspections lightly.  There were several situations identified by the inspector where source documentation at the site did not match what was reported in the Case Report Forms.  Yes a monitor should have caught this prior to the inspection, but the site is ultimately responsible for the data.  Furthermore a site may argue that mistakes happen and the data may have gotten somehow transposed, but even small differences between the source and CRF could have significant implications for the data in a trial.  Morale of the story: take your time and be diligent when transcribing data from source to CRF.

  P.S. The FDA inspector also noted that corrections were made after a subject completed a trial without explanation.  It does look suspicious and I can completely agree with the inspector on questioning the legitimacy of the data in that situation.

2) Not properly consenting subjects.  Their were several violations of this genre reported, but one that I would like to point out is as follows: 

 "In a letter dated March 1, 2006, the IRB informed your site that they had approved informed consent form version 021706 which included modifications that were made to the risks section. The IRB’s letter further stated that current and new subjects were to sign the revised informed consent form. There was no documentation found during the FDA inspection to show that your site re-consented Subject #0004, who was still enrolled into the study at the time, with the revised informed consent form to inform them of the possible additional risks the subject may experience."


   This says that the site received notification from the IRB of a amended informed consent and that all currently enrolled subjects (as well as new ones) should sign the amended consent to inform them of newly identified risks of study participation.  For some reason this re-consenting did not happen (whether it was accidental or not is not the point here), and the investigator is the one who takes the heat because the participant was never made aware of new risks that could have affected his/her willingness to continue participating in the trial.  Again, mistakes do happen but the FDA will not take them lightly if the safety of subjects are at stake (if at all).  Morale of this story: take whatever measures are necessary to ensure that the recommendations of the IRB are followed with regards to the consent of subjects participating in clinical trials.




  Notice I did not say that the sponsor has power over the administration of the consent....more on that later...
   
   

Rescue Me

How many of you have heard sponsors say that accountability of rescue medication is not required because the study endpoints, data, etc are not concerned with that information?  If you haven't consider yourself lucky because I believe that anybody who tells you that is not correct and is potentially hanging a site out to dry with that advice.

Rescue medication, if dispensed by the site pharmacy, is technically investigational  product and subject to the same rules and accountability logs as the actual study medication.  For example, if a sponsor for a GERD study provides Gelusil or Tums as the rescue medication to be dispensed by the site, then the site is responsible for ensuring that subjects are dosing properly and not causing themselves harm by overusing the medication.  So what's the big deal, you ask?

Well, these medications/products are being dispensed as part of a government regulated clinical research trial and the subject is receiving the medication free of charge from the site.  Basically this means that the product should be considered and evaluated under the same umbrella as the actual investigational product being examined in the trial.  Taking 50 Tums that are expired and have been stored on a hot windowsill at the coordinator's desk with no regards to temperature monitoring could be as risky for trial participants as unmonitored study drug.  Okay, so maybe that was an over-exaggeration, but the bottom line is that any medication provided by the site for ANY use in an clinical trial is regarded as investigational product, no matter what your sponsor says.

So, to make a long story short and to potentially avoid ssues down the line, (I'm sorry if this leads to more work) consider the following:

ALWAYS keep a master accountability log of ANY medication leaving/returning to the site, even if the sponsor says you do not have to do so.  Thank me later if you get audited.

Monitor compliance with the rescue medication at each visit, even if not required by the protocol.  Diligence in this regard is important even if it takes up more of your time

Thoroughly educate the subject on the use of the rescue medication, even if it is something with which they are familiar.  Think Acetaminophen for example: most people know you can get extra strength over the counter, but how many are aware of the maximum dosing guidelines (4 grams by the way, at this time but subject to change based on FDA recommendations)?

These suggestions will serve the site's purposes better if they are followed, and sponsors can learn from it too because if they are providing sites with rescue medication, then they should very well be treating it just like an actual study medication...no ifs, ands, or buts.

Now Hear This

As discussed at length on EyeonFDA blog, there will be a public hearing on the use of social media and its use in the pharma/biotech industry.  This FDA hearing will take place on November 12 and 13, and really has been a long time coming.  Social media has been a sort of unregulated medium for all sorts of things in the pharma world, from advertising to clinical trial recruitment to DTC (direct to consumer) publications.  It seems that the FDA has been slow to react to the viral potential of social media, issuing a warning letter here and there when it comes across a piece of marketing that it finds obtrusive or provides nebulous information to consumers about a product.  Just like with other forms of marketing, unregulated social media has the potential to take advantage of an uneducated public, which is especially dangerous when speaking of pharmaceuticals  There should be some sort of formal documentation from regulatory authorities about how social media can and should be used by the pharma/biotech industry for various purposes.

I don't think that the use of social media by pharma should be banned because Twitter, Facebook, and blogging can be very effective ways of properly informing the public about a topic/product or even inviting public opinion in an informal fashion.  There are lots of companies that use the above mentioned social media platforms to keep consumers up to date with the company and its products.  However pharma has a bad reputation of using media in less than completely ethical fashions to benefit itself, so we cannot rely on the industry to regulate itself.  Hopefully this hearing next month will elicit some good, solid recommendations (maybe even a formal guidance document) from the FDA on what can and can't be done with social media because like it or not, social media is here to stay.

For some light reading, check this link to the official FDA release

Medical Records?

There are quite a few opinions out there on whether or not clinical research sites should request medical records for research participants that are not patients of the research facility.  Some people argue that without medical records there is no way to complete assess an individual's eligibility for a trial and there are absolutely no exceptions to this.  On the other hand some sites may argue that medical records are not necessary unless specific documentation about a condition (ie arthritis, cancer, MS) is required.  These sites may say that most records contain lots of extraneous information that is not clinically relevant to the trial, but must be documented in such a way that it greatly waters down the quality of the data from the site.  Not to mention that it leaves plenty for an overzealous monitor to scrutinize for possible study entry violations. 

So who is right?  The FDA does not explicitly require that copies of medical records be obtained for clinical trial participants, but good clinical practice would strongly advise that attempts to request records are made prior to enrolling a subject in a trial.  The federal regulations do mention that a clinical investigator needs to collect a complete and accurate case history of a subject and that this history does include past/current medical conditions, but those rules do not indicate how this history should be acquired.  In my opinion a site is more than welcome to use a medical history interview in the clinic to acquire a case history, but there should absolutely be some sort of documentation signed off by an investigator that the interview was deemed to be accurate/complete if this practice is utilized.  However I think that if copies of medical records can be obtained, that is always preferable to ensure total collection of a case history and to protect the site in the event of an audit (even if it does open doors to much more work by site staff in recording extra medical conditions, having an investigator review those records, and fighting with monitors over questionable findings).    

I guess the take home message here is that while the clinical investigator is ultimately responsible for the eligibility of research participants, there is no substitute for outside, objective information that verifies a subject's case history.  Most clinical research staff (investigator's included) are truly trying to do a good job with honest data collection and documentation, but there are those out there who will leave out exclusionary conditions or medications just to see a person enrolled in a trial (remember, this is one reason that audits occur). With all the discussion lately about the FDA tightening the reins on clinical research, it's more important than ever for research sites to ensure that GCP is followed tightly and medical records are an excellent way to prove this.

Noted.

I find that one significant gray area in clinical research is the use of the Note to File.  There are numerous situations that may or may not require the use of an explanatory Note to File, depending on who and when you ask.  Generally the Note to File should be reserved for situations where something has occurred over the course of the study that cannot be explained sufficiently using the documentation already present in the chart (or file/crf/etc...).  The point of the Note to File is to provided clarification so that an auditor or outside reviewer can understand a situation, see what was done to correct it, and move onward with few or no questions.  The problems with Notes to File, however, that I see are twofold:

1) Notes to File (and one of my pet peeves is incorrect pronunciation of the plural...it's Notes to File, not Not to Files.  Welcome to middle school grammar) are often written poorly with too little or too much information. 

2) Charts will have either too many or too few Notes to File to explain situations in the trial.

 Notes to File should ideally be concise, well written grammatically, and informative.  They should not be filled with flowery language and should not go into great detail explaining something that an auditor would already know.  For example auditors should understand the protocol, so one should not describe protocol procedures related to a note to file in great detail because the note will become too wordy and likely confusing to the reader.  Ideally, a Note to File will identify what the problem is, how it was fixed, and what was done to prevent it from occurring in the future.  Anything that goes outside of this is a furbelow.


Secondly one should exercise good judgement when creating Notes to File.  Too many of them will raise a red flag to an auditor that there were many mistakes in a study, too few will not adequately address mistakes or issues with the data/study files/etc. A good rule of thumb is to write a Note to File when a situation arises that cannot be explained by a progress note (ex. regulatory binder issue) or when a mistake occurs that has implications for the entire study (ex. consent error or dosing error) if not corrected. 

It is confusing for inexperienced coordinators to decide what merits a Note to File, but it is important to learn so make sure you ask an experienced staff member if you have questions.  Remember  that even though a Note to File is useful for explaining sticky situations,  it is also an indication of a discrepancy or mistake in a trial and should not be used liberally.