Friday, January 20, 2006

Drug industry officials see room to improve safety - or do they?

I've written about the substantial inadequacies of the postmarketing drug safety surveillance "system" here , here and here (note: the word "system" in this context is a bit of an oxymoron).

The Gartner Group utilized some of my opinions in formulating their report "Predicts 2006: Life Science Manufacturers Adapt to Industry Transitions" (unfortunately available only to subscribers). I'd been interviewed in the report's preparation after a Gartner principal had seen my writings on the subject.

An article has now appeared at Reuters stating that the system for monitoring drug side effects only needs "focused treatment" (?). It does not need "major surgery", only additional funding of the FDA bureaucracy and 'better ways to collect data.' The latter on it face seems to mandate exploratory surgery due to problems such as this.

Excerpts from the article:
Drug industry officials see room to improve safety
Thu Jan 19, 2006 04:06 PM ET
By Lisa Richwine

WASHINGTON, Jan 19 (Reuters) - The U.S. government's monitoring of drug side effects could be strengthened but does not need the major changes critics have advocated, industry officials told an expert panel on Thursday.

Additional funding for the Food and Drug Administration and better ways to collect and analyze safety information could help improve detection of unexpected problems after drugs reach the market, the officials said.

"Drug safety in the U.S. definitely needs improvement. It needs a careful, focused treatment, but I would argue not radical exploratory surgery," said Geoffrey Levitt, chief counsel for regulatory and research at Wyeth (WYE.N: Quote, Profile, Research) .

A string of serious side effects linked to widely used prescription drugs has led to calls for major changes, such as moving safety oversight to an independent board outside the FDA.

Industry officials who spoke to an Institute of Medicine (IOM) panel said they felt that that proposal, as well as suggestions that companies re-apply for approval five to 10 years after a drug launch, were unnecessary and could have negative consequences.

"We think that moving oversight of safety to another federal agency seems extreme and could significantly stall incremental progress," [translation - stall profits? -ed.] said James Nickas, senior director of development at Genentech Inc (DNA.N: Quote, Profile, Research) .

The panel will consider the input as it prepares a report on drug safety oversight that is scheduled to be completed this summer. The FDA requested the investigation last year after the agency was criticized as being slow to respond to signs of problems with antidepressants and Merck & Co. Inc.'s (MRK.N: Quote, Profile, Research) withdrawn painkiller Vioxx ...

"Exploratory surgery" is a metaphor that implies a major problem exists but is not yet identified nor understood, and exposure of the internals are required to establish a diagnosis and render treatment. My belief is that is exactly what's needed to bring the paper-based, 1970's approach to postmarketing drug safety surveillance into the 21st century.

One major improvement would be for pharmas to become more aware of the field of Medical Informatics, whose cross-disciplinary experts are working to move the U.S. towards national electronic health records. Such a resource is perhaps the best (if not only) way to move drug surveillance from the hit-or-miss systems currently in place (e.g., FDA MedWatch and use of payor billing information) to a concurrent, comprehensive system of data flow on this problem.

Yet I still see job postings such as this from my former employer, Merck (now struggling to regain the public's confidence on clinical data issues). I sat on a "talent management" team there whose role was to suggest improvements in talent-seeking. On a number of occasions I advocated for medical informatics experts cross-trained in both medicine and biomedical information science for clinical data management and IT-related leadership roles. However, this type of posting is common throughout all pharmas:

Worldwide Clinical Data Management Operations Manager - Global Data Operations

The Manager of WCDMO (Worldwide Clinical Data Management Operations) is responsible for providing all aspects of supervision to individuals supporting a functional area within WCDMO ... This position is in the Global Data Operations area.

Education: BA, BS or MS degree, preferably in Biology, Nursing, Computer Science, or related discipline. Experience: At least 8 years' experience in data management, clinical research, and/or database design and development. At least 3 years of supervisory and/or management experience. Six Sigma and PMI certification is a plus.

Knowledge and Skills: An overall working knowledge of the clinical development process. Knowledge of database structures and available tools to manage, extract, and report data. A basic knowledge of statistical concepts.

Baccalaureate or masters degrees (but not doctoral or postdoctoral credentials), six-sigma ("management metaphysics") and a "working knowledge of clinical development" are nice credentials, but represent basically the same credentials that might have been sought in, say, the 1970's.

I also note the background of the Wyeth chief counsel for regulatory and research who argues for "focused treatment" of the "system" (which reflects a 'let's not rock the boat too much' philosophy):

Geoffrey Levitt
Vice President and Chief Counsel, WYETH

Mr. Levitt is Vice President & Chief Counsel, Regulatory and Research at Wyeth Pharmaceuticals, Inc., where he is responsible for a wide range of legal/regulatory issues related to FDA compliance and clinical research. Prior to joining Wyeth in April 2001, Mr. Levitt was a partner and co-chair of the food and drug law group at Venable, Baetjer, Howard & Civiletti in Washington, DC. He has written and lectured extensively on food and drug law, and is a past member of the editorial board of the Food and Drug Law Journal.

Impressive legal credentials, but no apparent credentials in the medical and scientific domains, nor in biomedical information science.

As has been mentioned in this blog in the past and at sites such as this, when non-medical personnel decide they know best what's needed to cure healthcare's ills, the results are not often optimal.

I would add that a lawyer employed by industry would likely have significant, inherent conflicts of interest regarding matters of such import where impartial clarity is essential.

-- SS


Anonymous said...

As a non-medical person involved in Health Policy I object to your position. I think the bigger point you are trying to make is that too often decision makers in healthcare ignore, or are unaware of, the medical implications of their decisions. But on the flipside, all too often medical personnel ignore issues of societal agency that have, by default, fallen on their shoulders--while simultaneously opposing any reasoned discussion of who should take on societal agency roles.

InformaticsMD said...

> I think the bigger point you are trying to make is that too often decision makers in healthcare ignore, or are unaware of, the medical implications of their decisions.

In the field of biomedical information science, most certainly. For instance, putting desktop computers with big fans on the ceilings of small ICU rooms to become colonized and spread infections at a time of increasing drug resistance. It doesn;t get much worse than that (one of several examples at ).

I should point out that my posting focuses on the area of biomedical information science (medical informatics). I am not directly addressing other domains in healthcare in the post.

However, I could argue that if you as a non-medical person involved in Health Policy had a medical background complementing your primary field, your decision-making ability would be enhanced.

In Medical Informatics we take the time and effort to train in two specialties: clinical medicine, and clinical computing. This provides insights that people, in my experience, who have trained in one or the other (or in neither) simply do not often have.

As to stereotyping of clinicians, I addressed that some years ago (again, focused in the domain of healthcare informatics) at .

Clinicians are a varied bunch.

-- SS