Transcript Powerpoint file

BETTER APPROACHES
TO MONITORING
MEDICAL PRODUCT
SAFETY ARE NEEDED
1
EVALUATING THE SAFETY OF
MARKETED MEDICAL PRODUCTS
IS AN “ORPHAN” RESEARCH
AREA
• FDA has responsibility, but
– internal resources have been primarily
directed to evaluation of new products
– has minimal funds for extramural research
• Pharmaceutical industry has
responsibility, but
– internal resources are primarily directed to
evaluation of new products
– “Best practices” not established; FDA has
little basis to mandate much besides
reporting and case follow-up
2
ROLE OF ACADEMIA
• Research of academic scientists primarily
supported by grants
• NIH can stimulate research in certain
areas by requesting targeted applications
• Development of statistical methodology in
medical research has focused on areas
of greatest interest to NIH: clinical
trials, epidemiological research
• Little or no grant support for methods in
post-marketing safety surveillance
3
EVALUATING PRODUCT
SAFETY IS OF INCREASING
IMPORTANCE
• Availability of new products is rapidly
increasing; many treat previously
untreatable chronic conditions
• More people are using medical products,
and multiple products than ever before
– aging population
– improved therapies (non-sedating
antihistamines, inhaled steroids, SSRIs)
– new therapies for previously untreatable
conditions (hair loss, erectile dysfunction,
PMS)
4
RECENT STUDIES HAVE
RAISED PUBLIC
CONCERNS
• Deaths due to adverse drug
effects
• Medical errors
5
CURRENT PRACTICE:
Spontaneous Reporting
• Passive surveillance
– physicians or consumers voluntarily
report adverse effect to product
manufacturer or FDA directly
• All reports end up in FDA data base
• Huge numbers per year
– 250K: drugs and therapeutic biologics
– 100K: medical devices
– 13K: vaccines
6
CURRENT PRACTICE:
Active Surveillance
• In some cases, manufacturer does
“phase 4” study after product is
marketed
• An agreed-on number of patients is
treated with new product, with
active follow up by manufacturer
• Annual reports to FDA; rapid
reporting of serious events that
were previously unobserved
7
CURRENT PRACTICE:
RETROSPECTIVE STUDY
• Identify cases of interest in available
data base
– Health maintenance organization
– Medicaid
– National Death Index
• Define and identify a reasonable control
group from same or other data source
• Study association of exposure and
outcome using case-control methodology
8
WHAT TO DO WITH
SPONTANEOUS REPORTS
• Several simple methods have been
proposed over the years
• All have important limitations; none
is considered “standard” practice
• Recent development of Bayesian
data-mining approaches shows
promise for signal identification
9
BIG QUESTIONS
• Is there more that could be done?
• Are there other, as yet untapped,
sources of data that would be
helpful?
• How can we promote more research
in this area?
10