Académique Documents
Professionnel Documents
Culture Documents
Quality Standards for Medicines, Supplements, and Food Ingredients throughout the World
Transdermal Systems
1
<724> Drug release
• April 1, 2005
2
<724> Drug release
• Apparatus 6 – Cylinder
3
<724> Drug release
4
<724> Drug release
Transdermal systems
5
TDS LEAK TEST
rationale for the need for a leak test in transdermal patches
Gordon L. Flynn
Professor Emeritus, University of
Michigan College of Pharmacy
An apology!
6
USP Mission Statement—Promoting the Public Health
“USP‐NF is published in continuing pursuit of the mission of the
United States Pharmacopeial Convention (USP), which is:
to promote the public health and benefit practitioners and
patients by disseminating authoritative standards and
information developed by volunteers for medicines , other health
care technologies, and related practices used to maintain and
improve health and promote optimal health care delivery.”
I hope to convince you that never in the past has the primary packaging of a
drug delivery system been such an problematic issue with respect to patient
safety than with the delivery of a low therapeutic index drug out of a form‐fill‐
seal transdermal delivery system; this is because leakage in the course of drug
delivery from such systems expands the area of delivery, increasing the rate of
delivery, and, to date, leakage has not been under adequate control.
NEVER BEFORE BASED ON THE USP‐NF of 2005:
<1> Injections — Packaging: validation must include verification that package
prevents microbial contamination (this a zero tolerance situation)
<381> Elastomeric Closures for Injections: … evaluation … should be made by
appropriate additional specific tests to determine the suitability of an elastomeric
closure for its intended use
<601> Aerosols, Nasal Sprays, etc. — Leakage Test: … weigh each container … allow
each container to stand in an upright position at a temperature of 25˚ for not less
than 3 days … again weighing each container
<661> Containers: — … many Pharmacopeial containers are of such nature … as to
require the greatest attention to the containers in which they are stored or even
maintained for short periods of time
<731> Loss on Drying: … the procedures set forth in this chapter determines the
amount of volatile matter of any kind that is driven off under the conditions
specified
<905> Uniformity of Dosage Forms — Content Uniformity: … transdermal testing by
content uniformity of individual units
<1146> Packaging Practice: as it relates to transdermal units as opposed to single
solid units
<1150> Pharmaceutical Stability: … as it relates to the integrity of transdermal units
<1151> Pharmaceutical Dosage Forms — Transdermal Systems: … different types of
systems not spelled out
7
Let me start our by saying that there are two fundamentally different types
of transdermal systems, form‐fill‐seal patches (liquid reservoir patches) and
matrix (monolithic) patches. Each comes in a range of sizes and each has a
range of phase configurations. The feature that most distinguishes these
two types of systems is that liquid reservoir patches (pouched patches)
contain the bulk of their drug in a walled‐off fluid, a relatively free‐flowing
medium, while matrix patches contain their drug in either the adhesive
layer, a thick, non‐flowing polymeric phase, or in a solid polymeric reservoir.
The two types of systems share some problems. They don’t adhere to the
skins of some individuals as firmly as they should; there is lifting of edges,
particularly when patches are worn multiple days. And all too often they
detach during showers, etc. To comparable extents, delivery from them
suffers from percutaneous absorption variability inherent in the treated
population.
However, in one important way they differ markedly. Form‐fill‐seal patches
are known to leak; matrix patches don’t share this problem.
drug
backing
layer
reservoir to illustrate
the critical
heat‐
sealed
difference
bioadhesive rim
internal
release polymer
liner layer
Form‐fill‐seal Patch
backing
layer
bioadhesive
release
liner
Matrix Patch
8
To understand the issue that
we are faced with today
during this session—an issue
that pertains to assuring that
individuals who are being
treated with transdermal
patches are being treated
with products that are safe to
use—a scientist has to come
to terms with the manners in
which patches are
manufactured and the
inherent system frailties and
weaknesses that their consider the
manner of manufacture form‐fill‐seal
creates in them. patch
9
The finished form‐fill‐seal patches are cut by machine from
the ‘web’, placed in sealed foil envelopes, and generally
packaged in boxes containing ≥ 5 foil‐sealed patches.
Quality must be designed in—difficult to do with form‐fill‐seal patches.
What problems are form‐fill‐seal patches are particularly prone to?
• Faulty heat seals during manufacturing
• inadequacy of production monitoring
• too high a production rate for adequate inspection
• lack of adequate post‐production leak testing (and method)
• Evaporation of volatile content in storage (resolved by packaging)
• during distribution/storage in warehouse and pharmacy
• while in patient’s hands under less well controlled conditions
• Rupture of faulty seals during patch patient use and wear
• lack of ruggedness due to climate (warmth, moisture, cold)
• pressure, particularly when sleeping
known form‐fill‐seal
patch defects
10
the incomparable danger of such defects requires our fullest attention
Let’s Be Honest! The major problem with form‐fill‐seal
(liquid reservoir) patches has been with the leakage of a
highly potent drug with a low therapeutic index out of the
liquid reservoir as the result of a faulty seal. The drug is
fentanyl.
Fentanyl is a drug with a low therapeutic index. As just one
‘proof’ of this, it has been learned ‘the hard way’ that
patches containing fentanyl can be lethal in otherwise use‐
conditioned patients if worn under a heating pad, heating
blanket, or in a hot tub. Reasonable activation energy
calculation indicates that such heating wouldn’t represent
so much as a doubling of the delivery rate. This danger is
spelled out in the drug’s prescribing warnings to physicians
and package insert warnings to patients.
11
Concerned about deaths associated with the proper use and also the abuse of
the then existing fentanyl form‐fill‐seal transdermal patch, four years ago I
submitted a citizen’s petition to the FDA.
The main emphasis of my stated concern related (relates) to the safety of
patients using fentanyl form‐fill‐seal dosage forms that were “legitimately
prescribed, properly dispensed, and properly applied.” In other words, I was
most concerned about situations in which no error of judgment was made by
the prescribing physician, in which the dispensing pharmacist correctly
interpreted the physician’s ‘order’ and dispensed the right dose (dosing rate)
patch, and in which the the patient applied the patch correctly.
My submission of the petition followed the first fentanyl patch recall due to
leakage of its drug‐containing gel. Fentanyl transdermal systems have been the
foremost source of gravely serious therapeutic mishaps. Yet let’s not loose sight
of the fact that, in transdermal form, fentanyl is a drug that allows patients
afflicted with severe chronic pain to live relatively normal lives. But when the
drug’s dose is increased by as little as a factor of two, it can produce a
respiratory depression so deep that a patient simply ceases to breathe.
In the citizen petition I specifically recommended that the FDA:
“a) not approve further liquid reservoir transdermal systems unless and
until the potential manufacturers of such systems provide convincing
evidence that the seals of the form‐fill‐seal patches they wish to
introduce are failsafe with regard to leakage,”
“b) review the manufacturing procedures and controls now in place for
the production of presently marketed liquid reservoir transdermal
systems to make sure that these are failsafe with regard to leakage,”
“c) review the overall safety of use of liquid reservoir patches now on
the market from standpoint of risk of harm to patients using such
patches appropriately and also from the standpoint of their relative
ease of drug abuse.”
I did this because in my view people’s lives were on the line. I’m here
before you today mainly as a result of filing this petition.
12
Before I speak to the situation of patch leakage, I’d like to point out that
there isn’t a situation that I know of that can’t be dealt with using a matrix
patch. Consider that:
1. Nitroglycerin, introduced as a form‐fill‐seal patch, can obviously be
bioequivalently delivered from a matrix system patch
2. Nicotine, introduced as a form‐fill‐seal patch, can be bioequivalently
delivered from a matrix system
3. Estradiol, introduced as a form‐fill‐seal patch, can be bioequivalently
delivered from a matrix system
4. Fentanyl, introduced as a form‐fill‐seal patch, can be bioequivalently
delivered from a matrix system
and, despite questionable claims that the above form‐fill‐seal systems
contain membranes that control delivery rates, the corresponding matrix
patches are just about the same size (have essentially the same delivery
area), have been proven bioequivalent, and yet don’t contain rate‐controlling
membranes.
The trick is to build matrix systems that remain saturated or near saturated in
terms of their drug over the greater fraction of time that they are worn.
Let’s consider the nature of the problem that we
do confront.
seeming defects
People ‘live in’
and ‘sleep in’
patches as the
one that’s
depicted to the
right (here you
see an internet
photo that is
touted as
representative
of a faulty
fentanyl patch)
13
disclosure and trial documents
published on the internet indicate
that the original 75 µg/hr fentanyl
patch has the following features
A particular problem with the original form‐fill‐seal fentanyl
patch, again from documents found on the internet, seems to
have been the formation of “unseal channels” due to “stringer
leakers” formed in the course of filling the patch pouches.
Apparently, a string of gel occasionally trailing the pouch fill
operation deposits along the seam that is momentarily to be
heat sealed. Consequently, the heat seal doesn’t ‘take’ at the
locations where gel is deposited. Extrusion of gel from the
weak or defective seals, particularly when placed under
pressure, is surely a problem.
We can consider what stresses a worn patch has to bear and
testing that might be performed, but keep in mind that quality
has to be designed into a patch—the production of a patch—
and can’t really be belatedly tested in.
14
Given the propensity form‐fill‐seal patches have to leak, how much
pressure must a patch withstand while in use? This is something that
doesn’t seem to have been adequately addressed. Consider that the
body surface area of a male who weighs 75 Kg (163 lb) and is 177.8 cm (5’
10”) tall is about 1.92 m2 (19,200 cm2)
Laying prone, the male would have something on the order of 40% of his
body in contact with surface he is lying on. Using this figure, he’d have
7,680 cm2 of his body in continuous contact with the surface (note that
with less contact the numbers below become even larger).
He would thus have his 75,000 gm weight distributed over 7,680 cm2 or,
on average, about 9.77 gm would be pressing down on each cm2 of
contact area. Since 1 pound = 453.6 gm, this amounts to 0.0215 lb/cm2
A 10 cm2 patch would experience, on average, something like 97.7 gm of
weight pressing up against it. This amounts to about 0.2153 lb. The
patch’s seal would have to safely withstand such pressure. Using a safety
factor of 10, the weight would be > 0.2 lb/cm2 or > 2 lb/10 cm2 patch, not
an inconsequential weight relative to forcing leakage.
As shown in this drawing that
speaks to emergency
accommodations in a school
gym in the course of a
hurricane or other natural
disaster, people sleep in a
variety of positions. Should
any of them be wearing
patches, they’d be rolling
around on them for hours at
a time. Of course, wearers
tend to place patches over
body sites where the patches
are somewhat ‘protected’
and where there is less
weight suffered by them, but
in the course of twisting and
turning during sleep, patches
still face considerable weight
stress.
15
POINT # 1: There is a need for rigorous unstressed
testing of patches during production and, with
statistically supportable sampling, further testing
on post batch product:
1.continuous visible inspection
2.inspection aided by microscopic examination
of samples
3.Identification of leakage by placing sheets of
patches between adsorbent sheets of paper
4.swab patch samples to identify leakage (there
should be no drug outside the primary
envelope.
POINT # 2: There is a need for a post‐
manufacture test of proof of seal strength
under pressure on every batch of form‐fill‐seal
product that is made. Considering that, with a
drug like fentanyl, their has to be zero
tolerance for weak and faulty seals, the test has
to be statistically rigorous with respect to:
1.the means of sampling for such testing
2.the number of samples subject to post‐
production pressure testing (always with
limited sampling of a batch)
3.Pressure‐stressed patches should also be
swabbed
16
But to build quality in there’s a need to improve manufacturing methods and
in process controls so that the integrity of every patch is assessed as it is
coming off the production line. What kind of production surveillance is
needed?
It’s been pointed out that stringer leaker defects, fold‐over defects, and cuts
have been noted in the course of patch production. In addition, notably,
according to records, form‐fill‐seal patches produced in the past were visually
inspected by production staff teams, with two members of the production team
always operating as inspectors for 30 min periods at a time. The inspectors had
to judge the integrity of patches moving through their fields of vision at a rate
of about 4 patches per second. In process instrumental measurements also
appear to have been made in the course of production, but if this was so, these
weren’t anywhere close to adequate relative to preventing the formation of
leakers and thus protecting the public.
Instrumental methods of inspection have to be developed that adequately
identify and sort out defective patches from well formed ones. The commercial
instrument below was designed to inspect nicotine patches; it is presented just
as an idea of what is needed. As far as I can tell, in process controls aren’t an
area of USP ‘involvement’, but perhaps should be.
I’m informed that automatic
machine inspection might be
possible. This would seemingly
be the only way to really
counteract the problem
associated with identifying
leakers. The problem here is
serious enough that a task
force, a consortium of
surveillance machine
manufacturers, pharmaceutical
company scientists, and
perhaps academic scientists,
might be formed to address
the problem.
17
Here again is the
problem. You tell
me whether or not
you think you could
find all production
flaws in patches
moving through your
view at four per
second.
POINT # 3: Given that a leaky form‐fill‐seal patch systems
can result in death, there has to be much better control of
quality than now exists, if not for all patches, then at least
for patches containing potent, low therapeutic index drugs:
1.there is no question that superior inspection
machinery has to be developed—non‐destructive image
analysis
2.There is no question that visual human inspection isn’t
adequate for the kinds of defects that are known to
exist, especially when patches are moving past them on
a belt at 4 patches per second
3.Together the above two methods have to be failsafe
(have to lead to zero tolerance) to the same degree as
sterile products have to be failsafe (with respect to
microbes)
18
This is the end of my thoughts. I can see the
problem, and it’s a big problem, but, admittedly, I’m
not an engineer who might be able to solve it. I have
no background in fractal geometry, let along any idea
of how this mathematical tool might be statistically
applied to successfully monitor production. I’m
informed machines and software can be developed
that might make fractal geometry perform the task,
but can do nothing more than suggest the possibility.
THE END
And thanks to Mike
Houghton for his thoughts
on these matters!
Adhesion and Other Product Quality
Tests (Rationale for Inclusion)
Kris Derdzinski, Ph.D.
Principal Scientist
Ortho‐McNeil‐Janssen Pharmaceutical
September 14‐15, 2009
19
Transdermal Delivery Systems
y Physical units that deliver active ingredient at a desired
rate over a prolonged period of time
y Mechanism of delivery may be a passive diffusion or a
different mechanism depending on the systems design
y common designs currently marketed are transdermal
patches:
y a form‐fill‐seal type (reservoir systems)
y multilayer laminates with drug substance contained within a
solid phase (matrix), including drug in adhesive patches
y peripheral adhesive patches
y systems that use other mechanism for delivery
USP Workshop on Topical & Transdermal Drugs
(i.e. iontophoresis, microneedles, heat‐assisted delivery)
Sep 14-15, 2009
Adhesion
USP Workshop on Topical & Transdermal Testing
Drugs Sep 14-15, 2009 Adhesion Testing
Adhesion is Critical to Quality
y Adhesives are used to assure an intimate contact with the
patient skin and delivery of the desired dose per label
y Adhesives in TDS products must
Î allow for an easy removal of release liner before use
Î adhere properly to human skin upon application and
maintain the adhesion during the prescribed time of use
Î allow for easy removal at the end of use without leaving a
residue or causing damage to the skin (adverse affects)
Î be able to maintain the acceptable performance through
the end of product shelf‐life
20
Adhesion Quality Problems
y A review of adhesion related quality problems was published by
FDA scientists [Eur J. Pharm Biopharm, 64 (2006) 1‐8]
y Based on review of complaints received for various TDS (TDDS)
products the problems included:
Î failure of adhesive (lack of adhesion) during routine use when
worn over several days
Î need to use tape to hold TDDS in place (prevent loss of patch)
Î increased cost , need to use more patches due to poor adhesion
Î loss of efficacy due to unreliable adhesive properties
Î report of adverse events – skin damage during patch removal
Î patch damage during removal of release liner
Î loss of patch during use (stray patches in linens, hair , food…)
USP Workshop on Topical & Transdermal Drugs
Sep 14-15, 2009
Adhesion
USP Workshop on Topical & Transdermal Testing
Drugs Sep 14-15, 2009 Adhesion Testing
Adhesion Tests
y Adhesion tests are typically used in product development
by manufacturers and as in‐house quality control tests
y Three types of adhesion tests provide information relevant
to functional performance of adhesive
Î Peel force (from standard substrate / release line
Î Probe tack test
Î Shear strength (creep compliance)
y Generally in vitro tests that predict adhesion to human
skin are not available
USP Workshop on Topical & Transdermal Drugs
Sep 14-15, 2009
Adhesion
USP Workshop on Topical & Transdermal Testing
Drugs Sep 14-15, 2009 Adhesion Testing
21
Rationale revisited
y Adhesion tests are critical for efficacy, safety and quality
y Choice of specific tests method and acceptance criteria
will depend on the product design and specific
formulation
y Large variance of adhesion results does not preclude
setting meaningful criteria to assure consistent products
quality from batch to batch and similarity to the adhesion
performance for the registration batches
y Manufacturers’ knowledge on adhesion testing methods is
available and validated tests are available
USP Workshop on Topical & Transdermal Drugs
Sep 14-15, 2009
Adhesion
USP Workshop on Topical & Transdermal Testing
Drugs Sep 14-15, 2009 Adhesion Testing
References
Transdermal drug delivery system (TDDS) adhesion as a critical safety,
efficacy and quality attribute, Wokovich, A.M., Prodduturi, S., Doub,
W.H., Hussain,A.S., Buhse, L.F., Eur J. Pharm Biopharm, 64 (2006) 1‐8]
Skin adhesives and skin adhesion 1. Transdermal drug delivery systems,
Venkatraman, S., Gale, R., Biomaterials 19 (1998) 1119‐1136
Pressure‐sensitive adhesives for transdermal drug delivery systems,
Tan, H.S., Pfister, W.R., PSTT, Vol 2, No 2, Feb 1999
22
Thank you
USP Perspectives
Quality Standards for Medicines, Supplements, and Food Ingredients throughout the World
23
1974 Drug Bioequivalence Panel
• Berliner • Standards: no BE
• Cluff • BA causes therapeutic failures
• Doluisio • Procedures available
• Melmon • BA critical for some drugs/not all
• Nadas • GMPs and compendial standards inadequate
• Oates • New standards needed
• Riegelman • Research needed, including in vitro/animal
• Shideman models
• Zelen • Strengthen current standards/submit studies
• Robbins • Eliminate exemptions
• New organization
24
What’s the Further History
• General Guidance
• Criteria Guidance
• Bioanalytical Guidance
• Biopharmaceutics Classification System
Guidance
• Food-Effects Guidance
• Nasal Drug Products
• Oral Inhalation Drug Products (delayed)
• Topically Applied Drug Products (withdrawn)
• Dissolution Testing of Immediate Release Solid
Oral Dosage Forms (final)
• Extended Release Oral Dosage Forms:
Development, Evaluation, and Application of In
Vitro/In Vivo Correlations (final)
25
WHO Interchangeability Document: Table of
Contents
• 1. Introduction
• 2. Definitions
• 3. Documentation of equivalence for marketing
authorization
• 4. When equivalence studies are not necessary
• 5. When in vivo equivalence studies are necessary
and types of studies required
• 6. Bioequivalence studies in humans
• 7. Pharmacodynamic studies
• 8. Clinical trials
• 9. In vitro testing
• 10. References
• Three Tables: A, B, C
• A Table Indicates:
– Highest recommended dosage form strength
– Solubility
– Permeability
– BCS Class
– Dissolution Test Requirements
– Risks
– Indications and Comments
• ~ 130 Molecules
• ~ 60- 70% Suitable for BCS Biowaivers
• Implications
– BE Part of Drug Development
– Company Does It
– No CRO
26
And Calibration?
• Merriam-Webster: calibrate
– To ascertain the caliber of (as a thermometer tube)
– To determine, rectify, or mark the graduations of (as a
thermometer tube)
– To standardize (as a measuring instrument) by
determining the deviation from a standard so as to
ascertain the proper correction factors
– To adjust precisely for a particular function
27
Current Lots and Applications
Blister Pack
28
Lot P Collaborative Ranges
1 65 47 - 82 60 +/- 5
2 53 37 – 70 40 +/- 3
USP Publications
• Deng G., et. al.: The USP Performance Verification Test Part I: USP
Lot P Prednisone Tablets – Quality Attributes and Experimental
Variables Contributing to Dissolution Variance, Pharm. Res. 2007
• Liddell M., et al.: Evaluation of Dissolution Glass Vessel Dimensions
and Irregularities, Dissolution Technologies, Vol. 14 Issue 1, Febr.
2007
• Liddell M., et al.: Dissolution Testing Variability: Effect of Using Vessels
from Different Commercial Sources, American Pharm. Review, Vol.
10(6) Sept./Oct. 2007
• Nithyanandan P., et al.: Evaluation of the Sensitivity of USP
Prednisone Tablets to Dissolved Gas in Dissolution Medium using
Apparatus 2, Dissolution Technologies, Vol. 13 Issue 3, Aug. 2006
• Eaton J., et. al.: Perturbations Study of Dissolution Apparatus
Variables – A Design of Experiment Approach, Dissolution
Technologies, Vol. 14 Issue 1, Feb. 2007
29
USP Publications
• Hauck WW, Manning RG, Cecil TL, Brown WE, Williams RL. Proposed
change to acceptance criteria for dissolution performance verification
testing. Pharm Forum. 2007;33(3):574–579. Reprinted in: Dissolution
Technol. 2007;14(3): 8–12
• Hauck WW, DeStefano AJ, Brown WE, et al. Description of the
upcoming change in acceptance criteria for USP dissolution
performance verification tests. Pharm Forum. 2008;34(6):1630–1629.
Reprinted in: Dissolution Technol. 2009;16(1): 7–12
• Hauck WW, DeStefano AJ, Brown WE, et al. Change in criteria for
USP dissolution performance verification tests. AAPS PharSciTech.
2009;10(1): 21–26
• Weight 1% CV
• Hardness 6% CV
• Assay <1% CV
• Content Uniformity 2% CV
• Disintegrant Content 3% CV
• Disintegration Times ~10% CV*
• Dissolution Percentage <4% CV**
30
Intra-Lab Variances Smaller than Inter-Lab
80
%Dissolved at
30 minutes
75
Lot P
Apparatus 1
Sets of 6
70
Standard
errors
65
Approved
range
47 -- 82
60
55
One Lab
below range
50
in this chart
0 50 100 150 200 250 300
B l i nde d R un C ode
70
50
40
30
0 50 100 150 200 250 300
B l i nde d R un C ode
31
ISO Framework
Metrology—Traceability
A Std
TS YA = gA (X0)
CRM
X0 B Std TS YB = gB (X0)
32
National Traceability Chain
Calibration Labs
Katal
• katal is now adopted under the Convention of the Metre of 1875 by the highest
metrologic authority, the intergovernmental diplomatic CGPM, to express quantity
values of catalytic activity of enzymes and other catalysts in any field of science and
technology. Further derived kinds-of-quantity and derived units may be defined in the
conventional manners.
• Journals subscribing to the preferred use of SI units can now with the blessing of the
CGPM promote international standardization of units by including the katal in their
Information for Authors, thus facilitating the globalization of medicine.
33
USP Position
Brief Description
34
And What’s Next?
• <1151> (in revision)
• General Chapters for all five routes of administration, e.g., <1>
• Performance Test General Chapters for all five routes of
administration with reference to PVT
• PVT/reference materials for periodic PVT, e.g., web toolkit
• Science dialogues (e.g., in house calibrator)
• ? Something beyond dissolution
• Regulatory implications
– Chapter V—Drugs and Devices
• Subchapter A—Drugs and Devices/Adulterated Drugs and
Devices
• Section 501(a)(1) if substance; or (2)(A) if unsanitary
conditions; or (B) if drug; or (3) if container; or (4) if color
additive (A) or (B); or (5) new animal drug; or (6) animal feed
bearing or containing a new animal drug
• Section 501 (b) official compendium (three)
• Sections (c) through (i) or more
• Implications for US/other markets
Quality Standards for Medicines, Supplements, and Food Ingredients throughout the World
35
PRODUCT QUALITY AND
PERFORMANCE:
CLINICAL PERSPECTIVE
HOWARD I. MAIBACH, M.D.
PROFESSOR
DEPARTMENT OF DERMATOLOGY
UNIVERSITY OF CALIFORNIA, SAN FRANCISCO
94143-0989
I. WHY Quality/Performance
Tests?
1. Rational approach
2. Great theory
3. Underpinning – for in vivo data
36
II. CLINICAL OBSERVATIONS - ? FAIL
!
• Corticoid
• Why:
– high placebo role
– No phase 4 efficacy system
– “just” switch drug
– Likely potency – not failure
37
IV. WHY Drug Release Data?
Exemplar
Clinical Testing
• Fluocinonide
Innovator >Generic
• Replication awaited
38
Scientific Weakness
SOLUTIONS?
• Paired comparison
• Armitage Stats
• Bioengineering techniques
39
Why are We waiting?
• A mandate!
(the next Kefauver)
• The Concept:
Comparative effectiveness Research
(An Intern Med 151:203)
MaibachH@derm.ucsf.edu
40