Published Online: October 8, 2012. doi:10.
1001
1.0
0.9
0.8
Warning and/or Withdrawal
Probability of No Safety
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0 medications. JAMA. 2002;287(17):2215-2220.
0 1000 2000 3000 4000 4000 4000
Days Since Receipt of Notice of Compliance
Figure. Kaplan-Meier estimate of new active substance survival without a
safety warning and/or withdrawal. Sixty-eight products had serious safety
warnings only; 9 products had serious safety warnings and were then
withdrawn; and 7 products were withdrawn without prior serious safety
warnings.The red lines represent the 95% confidence intervals.
issue compared with a 34.2% (95% CI, 24.3-44.2) esti-
mate for an NAS with a priority review (P = .005, log-
rank test). Eighty-one NASs with a priority review that
were not major therapeutic advances were compared with
the 321 NASs with a standard review. The estimate of
this group of priority review NASs having a serious safety
issue was 36.0% (95% CI, 24.3-47.7) compared with
19.8% (95% CI, 14.8-24.8) for standard review NASs
(P=.004, log-rank test).
Priority reviews and standard reviews were assigned
to 42 and 45 NASs, respectively, for 5 serious diseases.
The estimate of the first group having a serious safety is-
sue was 31.1% (95% CI, 14.3-46.0) vs 34.9% (95% CI,
16.8-52.9) for the second group (P = .96, log-rank test).
Comment. Just fewer than one-fourth (23.7%) of all NASs
introduced between 1995 and 2010 had serious safety is-
sues. This result is similar to that reported by Lasser et
al.1 The difference between drugs with a standard ap-
proval and those with a priority approval in terms of their
safety record may be attributable to the shorter period
that the latter spends in the approval process and may
reflect deficiencies in Health Canada’s priority review
process.
Alternative explanations for the difference between
standard and review NASs are less likely. Priority re-
view drugs that were not major therapeutic advances were
still more likely to acquire serious safety issues than stan-
dard review NASs. If priority review drugs for serious dis-
eases were approved with a lower benefit to harm ratio,
they should be more likely to develop serious safety is-
sues than standard review drugs for the same diseases.
However, there was no difference between the 2 groups.
New products that offer major therapeutic advantages
should be embraced even with the significant lacunae that
exist about their safety, but because most NASs do not
fall into this category,3,4 clinicians and patients should
use these drugs very cautiously.
Joel Lexchin, MSc, MD
ARCH INTERN MED/ VOL 172 (NO. 21), NOV 26, 2012
©2012 American Medical Association. All rights reserved.
Downloaded From: http://archinte.jamanetwork.com/ by a J H Quillen College User on 05/27/2015
/archinternmed.2012.4444
Author Affiliations: School of Health Policy and Man-
agement, York University, and University Health Net-
work, Department of Family and Community Medi-
cine, University of Toronto, Toronto, Ontario, Canada.
Correspondence: Dr Lexchin, School of Health Policy
and Management, York University, 4700 Keele St,
Toronto, ON M3J 1P3, Canada (jlexchin@yorku.ca).
Financial Disclosure: None reported.
1. Lasser KE, Allen PD, Woolhandler SJ, Himmelstein DU, Wolfe SM, Bor DH.
Timing of new black box warnings and withdrawals for prescription
7000 2. Berlin RJ. Examination of the relationship between oncology drug labeling
revision frequency and FDA product categorization. Am J Public Health. 2009;
99(9):1693-1698.
3. Prescrire Editorial Staff. New drugs and indications in 2010: inadequate as-
sessment; patients at risk. Prescrire Int. 2011;20(115):105-107, 109-110.
4. Annual report 2009. Patented Medicine Prices Review Board website. http:
//www.pmprb-cepmb.gc.ca/english/view.asp?x=1340&mid=1187.Ac-
cessed June 9, 2012.
INVITED COMMENTARY
New Drug: Caution Indicated
I n assessing a new therapeutic drug, regulators at
Health Canada and the US Food and Drug
Administration (FDA) are entrusted with making
critical and difficult scientific judgments that can
affect the health of hundreds of thousands of patients
in a matter of months after product launch. In drug
evaluation, the cost of error may be high. Casualties in
pharmaceutical disasters are measured in tens of thou-
sands, and the population exposed to unsafe drugs
often numbers in the millions.1 In a new research let-
ter, Lexchin2 provides a valuable but basic report card
for the drug approval decisions at Health Canada over
a 16-year period.
The first lesson from Lexchin’s study is that serious
safety issues were not detected or fully understood at
approval for 1 out of 5 drugs, and 17 of the 434 drugs
approved were later withdrawn for safety reasons. The
results were worse for the subset of priority review
drugs—those evaluated under deadline pressure
because they were intended to treat a life-threatening
disease or were thought to provide an important thera-
peutic advance. Of those drugs, 1 out of 3 was approved
without a warning about a serious safety issue. A recent
report from the Institute of Medicine for the United
States suggests a future vision under which preapproval
testing is only the first installment in a life cycle re-
search program extending over many years.3 In a recent
article in the Archives,4 my colleagues and I reported on
a single year’s major postmarket safety actions by the
FDA and noted that 25 drugs received new boxed warn-
ings and that the 181 safety actions that were approved
in 2009 occurred a median of 11 years after initial
approval.
WWW.ARCHINTERNMED.COM
1681
 The scientific record required to support the ap-
proval of a new molecular entity, or in Canada a new ac-
tive substance, typically amounts to hundreds of thou-
sands of pages of densely written scientific reports. In vitro
studies explore the mechanism of action. The chemical
stability, human metabolization, and consistent manu-
facture of a new molecule require careful evaluation. Ben-
eficial effects, high doses, and long-term exposure are ex-
amined in animal models that may or may not predict
effects in humans. Hundreds to thousands of patients are
exposed in preapproval clinical testing. Although many
new drugs are intended for long-term use, the interna-
tional standard requires that only a few hundred pa-
tients be exposed to these drugs for 12 months or more.5
Regulatory agency reviewers are faced with a formi-
dable mass of scientific data that contain only modest
amounts of information from direct patient testing. There-
fore, important clinical questions may remain unan-
swered. There are increasing demands by legislatures, in-
dustry, medical professionals, and patients to make these
difficult decisions even more quickly in both Canada and
the United States.
The current FDA “Expedited Drug Development Path-
way” now includes provisions to reduce the number and
size of clinical trials, to accept more limited evidence of
efficacy, and to initiate drug review before the comple-
tion of testing. It also sets short review deadlines. In fis-
cal year 2011, a total of 16 of 35 new molecular entities
(46%) received a priority review from the FDA, 13 (37%)
received additional fast-track treatment, and 24 (69%)
were approved in the United States earlier than in Eu-
rope or Canada.6
Getting faster access to newly developed, less thor-
oughly tested drugs is at best a mixed blessing. For the
first 3 years after approval, new drugs should carry a spe-
cial warning akin to the black triangle used in Britain. It
should be prominent and mean to every physician, “New
Drug: Caution Indicated.”
Thomas J. Moore, AB
Published Online: October 8, 2012. doi:10.1001/2013
.jamainternmed.610
Author Affiliation: Institute for Safe Medication Prac-
tices, Horsham, Pennsylvania.
Correspondence: Mr Moore, Institute for Safe Medica-
tion Practices, 101 N Columbus St, Ste 410, Alexandria,
VA 22314 (tmoore@ismp.org).
Financial Disclosure: None reported.
1. Wood AJ. The safety of new medicines: the importance of asking the right
questions. JAMA. 1999;281(18):1753-1754.
2. Lexchin J. New drugs and safety: what happened to new active substances
approved in Canada between 1995 and 2010? [published online October 8,
2012]. Arch Intern Med. 2012;172(21):1680-1681.
3. Psaty BM, Meslin EM, Breckenridge A. A lifecycle approach to the evalua-
tion of FDA approval methods and regulatory actions: opportunities pro-
vided by a new IOM report. JAMA. 2012;307(23):2491-2492.
4. Moore TJ, Singh S, Furberg CD. The FDA and new safety warnings. Arch In-
tern Med. 2012;172(1):78-80.
5. The extent of population exposure to assess clinical safety: for drugs in-
tended for long-term treatment of non-life-threatening conditions, 1995. In-
ternational Conference on Harmonisation website. http://www.ich.org/products
/guidelines/efficacy/efficacy-single/article/the-extent-of-population
-exposure-to-assess-clinical-safety-for-drugs-intended-for-long-term-treatme
.html. Accessed August 16, 2012.
6. US Food and Drug Administration. FY2011 Innovative Drug Approvals. Silver
Spring, MD: US Food and Drug Administration; November 2011.
ARCH INTERN MED/ VOL 172 (NO. 21), NOV 26, 2012
1682
©2012 American Medical Association. All rights reserved.
Downloaded From: http://archinte.jamanetwork.com/ by a J H Quillen College User on 05/27/2015
RESEARCH LETTER
HEALTH CARE REFORM
Cost-effectiveness of Enhanced Depression
Care After Acute Coronary Syndrome:
Results From the Coronary Psychosocial
Evaluation Studies Randomized
Controlled Trial
D epression following an acute coronary syn-
drome (ACS) affects 2 in 5 patients and is one of
the most important psychosocial predictors of a
poor cardiovascular prognosis.1 In the Coronary Psycho-
social Evaluation Studies (COPES) randomized con-
trolled trial, we compared the effectiveness of enhanced de-
pression care, which comprised patient preference for
problem-solving psychotherapy, antidepressant use, or
both, through the use of a stepped-care algorithm, with
usual care in patients with ACS and persistent depressive
symptoms 3 months after discharge.2 The 6-month trial
demonstrated that enhanced depression care improved
patient satisfaction with treatment and reduced depres-
sive symptoms. However, the intervention’s impact on
health-related quality of life, health care utilization, and
cost-effectiveness has not been evaluated. To help bridge
this gap and inform decision making, we undertook a
cost-effectiveness analysis of enhanced depression treat-
ment in patients with ACS and persistent depressive
symptoms using results from the COPES trial.
See Invited Commentary
at end of letter
Methods. Data. We interviewed patients to determine
their antidepressant and anxiolytic medication use and
dose; ambulatory care visits with mental health spe-
cialists, cardiologists, and primary care physicians; and
hospitalizations for stable angina, unstable angina,
ST-segment elevation or non–ST-segment elevation
myocardial infarction, and congestive heart failure.
Hospitalizations were confirmed by medical chart
review and adjudicated by 2 board-certified cardiolo-
gists. Hospital electronic health records were also
actively surveyed for hospitalizations. Costs were esti-
mated using average wholesale drug prices and Medi-
care reimbursement rates. Standardized measures of
quality of life were obtained using the 12-Item Short
Form Health Survey (SF-12) and converted to health
utilities using the Short Form 6 Dimensions (SF-6D)
scoring algorithm.3
Statistical and Cost Analysis. Quality of life, health
care utilization, and cost outcomes were adjusted for
potential confounding by age, sex, ethnicity, race,
marital status, education, depressive symptom sever-
ity, type of ACS, and left ventricular ejection fraction
using linear regression models. To determine cost-
effectiveness, mean incremental costs and mean incre-
mental quality-adjusted life-years were estimated
using 6-month outcomes. We performed nonparamet-
WWW.ARCHINTERNMED.COM