Martin E. Sandbu: Dicing with death?

A case study of Guidant Corporation

Dicing with death?

A case study of Guidant Corporation’s
implantable defibrillator business

Martin E. Sandbu
Senior Fellow, Zicklin Center for Business Ethics Research
Wharton School, University of Pennsylvania

with research assistance from Jeisun Wen

Revised September 2008

Copyright © 2008 by Martin E. Sandbu

Summary:
Guidant Corporation (today part of Boston Scientific) was a manufacturer of implantable
defibrillators—life-saving devices implanted into patients to deliver electric shock therapy to
irregularly beating hearts. In 2005, a young man implanted with one of Guidant’s
defibrillators, the Prizm 2 DR model 1861, died when his device failed to deliver its electric
pulses as he was suffering cardiac arrest. The cause was a short-circuit which Guidant had
known could affect units of this particular model manufactured before April, 2002. Guidant
had changed its manufacturing process to eliminate the potential malfunction, but had
chosen not to inform doctors. The stated reason was that the risk of malfunction was
minuscule, and that publicizing the malfunction would likely encourage some patients to
undergo replacement surgery, which carries risks of complication and even death at least as
great as the probability that a Prizm 2 DR unit would malfunction. Guidant was also in a
critical stage of takeover negotiations with the healthcare products giant Johnson &
Johnson, which had made an offer to acquire Guidant.

1
 Martin E. Sandbu: Dicing with death? A case study of Guidant Corporation

1. Introduction1
Joshua Oukrop, of Grand Rapids, Minnesota, was an active teenager who enjoyed an
outdoors lifestyle. When he was seventeen, Joshua began to experience bouts of faintness
and some times total blackout. His physician, Dr. Barry J. Maron, diagnosed him with
hypertrophic cardiomyopathy, a heart condition in which the wall of the heart is unusually
thick. The condition put Joshua at a high risk of dying young due to heart failure.

Dr. Maron recommended that Joshua undergo surgery to receive a heart defibrillator, an
electronic device designed to deliver an electrical shock in case the heart stops beating.
Joshua’s father, Lee Oukrop, suffered from the same heart condition and had already had a
defibrillator implanted. “I promised my son it would save his life,” Lee Oukrop said.

In October 2001, Joshua underwent surgery and received a Prizm 2 DR 1861 defibrillator
made by Guidant Corporation. Soon after, he returned to hiking, bicycling and
snowboarding back home in Grand Rapids, and was able to carry on his previous active
lifestyle. Joshua eventually matriculated into Bemidji State University in Minnesota, hoping
one day to become a teacher.

The impact on Joshua’s life of having the defibrillator implanted in his chest was minor.
Every three months, he would see Dr. Maron to have his device checked and maintained.
“Each time, it was normal,” Dr. Maron said.

In March of 2005, six weeks after his last checkup, Joshua went to Utah with his girlfriend
for spring break. During a mountain bike ride, he started complaining of fatigue. He
stopped, got off his bike, collapsed, and died of cardiac arrest.

Later tests by Guidant Corporation showed that his defibrillator, the Prizm 2 DR 1861, had
failed to deliver its lifesaving jolt when Joshua needed it. The cause of the malfunction was a
short-circuit that would not have been detectable in advance. It could have happened during
or before Joshua’s cardiac arrest, but because of it, the device was unable to save Joshua’s
life.

Two months after Joshua’s death, Guidant officials met with Joshua’s doctors to explain what
went wrong. The doctors learned that the company had already been aware of the possibility
of a short-circuit in the Prizm 2 DR model 1861, since they had observed it in other,
returned, devices (but no other patients were thought to have died due to the malfunction).
In fact, Guidant had changed its manufacturing process twice in 2002 to address the
problem, which had been eliminated after the changes. In compliance with the law, the
company had informed the federal Food and Drug Administration (FDA) of the
manufacturing changes, but was not required by law to inform doctors, and had decided not
to do so.

1The information in this section draws from Barry Meier, “Maker of Heart Device Kept Flaw From Doctors,”
New York Times, 24 May 2005, and Maura Lerner, “Hunting down dangers to the heart,” Minneapolis Star Tribune,
24 July 2005.

2
 Martin E. Sandbu: Dicing with death? A case study of Guidant Corporation

Joshua Oukrop’s doctors, Dr. Barry J. Maron and Dr. Robert G. Hauser, were outraged.
They told the New York Times that they would have replaced they unit if they had known of
the potential malfunction. In response, Dr. Joseph M. Smith, Senior Vice President and
Chief Medical Officer of Guidant’s Cardiac Rhythm Management Division, argued that the
risks associated with replacing the device, as many patients would be likely to do if the
company publicized the problem, were much higher than the risk of malfunction, which
remained very small and within product specifications.

2. The genesis of a dilemma

Guidant Corporation was originally a spin-off from Eli Lily and Company. In 1994, it
adopted the name Guidant—a name derived from the word guide, one who shows the way by
leading. It made its first initial public offering (IPO) on December 14, 1994, on the New
York Stock Exchange (ticker symbol: GDT). 2

In the following ten years, Guidant experienced strong growth in demand for its medical
devices, partly due to the aging population. From 1994 through 2004, sales went from just
under $900 million to $3.8 billion—an average growth rate of 16% per annum. Market
capitalization grew from $1 billion to $23 billion as of the end of 2004. By the same year,
Guidant employed over 12,000 individuals, and more than 2 million patients worldwide were
treated by its therapies.

Guidant was the first company in the world to introduce the automatic Implantable
Cardioverter Defibrillator (ICD), a device that restores a normal heartbeat for patients with
abnormally fast and life-threatening heart rhythms, which can lead to sudden cardiac arrest.
Guidant also pioneered the world’s first cardiac resynchronization therapy defibrillator (CRT-
D) for heart failure (the gradual weakening of the heart muscle that affects more than 20
million people worldwide).

Of its various business groups, Guidant’s Cardiac Rhythm Management business in

particular has consistently grown faster than the market, which increased by a total of 14
percent over the last decade. In 2004, Guidant’s implantable defibrillator sales reached $1.8
billion, an 18 percent year-over-year increase, fueled in large part by cardiac
resynchronization therapy defibrillators for heart failure patients. Sales of implantable
defibrillator systems make up 47% of total revenues (Exhibit 1).

In December of 2004, Guidant found itself courted by healthcare products giant Johnson &
Johnson which sought to acquire the medical devices manufacturer to bolster its portfolio of
medical products and enter a growing market in which it had little presence. On December
15th, 2004, Johnson & Johnson announced that it would offer Guidant shareholders 25.4
billion dollars in cash and stocks to take over the company, corresponding to $76 per share. 3

2The facts in this and the next three paragraphs are taken from Guidant Corporation’s Annual Report 2004,
available on http://web.archive.org/web/20060312024230/www.guidant.com/investors/annualreport/
Guidant_AR2004.pdf (accessed 9 September 2008).
3Guidant press release 15 December 2004. Available on
http://bostonscientific.mediaroom.com/index.php?s=64&item=191 (accessed 9 September 2008).

3
 Martin E. Sandbu: Dicing with death? A case study of Guidant Corporation

After the death of Joshua Oukrop in March, Guidant came under pressure due to media
attention to the case. On June 17th, 2005, the New York Times published its story about
Joshua and interviews with both his doctors and Guidant executives. The same day, Guidant
issued an advisory to physicians with information about the possible malfunction. 4 In the
advisory, Guidant recommended that patients do not replace their ICDs prior to the
appearance of normal replacement indicators. Despite the recommendation, Guidant
offered a free replacement device to any of its patients deciding to undergo surgery to
remove a PRIZM 2 DR 1861 unit produced before the April 2002 manufacturing
modification, as well as up to $2,500 of unreimbursed medical expenses incurred by patients
associated with replacement surgery. In a follow-up advisory issued on August 8th, Guidant
reported that no additional clinical failures had been observed with the model.5

By summer 2005, the deal with Johnson & Johnson had still not been closed. With the media
fallout from the malfunction in its Prizm 2 model, Guidant’s position was deteriorating fast.
Negotiating hard, Johnson & Johnson began to call for a lower price than the one originally
offered, citing the trouble Guidant was experiencing over its defibrillators, and the possible
financial ramifications, as the reason for the revaluation. Unless another company made a
competing offer for it, Guidant officials knew that they would have to accept a significantly
lower offer from Johnson & Johnson in a renegotiated deal. As expected, on November 15th,
2005, Johnson & Johnson revised its offer down to 21.5 billion dollars or $63.08 per share. 6

3. How ICDs work

Implantable Cardioverter Defibrillators (ICDs), the type of implanted device that was used
by Joshua Oukrop, constitute the mainstay of Guidant’s business. These devices help
regulate heart rhythm through the use of electrical charges, preventing potentially fatal
consequences of an irregular heartbeat. Guidant’s ICDs are surgically placed under the skin
and powered by batteries. They are often used in patients who are at risk of sudden cardiac
arrest, and can provide doctors with critical, detailed information regarding each episode in
which the ICD was activated to deliver therapy.

ICDs have grown increasingly complex since their inception, but generally share a number
of similar components: a computer microprocessor, a battery, and leads7 that all work
together to detect cardiac activity and deliver stimulating energy. The computer component
has become increasingly sophisticated, often combining diagnostic software, hundreds of
programmable options, and combined pacing, defibrillating, and heart failure therapy.

ICDs require replacement every 5 to 6 years depending on the model being used. Factors
that play into the need for replacement may include hardware and software malfunctions,

4Available on http://web.archive.org/web/20051028093844/www.guidant.com/physician_communications/
PRIZM2_DR.pdf (accessed 9 September 2008).
5Available on http://web.archive.org/web/20060312021524/www.guidant.com/physician_communications/
PRIZM2_Update.pdf (accessed 9 September 2008).
6 Kwan Yuk Pan, “Guidant rises on J&J revised bid,” Financial Times, 16 November 2005.
7 The wires that lead from the ICD into the heart, used to detect abnormalities and deliver electroshock
therapy.

4
 Martin E. Sandbu: Dicing with death? A case study of Guidant Corporation

frequency of therapy, and environmental conditions; however, the dominant reason for
almost all ICD replacements is battery depletion. Towards the end of its expected life, the
failure rate of ICDs increases exponentially, and so patients are often advised to replace their
units before this happens.

The ability of ICDs to prolong the expected lifespan of its users has been well documented
in medical literature. In the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT)
sponsored by the National Heart, Lung and Blood Institute (NHLBI), 2521 patients with
heart failure were randomly assigned to receive a placebo treatment, medication with the
drug amiodarone, or an ICD.8 After 5 years, the cumulative mortality rate was 36.1% in the
placebo group, 34.0% in the amiodarone group, and 28.9% in the ICD group. While the risk
of death in the group receiving drug medication was not statistically different from the
placebo group, ICD therapy was associated with a statistically significant reduction in the risk
of death of 7.2 percentage points. The results of the study are illustrated in Exhibit 2.

According to another study, “fewer than 20% of the 1.6 million US patients with recognized
indications for ICD implantation as defined by the Center for Medicare and Medicaid
Services have a device implanted. Another 600,000 US patients have heart failure but none
of the Center’s indications for ICDs; They may also be candidates for ICDs according to the
SCD-HeFT trial, but almost none of them have received an ICD.”9 The potential of ICDs
to benefit millions of Americans with deadly heart conditions has yet to be fully realized.

4. A faulty product?
The increasing complexity of ICDs and the intrusive operation required to implant them
raise obvious concerns regarding the safety of the devices. As with any mechanical device,
failures will occur. Batteries drain over time, decreasing reliability. Leads are exposed to the
mechanical forces of a beating heart and their wires and insulation may wear out over time.

On July 1st, 2005, the FDA issued a report that identified Guidant’s June 17th advisory
regarding the Prizm 2 DR device as a Class I recall. 10 Devices carrying this classification are
those believed by the FDA to have the potential to cause serious adverse health
consequences (i.e. death) in the case of malfunction.

In mid-September 2005, Guidant issued a comprehensive Product Performance Report to

make quality and performance information publicly available. Details concerning the failures
experienced by the Prizm 2 DR were also included in the document.11

8Gust H. Bardy, Kerry L. Lee, et al., “Amiodarone or an Implantable Cardioverter–Defibrillator for Congestive
Heart Failure,” The New England Journal of Medicine, Vol. 352, No. 3 (January 20, 2005), pp. 225–237.
9J. David Burkhardt and Bruce L. Wilkoff, “Malfunctions in implantable cardiac devices: Putting the risk in
perspective,” Cleveland Clinic Journal of Medicine, Vol. 72, No. 9 (September 2005), pp. 736–744.
10“FDA Updates Consumers on Guidant Corporation’s Implantable Defibrillators,” FDA Press Release
P05-37, 1 July 2005. Available on http://www.fda.gov/bbs/topics/NEWS/2005/NEW01198.html
(accessed 9 September 2008).
11Available on http://web.archive.org/web/20051126074952/www.guidant.com/physician/
product_performance_report.pdf (accessed 9 September 2008). A summary table is reproduced in Exhibit 3.

5
 Martin E. Sandbu: Dicing with death? A case study of Guidant Corporation

The product performance report counts the total number of U.S. registered implants for the
device at 43,000. From these, 124 non-premature battery depletion related failures had been
confirmed, or 0.29%. (Only 63 of the failures, or 0.15%, were of a nature that compromised
the electroshock therapy and could result in the death of the patient). Thus the overall
failure rate must is very low, in particular in comparison with the significant reduction in
mortality that devices provide. Furthermore, 26 of the failures that compromised therapy
were due to the short-circuit that proved fatal to Joshua Oukrop. Apart from the newly
discovered short-curcuit, therefore, the rate of confirmed failures in the U.S. population of
units is 0.23% (or 0.11% if we exclude the failures that did not compromise therapy).

The short-circuit has only occurred in units manufactured prior to April 2002, when
Guidant changed its manufacturing process to eliminate the problem, and is not thought to
be possible in units produced at later dates. No failures have been observed in post-April
2002 products. In addition, Guidant improved its manufacturing process again in November
2002 to provide further safeguards against manufacturing defects. The worldwide number of
Prizm 2 DR unites produced before these changes was 26,000, of which 14,000 remained
implanted in U.S. patients as of August 2005. The manufacturing process changes were
reported to the Food and Drug Administration, in compliance with the law; however,
Guidant was not required to publicize either the changes or the reason for them. In
particular, they were not legally required to divulge the potential malfunction to doctors.

Guidant had documented a total of 28 cases of short-circuits from the 26,000 vulnerable
implants (the 26 in the United States mentioned above, and two more abroad). This
translates to a failure rate of 0.11% among vulnerable units, which would come on top of
the 0.23% failure rate for Prizm 2 DR devices from other causes. Naturally, the risk could be
higher, since not all failures are necessarily noticed and documented. In returned product
testing, Guidant examined 1,005 devices manufactured on or before April 16, 2002. Out of
this sample, four failures (0.40%) were provoked. Guidant noted, however, that these devices
represent only a small, non-random sample of the implanted population, and may not be
representative of the rest of the active units. The company estimated the risk of short-
circuits in pre-April 2002 units between 0.10% and 0.24%.

5. The dangers of information

In response to criticisms from Joshua Oukrop’s doctors, Guidant pointed to the risks
associated with undergoing surgery to replace an ICD before its normal expiration. The risk
of complications under or after such surgery is indeed considerable. A study published in
the medical journal Heart estimated the total risk complications from either first implant or
replacement surgery to be on average 1.9%.12 The different types of complications include
wound infection, erosion13, lead displacement and/or malfunction, non-healing wounds, and
can in rare cases cause death. The risk of complication from replacement was in fact much
higher than for first implants. Of the 245 patients who underwent replacement surgery there
were a total of 16 complications (6.5%) of which one death (0.41%).

12A. A. Harcombe, S. A. Newell, P. F. Ludman et al., “Late complications following permanent pacemaker
implantations or elective unit replacement,” Heart, Vol. 80 (1998), pp. 240–244.
13 Deterioration of tissue membranes due to physical abrasion or inflammation.

6
 Martin E. Sandbu: Dicing with death? A case study of Guidant Corporation

On average, the increase in the risk of death attributable to possible short-circuits in the
Prizm 2 DR—between 0.10 and 0.24 percentage points, on Guidant’s estimates—was
therefore comparable to the 0.41% risk of death due to early replacement surgery. Both
numbers also imply no more than tiny marginal changes to the average risk of death among
heart failure sufferers with ICDs—28.9% over five years, which is the typical lifetime of a
device, as against 36.1% for those without one. We can put the numbers differently. Of the
14,000 U.S. patients who still had a vulnerable Prizm 2 DR unit implanted, an expected
number of 14 to 34 of them might be affected by the short-circuit. If all 14,000 chose to
undergo early replacement surgery, those 14 to 34 would be saved, but means an expected 57
people would die in surgery (and many more would suffer non-lethal complications).14

Guidant’s argument for why they had not publicized the manufacturing defect referred to the
consequences of doing so on the health of patients who had the device implanted. The
numbers above support the contention that publicizing the problem might have led to worse
consequences than Guidant’s chosen course of action. Patients, worrying about the short-
circuit, might demand replacement surgery before normal schedule, which is as risky as
leaving the device in or even riskier when counting non-lethal surgery complications. From a
utilitarian point of view, this constitutes a reason not to divulge the information, since more
lives might be lost from the ensuing replacement surgeries than from the malfunction itself.

Even in the best of cases, the minute improvement in net risk that replacement surgery
might provide, raises the question of whether it warrants the cost of a new device and the
surgery itself. The device alone may cost $22,000. 15 For people who qualify for Medicare, the
government reimburses up $30,000. But otherwise (and above the reimbursement limit),
patients and their insurance providers are left to pick up the tab. After issuing its advisory,
Guidant announced that it would be bear the cost of the replacement units and pay patients
up to $2,500 to defray costs of unreimbursed medical care. No matter who foots the bill,
however, marginally reducing the risk of death by replacement surgery comes at a high price.

6. Legal and ethical context

The federal government regulates the manufacturing and use of medical devices under the
1990 Safe Medical Devices Act and the 1992 Medical Device Amendments. Post-sale
surveillance of medical devices was strengthened by requiring health care facilities to report
device-related serious injuries or deaths, by establishing tracking of certain high-risk devices,
and by giving the FDA authority to require tracking for any other device. Manufacturers were
also required to report to the FDA any device malfunction that could cause significant injury
to a patient. As a result of this legislation, the FDA has received numerous device-related
adverse event reports. The law, however, left it to Guidant’s discretion whether to inform
doctors of the potential malfunction.

14The expected number of surgery deaths is somewhat misleading, since patients eventually need replacement
surgery anyway when the unit wears out for normal reasons. Still, an increased pace of replacement surgery
implies more replacement surgeries overall, so that over a given lifetime, many people will end up with one
more replacement surgery (with the attendant exposure to risk) than otherwise. If this is half the population,
the expected number of surgery deaths is 29.
15 Cinda Becker, “Stuck with the check?,” Modern Healthcare, Vol. 35, No. 26 (27th June 2005), p. 20–22.

7
 Martin E. Sandbu: Dicing with death? A case study of Guidant Corporation

Guidant exercised that discretion with the decision not to divulge information about the
manufacturing defect. They later defended that decision based on the risks of replacement
surgery, presented in the previous section.

Joshua’s doctor, Dr. Maron of Abbott Northwestern Hospital in Minneapolis, refused to

accept Guidant’s utilitarian argument. “It is a statistical argument that has little to do with
real people,” Dr. Maron said. He also said that the numbers reported to Guidant might
understate the situation because product problems could go undetected or might not be
reported.16

The Heart Rhythm Society responded to these events by calling for a task force to examine
and develop new guidelines to better protect patients. In a press release, Anne B. Curtis, MD,
president of the Heart Rhythm Society said, “Patients need to discuss the variety of
treatment options available with the heart rhythm specialist overseeing their care. Each
patient is unique and the decision regarding ICD treatment based on recall information from
the manufacturer and the FDA will depend on the patient’s specific medical condition.”17

In an article published by the Journal of the American Medical Association, Dr. William H.
Maisel writes, “Manufacturers are required to report to the FDA any device malfunction that
causes or could cause significant injury. The decision for a manufacturer, however, of
whether or not to notify physicians, patients, or the public about an observed malfunction is
less straightforward. Historically, judgments have been made on a case-by-case basis by
considering factors such as rate of malfunction, likelihood of patient injury, cause of device
failure, and potential to mitigate the problem with an intervention. Because of the enormous
financial consequences of the manufacturer’s decision, there is also an inherent conflict of
interest.”18

Conclusion
When Guidant found out about the malfunction, it had to choose whether or not to inform
the public about the short-circuits in the Prizm 2 DR. While it complied with its legal
obligation to report manufacturing changes to the FDA, it did not go beyond what
disclosure the law demanded, and only released its findings once the the New York Times was
about to reveal them. A utilitarian argument for withholding the information exists: More
lives could be lost, and certainly more complications could be caused, if the information was
widely publicized than if the findings were kept under wraps. Guidant had the additional
concern that media coverage on the problem would upset the takeover negotiations with
Johnson & Johnson at a critical stage. Many patients and medical professionals,, meanwhile,
claimed that they had right to be informed. Did Guidant do the right thing?

16 Barry Meier, “Maker of Heart Device Kept Flaw From Doctors,” New York Times, 24th May, 2005.
Heart Rhythm Society, “Heart Rhythm Society to Develop Guidelines on ICD Recalls: Task Force to Begin
17

Drafting Recommendations on Manufacturer Notification Standards,” Press release 22 June 2005.

18William H. Maisel, “Safety Issues Involving Medical Devices Implications of Recent Implantable
Cardioverter-Defibrillator Malfunctions,” Journal of the American Medical Association, 24/31 August, 2005.

8
 Martin E. Sandbu: Dicing with death? A case study of Guidant Corporation

Exhibit 1: Product composition of Guidant’s sales

nse Sales by Product

Year ended December 31, 1994

17%

27%
$516

54%
$515

2%
$411

Year ended December 31, 2004

47%
19%

12%

14%
8%

Implantable Defibrillator Systems

Pacemaker Systems
Angioplasty Systems
2 ’03 ’04 Cardiac Surgery, Biliary, Peripheral, Carotid
Systems and other
Coronary Stent Systems

Source: Guidant Corporation Annual Report 2004

9
 Martin E. Sandbu: Dicing with death? A case study of Guidant Corporation

Exhibit 2: Therapies in Sudden Cardiac Death

The new england journal of medicine
in Hearth Failure Trial (SCD-FeFT)

Hazard Ratio (97.5% CI) P Value

Amiodarone vs. placebo 1.06 (0.86–1.30) 0.53
ICD therapy vs. placebo 0.77 (0.62–0.96) 0.007
0.4
Placebo
(244 deaths; 5-yr event rate, 0.361)

0.3 ICD therapy

Amiodarone (182 deaths; 5-yr event rate, 0.289)
Mortality Rate

(240 deaths; 5-yr event rate, 0.340)

0.2

0.1

0.0
0 12 24 36 48 60
Months of Follow-up
No. at Risk
Amiodarone 845 772 715 484 280 97
Placebo 847 797 724 505 304 89
ICD therapy 829 778 733 501 304 103

Figure 1. Kaplan–Meier Estimates of Death from Any Cause.

CI denotes confidence interval.

Source: Figure 1 in Gust H. Bardy, Kerry L. Lee, et al.,

“Amiodarone or an Implantable Cardioverter–Defibrillator for Congestive Heart Failure,”
follow-up occurred in 188 patients (11 percent) in Kaplan–Meier mortality curves are shown in Figure
The New England Journal of Medicine, Vol. 352, No. 3 (January 20, 2005), pp. 225–237.
the drug groups. The median time from randomiza- 1. The relative risk reduction of ICD therapy as com-
tion to crossover was 26.7 months. pared with placebo was 23 percent, and the absolute
reduction at five years was 7.2 percentage points.
icd shocks
Of the 829 patients in the ICD group, 259 (31 per- prespecified subgroups
cent) were known to have received shocks from their Mortality curves and hazard ratios for the compari-
device for any cause, with 177 (68 percent of those son of placebo with amiodarone and with ICD ther-
shocked, or 21 percent of the ICD group) receiving apy according to the prespecified subgroups defined
shocks for rapid ventricular tachycardia or fibrilla- by the cause of CHF and NYHA class are shown in
tion. During five years of follow-up, the average an- Figures 2 and 3, respectively. There was no interac-
nual rate of ICD shocks was 7.5 percent. For appro- tion of either amiodarone therapy (P=0.93) or ICD
priate shocks only (i.e., shocks for rapid, sustained therapy (P=0.68) with the cause of CHF. The inter-
ventricular tachycardia or fibrillation), the average action between amiodarone and NYHA class was
annual rate of ICD shocks was 5.1 percent. significant (P=0.004). Among patients with NYHA
class III CHF, there was a relative 44 percent increase
primary outcome in the risk of death among patients in the amioda-
A total of 666 patients died: 244 (29 percent) in the rone group, as compared with those in the placebo
placebo group, 240 (28 percent) in the amiodarone group (hazard ratio, 1.44; 97.5 percent confidence
group, and 182 (22 percent) in the ICD group. As interval, 1.05 to 1.97). Among patients with NYHA
compared with placebo, amiodarone therapy was class II CHF, no excess risk of death was associated
associated with a similar risk of death (hazard ratio, with amiodarone therapy, as compared with placebo
1.06; 97.5 percent confidence interval, 0.86 to 1.30;10(hazard ratio, 0.85; 97.5 percent confidence inter-
P=0.53) and ICD therapy was associated with a de- val, 0.65 to 1.11).
creased risk of death (hazard ratio, 0.77; 97.5 per- The interaction between ICD therapy and NYHA
cent confidence interval, 0.62 to 0.96; P=0.007). class was also significant (P<0.001). Among pa-
LURE DETAILS: PULSE GENERATORS
Martin E. Sandbu: Dicing with death? A case study of Guidant Corporation

) Exhibit 3: Guidant product performance report for

Ventak Prizm 2 DR model 1861

VENTAK PRIZM 2 DR Model 1861

Total Registered U.S. Implants: 43,000
U.S. registered implants) U.S. Confirmed Failures: 124 (0.29% of U.S. registered implants)
U.S. Approval Date: August 2000
Average Device Age: 30.1 mo.



 
MED .O#ONFIRMED #ONFIRMED
ES &AILURES &AILURES

herapy Therapy Total Therapy Therapy Total

vailable Compromised Available Compromised

– – 0 Electrical 16 7 23
14
Capacitor 13 3
1 1 2 18
Integrated circuit 1 2
1 – 21
Device tones 2 –
– 1 24
Integrated circuit – 1
– – 0
19
Resistor – 1

1 – 1 Mechanical 12 36 48
1 –
7
Short circuit (PRIZM 2 DR) – 26
(Advisory issued)
2 1 3 12
Seal plug 9 9
15
Header 2 1
11
Setscrew 1 –

Software 8 – 8
25
Software download 7 –
27
Impedance measurements 1 –

Other 25 20 45
Non–patterned 13 17
28
Battery depletion 8 3
26
Power on reset 2 –
29
Battery voltage at implant 2 –

U.S. Confirmed Failures 61 63 124

Source: Guidant Corporation Product Performance Report 2005

11
 Martin E. Sandbu: Dicing with death? A case study of Guidant Corporation
Downloaded from heart.bmjjournals.com on 6 December 2
Exhibit 4: Late complications following
Late complications of pacemakers
ICD implantation or replacement surgery
Table 2 Number of complications in the late period following pacemaker implantation or procedure. P
elective unit replacement electrodes w
Electrode
intravenous v
Infection Erosion problems Miscellaneous Totals procedure. S
First implants
hexidine gluc
Single chamber procedures w
(n = 1985) 9 (0.4%) 8 (0.4%) 3 (0.2%) 5 (0.2%) 25 (1.3%)
Dual chamber
(n = 391) 4 (1%) 2 (0.5%) 1 (0.3%) 2 (0.5%) 9 (2%) OPERATORS
EUR (n = 245; 10 dual There were 3
chamber) 5 (1.6%) 10 (4.1%) 1 (0.4%) 0 16 (6.5%)
this study (ta
Total (n = 2621) 18 (0.7%) 20 (0.8%) 5 (0.2%) 7 (0.3%) 50 (1.9%) operators as
Fourteen wer
EUR, elective unit replacement.
seven were
Source: Table 2 from A. A. Harcombe, S. A. Newell, P. F. Ludman et al., implants), six
stored: patient details, indications for pacing,
“Late complications following permanent pacemaker implantations or elective unit replacement,”
operator,
Heart, venous
Vol. 80 (1998), access route, pacemaker gen-
pp. 240–244.
performed few
erator, electrode type, and sensing and pacing and five wer
thresholds. Subsequent complications are re- fewer than 12
corded on the same database. All single and
dual chamber pacemaker implantations and STATISTICS
EUR procedures during the study period (11 Complication
years) were analysed for complications occur- number of c
ring up to one year after the last implantation implanted, an
(that is, up to 31 December 1995). Late com- rates are give
plications were defined as those occurrences Pearson !2 s
requiring operative intervention later than six Comparisons
weeks after the procedure. Complications and EUR ha
which could be dealt with by reprogramming portion of sin
were not included. In studies of early complica- in each group
tions the first six weeks postimplantation have Probability (p
been used to define the early period, except in sidered signifi
that of Aggarwal et al, where the first two
months were studied.4 Unit replacements
required as a result of complications (early or Results
late) were excluded. The definitions used in In total there
this study have been described previously.1 complications
Implantation was performed in the radiology to 97 years),
department under local anaesthetic after pre- male, giving a
medication with an oral benzodiazepine. From confidence in
January 1984 to February 1990, no antibiotics at a median o
years) after u
were given.12 From March 1990, patients were
(table 2).
treated with oral antibiotics (flucloxacillin, or
erythromycin in case of allergy) for 48 hours,
the first dose given immediately before the COMPLICATION