Persons using assistive technology may not be able to fully access information in this file.

For assistance, e-mail biolincc@[Link]. Include the Web site and filename in your messa

V-1

The BARI Protocol

Protocol for the Bypass Angioplasty

Revascularization Investigation
In this supplement we present the design of the Bypass Angioplasty Revascularization
Investigation (BARI). The BARI team of investigators, data coordinators, staff, and support
committees are committed to providing the highest quality data to scientifically test the
proposed hypotheses. Considerable controversy exists about the extension of percutaneous
transluminal coronary angioplasty (PTCA) into the arena of therapy for multivessel coronary
artery disease. Although coronary artery bypass surgery has been thoroughly compared with
medical therapy in randomized trials, studies of PTCA to date have been observational in
nature. The presumption that the results of trials of coronary bypass surgery can be applied to
the use of PTCA is not established; this creates a dilemma that requires resolution in the most
objective and scientific manner. In the Wangensteens's book (-The Rise of Surgery, 1978), there
is an excellent discussion of the controversy that surrounded the work of Semmellweiss. In one
of the earliest efforts to bring science to the bedside and to influence medical practice, he
investigated childbirth outcome in terms of whether the physician's hands were washed before
delivery. Those times were filled with extreme reluctance on the part of the medical
establishment to accept the observations of Semmellweiss. As the Wangensteens suggest, "Trial,
not debate, is the proper manner in which to resolve a question of this kind." It is in this spirit
that the BARI investigators are conducting the BARI trial. To the best of our ability, we will
provide information that will benefit patients and physicians in making decisions on revascu-
larization procedures. (Circulation 1991;84[suppl VI :V-1-V-27)

T he application of invasive therapy to the treat- been designed to accomplish this through random
ment of severe coronary artery disease assignment of revascularization strategy and system-
(CAD) has progressively increased during atic follow-up over 5 years. Although the primary
the past decade. In the United States in 1988 approx- clinical indication for revascularization varies among
imately 250,000 patients with CAD were treated with BARI patients (symptom relief or treatment of pro-
percutaneous transluminal coronary angioplasty found ischemia), all patients are judged to be at
(PTCA), and nearly as many patients received coro- relatively high risk for subsequent cardiac events.
nary artery bypass graft (CABG) surgery (National In addition to the clinical trial component, all
Center for Health Statistics, 1988; personal commu- patients who are eligible but refuse random assign-
nication). These numbers represent dramatic in- ment are asked to participate in the BARI Registry.
creases in the use of invasive procedures compared The registry also contains a 5-10% random sample of
with those of 1980, when 6,000 PTCAs and 137,000 those who are deemed ineligible for random assign-
CABG operations were performed.' Therefore, cor- ment because they are considered angiographically
onary revascularization is an important component of unsuitable for PTCA and/or CABG.
health care costs in the United States, with current Evidence from this clinical trial will provide a
direct costs easily exceeding the prior estimate of scientific basis for choosing PTCA or CABG as the
over $5 billion for CABG alone in 1984.2 initial revascularization treatment of severe multi-
At the present time, the choice between PTCA and vessel coronary disease.
CABG for patients with multivessel CAD who need
revascularization and who are suitable for either Specific Aims
procedure represents a clinical dilemma because the BARI is a comparative study of PTCA and CABG,
relative indications for PTCA and CABG in these the two most prevalent revascularization methods
patients are not yet clearly defined. used to treat advanced CAD. The study focuses on
Before informed therapeutic choices between the the treatment of patients who have multivessel dis-
two procedures can be made, controlled studies are ease and severe angina or ischemia, those who re-
needed to objectively compare the benefits and risks quire revascularization, and those who are suitable
of an initial strategy of PTCA versus CABG in for either procedure. The primary aim of BARI is to
appropriately selected patients. The Bypass Angio- test the hypothesis that an initial strategy of PTCA in
plasty Revascularization Investigation (BARI) has these eligible patients compared with CABG does
V-2 Supplement V Circulation Vol 84, No 6 December 1991

not compromise clinical outcome during a 5-year groups will be defined by the number of significantly
follow-up period. stenosed vessels, the number and proportion of myo-
cardial territories with a jeopardized coronary sup-
Aims of the Randomized Clinical Trial ply, the complexity of the lesion and vessel anatomy,
Because CABG is of established benefit for pa- and the degree of left ventricular function.
tients with severely symptomatic multivessel disease,
the use of an alternate treatment strategy, regardless Aims of the Bypass Angioplasty Revascularization
of its potential efficacy, must not impose a greater Investigation Registry
risk of mortality than CABG. For this reason, and to The BARI Registry will include eligible patients
provide a reliable end point for calculation of sample who refuse random assignment as well as a 5-10%
size, mortality is the primary end point of the trial. sample of patients who are excluded from the trial
The sample size was selected to enable BARI to rule based on angiographic criteria. The group of eligible
out with high probability that the 5-year mortality patients who refused random assignment will lend
rate with PTCA exceeds the 5-year mortality rate itself to the investigation of selection factors involved
with CABG by more than 2.5%. (For a discussion of in the choice of PTCA versus CABG. In addition, we
sample size calculations, see Appendix 3.) Although will be able to compare the results of treatment
mortality is essential in assessing the safety of the selected by choice with that selected by random
PTCA strategy in patients with multivessel CAD, allocation. The Registry patients who are ineligible
other end points of clinical outcome are of critical for random assignment because of angiographic cri-
importance, particularly if there is no difference in teria will be used to assess how angiographic exclu-
mortality between the two treatment strategies. Be- sion practices differ across sites, which treatment
cause the acceptable difference in mortality is small, such patients actually receive, and the long-term
a large sample size is required. This large sample size outcome with the given treatment.
also provides sufficient power to examine treatment
differences in rates of myocardial infarction (MI), Patient Selection
repeat revascularization, and recurrent severe angina Inclusion Criteria
or ischemia. Distribution of exercise capacity, ven-
tricular function, and need for medication will also Patients included in BARI must be representative
be compared by treatment assignment at selected of those who have multivessel coronary disease and are
follow-up points. treated for severe angina or myocardial ischemia in
In addition, BARI will provide much-needed an- current clinical practice. Specifically, eligible patients
swers to questions concerning the economic and psy- must meet the following criteria: clinically severe
chosocial aspects of myocardial revascularization. The angina or objective evidence of ischemia that requires
economic and quality-of-life consequences of PTCA the need for a revascularization procedure, angio-
and CABG strategies over a 5-year follow-up period graphically documented multivessel coronary dis-
will be compared. BARI will provide estimates of ease, suitability for both PTCA and CABG, and
initial and continuing indirect and direct costs of the informed consent for random assignment.
two procedures and provide measures on quality of
life. These critical data are collected in-depth at seven Patient Screening for Exclusion Criteria
participating BARI centers in an ancillary study of The population that is considered for BARI con-
economics and quality of life (SEQOL). This study is sists of those patients who undergo diagnostic coro-
funded by the Robert Wood Johnson Foundation, nary arteriography in a BARI institution and may
having had initial funding from Advanced Catheter include patients with off-site angiograms.
Systems. In addition to SEQOL, data on the number Exclusions from screening. Patients are first evalu-
of significant cardiac hospitalizations, employment ated for entry on the screening log. A patient is
status, and limitations of activities are collected for all considered eligible for screening and is placed on the
BARI patients. (Also see resource use and quality of screening log, allowing the patient to be tracked
life [p V-8].) through the remainder of the screening system if the
Although BARI is designed for the overall com- patient has none of the following exclusions: absence
parison of PTCA versus CABG as the initial strategy, of significant coronary disease, primary congenital
it will also provide comparative data for predeter- heart disease, primary valvular heart disease, primary
mined subgroups, which are defined by the various myocardial heart disease (including patients with a
clinical and angiographic presentations. Clinical sub- ventricular aneurysm, which requires surgery), prior
groups of special interest are those with unstable PTCA or CABG, single-vessel CAD, and/or an age
angina, stable Canadian Cardiovascular Society Clas- .80 years.
sification functional class III or IV angina, or class I Clinical and major angiographic exclusions. Patients
or II angina in the presence of either documented placed on the screening log are evaluated for clinical
ischemia or recent Q wave MI. Angiographic sub- and major angiographic exclusions. If none of the
following exclusions are present, the patient is con-
Definitions for terms printed in bold italics may be found in the sidered clinically eligible: age <17 years, geographi-
Glossary; see Appendix 2. cally inaccessible or unable to return for follow-up,
 The BARI Protocol V-3

No CAD technical expertise and considerations of patient safety.

Sgle VD On the basis of angiographic findings, patients are
Prior Procedure
Congenita Heart Disease excluded at this point if the patient is judged to be
Valvular Heat Die unsuitable for PTCA and/or CABG. (The criteria for
Prmary Myocardial Disease this decision are described on pp V-3-V-6, PTCA and
Age280 years CABG guidelines, respectively.)
Because the BARI investigators recognize the
limitations of trying to prospectively describe all the
Left Main Disease features that define suitability for PTCA and CABG,
Insudfcent Symptoms
Emergency Revasc. the final evaluation for eligibility involves the subjec-
Other tive judgment of both the surgeon and the angio-
plasty operator. This results in a certain degree of
diversity across centers, an additional strength of the
trial that allows the results to be applicable to the
broad group of patients considered for angioplasty.
Baseline data collected on all patients will allow
BARI to characterize in detail angiographic findings
of patients considered clinically eligible for the trial.
The population excluded at the final angiographic
level will be studied by using a 5-10% random
sample that is selected for inclusion in the registry.
Informed consent. Eligible patients who sign an
informed consent for random assignment are entered
in the clinical trial. Those who refuse random assign-
ment but sign an informed consent for follow-up are
entered in the registry. Those who refuse follow-up
are recorded, but no further data are collected.
Prototype consent forms for randomly assigned
and registry patients are shown in Appendix 8. Local
FIGURE 1. Flowchart showing patient screening mechanism. institutional review boards may prefer a modification
A full list of exclusion criteria appears within the text of the of the consent forms to provide additional informa-
Protocol (seepp V-2-V-3, Patient Selection). *Other screening tion. Any local change in the consent forms must
log exclusions include noncardiac illness expected to limit meet Public Health Service requirements.
survival, age <17, geographic inaccessibility, clinical contrain- Random assignment of patients. The Coordinating
dications to PTCA or CABG, inability to understand or Center (CC) prepared the sequence of random as-
cooperate with protocol, enrolled in competing study, techni- signment of patients to the treatment groups before
cally unsatisfactory angiogram, extensive ascending aortic their enrollment. Random assignment was stratified
calcification, primary coronary spasm, concomitant major by clinical site, and within each clinical stratum,
surgery required, and pregnancy. blocks of varying length were used. The sequence of
random assignment was verified and incorporated
insufficient angina or objective evidence of isch- into the BARI computer system for each clinical site.
emia, unstable angina or acute MI, which requires Details of the design and implementation of random
emergency revascularization, left main stenosis assignment are presented in Appendix 4.
>50% or of a character that precludes angioplasty, Participation in other trials. Randomly assigned
noncardiac illness that is expected to limit survival, BARI patients may not participate in any other
extensive ascending aortic calcification, primary clinical trial while they are participating in BARI.
coronary spasm, inability to understand or cooper- However, registry patients are free to participate in
ate with protocol requirements, coronary angio- other studies. BARI clinical sites may participate in
gram that is technically unsatisfactory, suspected or other studies for which BARI-eligible patients are
known pregnancy, enrollment in a competing clini- also eligible, but they must continue to meet their
cal trial, contraindication to CABG or PTCA be- commitment to-randomly assign an adequate number
cause of a coexisting clinical condition, and/or of BARI patients.
concomitant major surgery that is required (e.g.,
aortic and/or mitral valve surgery, carotid endarter- Angioplasty Guidelines
ectomy, and/or resection of left ventricular or ab- Criteria for Acceptable Candidacy to Percutaneous
dominal aortic aneurysm). Transluminal Coronary Angioplasty
Final angiographic exclusion. Clinically eligible pa- To be suitable for PTCA, a patient must have
tients are evaluated for final angiographic eligibility. anatomic characteristics associated with a reasonable
The surgeon and angioplasty operator assess the pa- probability of successful balloon dilatation. The ar-
tient's suitability for each procedure according to their teries that are patent but significantly narrowed
V-4 Supplement V Circulation Vol 84, No 6 December 1991

should be able to sustain prolonged occlusion with- rin (unless contraindicated) and other medication
out the development of cardiogenic shock. Although deemed appropriate for the clinical status of the
the potential for complete revascularization is not a patient. Each patient should be fully heparinized
requirement for entry of patients in BARI, there during the procedure.
must be a reasonable probability for successful relief PTCA should be performed in a cardiac catheter-
of the major stenoses presumed to be contributing to ization laboratory that is capable of providing high-
active myocardial ischemia. These criteria are judged quality video images with immediate replay (includ-
by a BARI-certified angioplasty operator before ran- ing 4- or 5-in. image-intensifier modes), biplane
dom assignment. Reasons for excluding patients are imaging or rapidly available orthogonal single-plane
documented on the Angiographic Exclusion form images, compound angulated projections, hemody-
and include the following: PTCA of (a) vessel(s) namic monitoring, and high-quality film processing.
responsible for ischemia is unlikely to be successful A full range of commercially available guiding and
because of excessive tortuosity of vessels proximal to dilating catheters and guide wires should be avail-
able. PTCA may be performed by either the brachial
the lesion, excessive angulation within the lesion, or femoral approach.
excessive lesion length, total chronic occlusion, or Each procedure begins with a coronary cine angio-
inability to dilate because of excessive calcification; gram of the vessels to be dilated. At least two scout
and PTCA would be excessively dangerous because projections of each vessel are obtained. For each
abrupt closure is likely and would result in cardio- targeted lesion, PTCA is attempted with the goal of
genic shock, or a major side branch cannot be achieving <50% residual stenosis and normal TIMI
adequately protected. (grade 3) distal flow*.2 A PTCA procedure is consid-
Guidelines for Strategy in Percutaneous Transluminal ered completed when the patient is removed from
Coronary Angioplasty the cath lab table.
Considerations of patient safety or logistics may
Once a patient has been randomly assigned to require that the initial PTCA procedure be per-
PTCA, the procedure must be performed within 2 formed over more than one session. If this approach
weeks. Clinical information derived from the medical is used the decision to do so must be made by the end
history, physical examination, ECG, and noninvasive of the first procedure, and all subsequent procedures
stress testing, coupled with the results of coronary should be performed within 2 weeks after the first.
and left ventricular cine angiography, determine the After PTCA cine angiography of treated arteries
strategy to be used for each patient assigned to should be repeated in the same projections as ini-
PTCA. The aim of the procedure is to maximize the tially used. Additional projections may be acquired as
effectiveness of PTCA in relieving ischemia and needed.
minimizing the risk of procedure-related untoward Heparin should be continued for 24-48 hours after
events. A lesion may be targeted for PTCA if all of the procedure in patients whose PTCA was per-
the following conditions are met: a stenosis repre- formed for total occlusion or was associated with
sents a 50% or greater diameter reduction by caliper lesion dissection, thrombosis, transient occlusion, or
measurement, the normal vessel diameter adjacent to in whom distal embolization was observed. Oral
the site of stenosis is > 1.5 mm, and the vessel calcium antagonist therapy should be continued for 4
supplies a sizable region of viable myocardium. Non- weeks, and a regimen of one aspirin tablet per day
significant lesions, lesions located distally or in small should be indefinitely continued.
arteries, and lesions in arteries that supply areas of In the case of abrupt closure, every effort should be
infarction are not routinely dilated. For each patient made to reestablish patency and flow and avoid MI.
a hierarchy of lesion priority is set in such a way that This effort may include emergency CABG or the use
PTCA is attempted first in lesions that are most likely of new technology devices.
to be responsible for the patient's ischemia. Before
revascularization of a patient in the study, the clinical Complications
importance and suitability of each lesion for PTCA is Patients are monitored for adverse events through-
categorized, and a treatment plan is specified by a out the hospitalization. These complications and
BARI-certified operator. These data will be used to their definitions are listed in Appendix 5. The need
define patient subsets for analysis of PTCA outcome for additional revascularization procedures, includ-
and to assess operator performance. ing emergency CABG and repeat PTCA for abrupt
Guidelines for the Percutaneous Transluminal reclosure, are also recorded. Each lesion that is
Coronary Angioplasty Procedure subjected to PTCA is assessed for the occurrence of
dissection and acute closure.
For randomly assigned patients, all PTCAs should
be performed by a certified BARI angioplasty oper- Guidelines for Repeat Percutaneous Transluminal
ator, and the initial PTCA must be performed by a Coronary Angioplasty
certified BARI operator. An experienced catheter- Patients who have an initially successful PTCA
ization laboratory staff should assist, and backup may undergo repeat PTCA when anatomic and clin-
cardiac surgical support must be immediately avail-
able. Preprocedure medication should include aspi- *TIMI flow criteria are given in the Glossary; see Appendix 2.
 The BARI Protocol V-5

ical circumstances are judged suitable. This may dence that the initial or repeat PTCA has been
occur under two circumstances: restenosis of a suc- unsuccessful. If PTCA would be particularly difficult
cessfully dilated coronary artery that is associated or associated with an increased risk of untoward
with clinical manifestations, either by recurrent an- events, CABG may be considered without repeat
gina or significant ischemia documented by objective PTCA. 5) Subsequent CABG may be necessary for
measures (see p V-8); and recurrent ischemia for the recurrence of symptoms or ischemia resulting
which repeat angiography indicates the development from the development of new CAD. Consideration
of new significant CAD that is responsible for the should be given first to repeat PTCA. If repeat PTCA
ischemia and amenable to PTCA. is judged to be inappropriate or not feasible, then
When PTCA is repeated, the procedure should be CABG should be considered if the symptoms are
performed by a BARI operator according to protocol disabling despite optimal medical therapy or if severe
guidelines. Indications for repeat PTCA will be moni- ischemia is documented.
tored carefully, particularly for patients participating in
a study of the 1-year follow-up angiography. Repeat Angioplasty Operator Certification
PTCA must be performed in accordance with BARI Criteria for certification include participation as an
indications for such procedures. Deviations are consid- independent operator in more than 300 elective
ered protocol violations and require defense and justi- PTCA procedures, of which at least 100 were multi-
fication by the responsible BARI investigators. vessel disease cases; demonstration of a success rate
Patients that are randomly assigned to CABG who per lesion of 85% or greater for subtotal lesions
have recurrent ischemia that is associated with by- among the last 100 cases; overall incidence per
pass conduit or native artery stenosis or occlusion for patient of PTCA-related acute myocardial infarction
which PTCA is deemed desirable should have PTCA or emergency CABG of 5% or less; and an overall
performed by a certified BARI operator in accor- mortality rate of 2% or less for elective PTCA
dance with the BARI protocol. patients.
To complete certification requirements, each BARI
Guidelines for Subsequent Coronary Artery Bypass angioplasty operator submits to the Central Radio-
Graft Surgery graphic Laboratory (CRL) preprocedure and postpro-
It is possible that patients who are randomly cedure films of five consecutive PTCA procedures
assigned to PTCA will have CABG before, during, or performed on patients with multivessel disease. The
after their initial PTCA procedure. To ensure the Central Radiographic Laboratory evaluates the films
appropriate use of CABG in such patients, the and determines the quality of the procedures, request-
following guidelines are recommended. 1) Once the ing additional films as needed. In-depth information
patient is randomly assigned, PTCA must be per- about the CRL may be found in Appendix F. The
formed within 2 weeks to minimize the time during current BARI certified angioplasty operators at each
which pre-PTCA crossover could occur. 2) During participating clinical site are listed in Appendix 8.
the initial PTCA procedure, patients who experience
closure of an artery that was previously patent but Classification of Outcome of Percutaneous
narrowed may require emergency CABG for relief of Transluminal Coronary Angioplasty
ischemia or infarction if they are refractory to repeat To define what constitutes a successful PTCA
angioplasty or medical therapy. The decision to pro- procedure is complex. One approach will be to assess
ceed with emergency CABG under this circumstance lesion improvement. The following are requirements
should be strictly based on the need to provide for complete lesion improvement: TIMI grade 3 flow,
appropriate patient care. 3) After an initially success- luminal diameter reduced by .20%, and residual
ful PTCA, indications for subsequent elective CABG stenosis of <50% diameter narrowing. If partial, the
are either recurrence or persistence of disabling lesion that is subjected to PTCA has all of the
symptoms that are accompanied by evidence of myo- following features: TIMI grade 3 flow and luminal
cardial ischemia resulting from inadequate or unsat- diameter reduced by >20% but residual stenosis of
isfactory PTCA and anatomy that is judged to be not <50% diameter narrowing. If there is no lesion
unsuitable for repeat PTCA. 4) Subsequent CABG improvement, neither of the above definitions for this
may be necessary for the recurrence of symptoms and improvement has been met.
ischemia after a period of symptomatic relief with Patients who undergo a PTCA procedure in which
evidence of restenosis of a previously successfully each diseased vessel is not dilated but all targeted
dilated coronary artery, as described in the previous vessels are improved is classified as "incompletely
section. Patients who require revascularization after revascularized by intent." Patients in whom PTCA
initial PTCA should first be considered candidates results are partially satisfactory (that is, not all tar-
for repeat PTCA. PTCA may be repeated more than geted vessels are improved) are classified as "incom-
once if there is a repetitive recurrence of symptoms pletely revascularized but not by intent."
or severe ischemia as defined. The decision to pro-
ceed with CABG in such patients should be based on Lesion Classification
the presence of angina or ischemia of sufficient In the analysis of PTCA outcome, it will be impor-
severity to warrant surgery in the presence of evi- tant to describe the characteristics of the lesions that
V-6 Supplement V Circulation Vol 84, No 6 December 1991

are targeted for treatment. The lesion classification extreme aortic calcification; and disease severe
system developed for use in BARI is described enough to warrant surgery.
below. Although the patient must meet the entry angio-
1) BARI class A-A lesion is considered to be graphic criteria for multivessel disease, the bypass of
class A if it exhibits all of the following characteristics additional arteries with only 50-60% luminal diam-
and has no class B or C characteristics: discreteness eter narrowing by visual assessment in a patient
with critical narrowing of <10 mm in length, vessel randomly assigned to CABG is permissible.
diameter adjacent to the site of stenosis of > 1.5 mm, Guidelines for Operative Management
lesion accessible and not excessively tortuous, subto-
tal occlusion, and concentricity with smooth borders. Once the patient is randomly assigned to CABG,
2) BARI class B - A lesion is considered to be class the procedure must be performed within 2 weeks.
B if it exhibits at least one of the following charac- Rigid control of all aspects of the management of the
teristics but none of those of class C: discreteness patient during and after surgery is not possible, but
the surgery form documents the techniques and
with critical narrowing 10-20 mm in length, recent methods used in the surgical management of these
(within 3 months) total occlusion, moderate vessel patients. Anesthetic techniques are not standardized.
tortuosity proximal to the lesion, irregular borders, Cold potassium cardioplegia (either crystalloid or
ostial location, significant calcification, lesion in bi- blood) is the protection of choice; however, cold
furcation, moderate vessel angulation within lesion, ischemic arrest may be routinely preferred by some
thrombus, ulceration, and/or eccentricity. surgeons or in special situations, and this decision is
3) BARI class C -A lesion is considered to be class left to the individual surgeon's judgment.
C if it exhibits any of the following characteristics: The internal mammary artery should be used for
excessive vessel tortuosity proximal to the lesion or revascularization of the left anterior descending coro-
excessive vessel angulation at its site, chronic (>3 nary artery whenever feasible. The choice of conduit
months) total occlusion (TIMI grade 0) or an un- for revascularization of other arteries depends on the
known period of total occlusion, critical narrowing of experience and judgment of the surgeon. Details of the
>20 mm in length, and/or inability to protect major cannulation technique, methods of myocardial preser-
side branches. vation, grafts used, aortic cross clamping, duration of
New Technology Devices cardiopulmonary bypass, perioperative medications,
and patient status are recorded.
The New Technology Committee monitors the Preoperative and postoperative use of antiplatelet
development of new techniques and devices such as drugs is recommended. For elective procedures 100
stents, cardiopulmonary support system, atherec- mg dipyridamole four times daily for the 2 days
tomy, and laser and recommends if and under what preceding CABG may be used, with daily aspirin
circumstances they can be used in BARI. No new intake after operation when deemed reasonable. In
devices, neither those of Investigative Device Exemp- the absence of evidence that long-term dipyridamole
tion (IDE) nor those of new technology that are therapy is essential after CABG, aspirin not exceed-
federally approved, are to be used as an initial ing 325 mg/day will be acceptable unless there are
strategy in randomly assigned patients. The devices contraindications to the use of antiplatelet drugs.
can be used in clinical situations such as abrupt
closure with hemodynamic compromise in which, Complications
based on local experience and judgment, the tech- Patients are monitored for adverse events through-
nique is in the best interest of the patient. This out their hospitalization. These complications and
restriction does not apply to registry patients. After their definitions are listed in Appendix 5.
completion of the initial single or planned staged
PTCA procedure, new technology devices may be Guidelines for Repeat Coronary Artery Bypass Graft
used if additional coronary interventions are re- Surgery or Subsequent Percutaneous Transluminal
quired. Coronary Angioplasty
Surgical Guidelines Repeat CABG should be considered on either the
recurrence of significant clinical manifestations that
Criteria for Acceptable Candidacy for Surgery suggest a need for further invasive therapy to relieve
Before random assignment, a BARI-certified sur- myocardial ischemia or on discovery of evidence of
geon must deem the patient suitable for CABG. profound ischemia with exercise testing (see Inclusion
Specifically, patients must have the following charac- Criteria, p V-2). Final decisions regarding the appro-
teristics: target vessels of an adequate size for inser- priateness of CABG or subsequent PTCA in patients
tion of a bypass graft (i.e., luminal diameter of >1 who have had prior CABG are based on angiographic
mm in all arteries to be bypassed); satisfactory distal evidence of graft narrowing or progressive and severe
runoff; no severe diffuse atherosclerotic involvement atherosclerosis in ungrafted vessels to a narrowing of
of distal coronary arteries including the absence of the luminal diameter of .50% by caliper measure-
multiple discrete severe lesions throughout the ment. Patients who have severe lesions in ungrafted
course of the artery to be bypassed; absence of coronary arteries or bypass grafts may be considered for
 The BARI Protocol V-7

Radom Initial Revasculari

AMgnt Completed
Tebphon Tebphone Tdephone
Conat Contact' C
tact
-_ 2-ek_~
Widow
Requied
BY
FIGURE 2. Follow-up chart for
randomly assigned patients. Patient
follow-up is described more fully on
0 2wks 4wks l4wks 6rnos 1 yr 2yrs 3yrs 4yrs_ 5yrs p V-7. Follow-up for regitry pa-
a a
tients occurs at identical intervals
except that the 4-14-week contact is
-_ Wmndow- omitted, and all follow-up is done
for
Iniial by telephone contact.
aui.c
visit
ainic vl VWjt b
ainiC Clinic Viit

a) Telephone Contact + ECG requested from physicn

b) Cinhc Vsit: Resing ECG, Exercse Test, Upid Levels

PTCA if recurrent symptoms or signs of profound performed at BARI institutions according to the
ischemia are thought to be related to lesions in vessels guidelines.
that are amenable to PTCA. Patients in BARI are followed for a minimum of 5
years from the time of entry into the randomly assigned
Surgeon Certification trial or registry. During this follow-up period nine
Each surgeon who performs CABG on patients major end points are ascertained: these are mortality,
who are randomly assigned in the BARI trial will MI, angina/chest pain, myocardial ischemia, subse-
require initial certification that is based on the fol- quent revascularization, resource use, quality of life,
lowing criteria: practice as an attending staff surgeon angiographic characteristics at 5 years, and left ventric-
for 3 years or more, majority of practice devoted to ular function at 5 years.
coronary artery surgery, most recent 100 consecutive Scheduled Follow-up
primary, elective, isolated CABG operations with a
mortality rate of no more than 2% (death within 30 Scheduled follow-up points are at 4-14 weeks
days of procedure) and an MI rate of no more than (randomly assigned patients only); at 6 months; and
4% (new Q waves within 30 days of procedure), at 1, 2, 3, 4, and 5 years after study entry. Randomly
performance as principal surgeon of 100 or more assigned patients alternate between clinic visits (4-14
CABGs with internal mammary artery grafts, and the weeks and 1, 3, and 5 years) and telephone contact (6
principal BARI cardiac surgeon at the participating months and 2 and 4 years), and registry patients have
BARI Clinical Unit is satisfied that the judgment, telephone contacts only (see Figure 2). During the
technical performance, results, and care after the clinic visit, randomly assigned patients undergo ECG
and exercise treadmill tests, and blood is drawn for
procedure meet current standards of the institution's fasting-state serum lipid levels. When the follow-up is
Department of Surgery. by telephone, an ECG is requested from the primary
Surgeon Participation physician. Data collected for the randomly assigned
and registry patients are identical with the exception
BARI surgeons are an integral part of the study. of the early follow-up, lipid levels, and exercise
They serve as representatives on all working commit- treadmill tests.
tees and share responsibility with the cardiologists Information that is collected at each contact in-
for adherence to protocol and policy decisions. The cludes symptomatic status, health behavior (diet,
surgeons who are currently participating in the BARI exercise, etc.), and quality of life. If a follow-up
trial are listed in Appendix 8. contact is not possible within the specified time
window, the evaluation takes place thereafter as soon
Patient Follow-up and End Point Ascertainment as possible. During each scheduled follow-up, events
Effective follow-up depends on the relationship that require data collection to document end points
developed by the BARI staff with referring physi- are identified.
cians and patients. All BARI centers make an effort
to actively involve referring physicians in the study. Ascertainment of End Points
This helps to ensure that therapy guidelines are At each follow-up contact all hospitalizations that
followed and that subsequent revascularization is occurred since the last contact are identified. A
V-8 Supplement V Circulation Vol 84, No 6 December 1991

Hospital Course form is required for the following dure, and Hospital Course forms, and preprocedure
conditions: cardiac hospitalization for 3 days or and postprocedure ECGs. Angiography in asymptom-
longer, cardiac procedure (coronary angiography, atic patients is discouraged except for the scheduled
PTCA, CABG), hospitalization of any length for angiogram at 5 years and the 1-year angiogram in
cardiac arrest, and hospitalization of any length for a patients enrolled in the 1-year angiographic study.
condition thought to be a complication of a revascu- No patient should undergo repeat revascularization
larization procedure. on the basis of anatomic findings alone. To be eligible
Data from the Follow-up and Hospital Course for angiography during follow-up, a BARI patient
forms are used to identify study end points. For each must have clinically severe angina, ischemia, or in-
end point additional data are collected as outlined farction as described by at least one of the following
below. criteria: unstable angina, CCSC III or IV angina,
1) Mortality-when a patient is identified as de- CCSC I or II angina and an ejection fraction of
ceased, the CC must be notified within 3 days by <50%, MI (Q wave or non-Q wave) after the initial
means of a Notification of Major Event form. In
addition, the following information must be sent to procedure, and/or severe ischemia on noninvasive test-
the CC: a mortality data form, a narrative report ing after initial revascularization. After angiography,
from the Principal Investigator that describes the the decision to perform a repeat revascularization
events surrounding the death and that is signed by procedure is left to the discretion of the investigator
the BARI surgeon and/or PTCA operator involved and the referring physician; however, the guidelines
with the case, a death certificate, a coroner's report for subsequent revascularization (given in detail on
(if autopsy is done), a cath lab report (if the patient pp V-3-V-7) must be followed in randomly assigned
died within 30 days of PTCA), a surgical report (if patients, and revascularization should be performed
the patient died within 30 days of CABG), any ECGs by a BARI surgeon or angioplasty operator.
performed within 24 hours of death, and any cardiac 6) Resource use-resource use is calculated by
enzymes drawn within 24 hours of death. This infor- using the mean number of significant cardiac hospi-
mation is used by the Morbidity and Mortality Classi- talizations. SEQOL is collecting bills from initial and
fication Committee (MMCC) to categorize each death. follow-up hospitalizations as well as data concerning
2) MI-when MI is suspected, all ECGs that are follow-up outpatient visits, tests, procedures, and
recorded within 3 days after the event are sent to the medications. Data are also being collected concern-
Central ECG Laboratory (in-depth information ing the wider economic impact of the two procedures,
about the Central Electrocardiographic Laboratory particularly on employment. These data will be ana-
[CEL] may be found in Appendix 7), and a Suspected lyzed to compare the profile of long-term costs and
MI report form is required. These data are used to benefits of PTCA and CABG.
determine whether the event was an MI. When data 7) Quality of life -at baseline and at each follow-up
are missing or inconsistent, the MMCC reviews all point, quality of life is assessed by using a brief ques-
available data to determine whether an MI occurred. tionnaire. Data on employment, self-care, social life,
MIs that do not result in hospitalization will be home life, sex life, and leisure activities are included.
identified through the yearly scheduled ECGs. The SEQOL ancillary study collects data on an array of
3) Angina -angina, which is classified as stable or quality-of-life measures at baseline and during follow-
unstable, is assessed at scheduled follow-up points up, including functional status, emotional health,
and during hospitalization. Stable angina is classified health concerns, and social functioning.
according to the Canadian Cardiovascular Society 8) Angiographic assessment - the extent of revas-
Classification (CCSC). Unstable angina is classified cularization will be angiographically assessed at the
according to the presence or absence of acceleration 5-year follow-up. The following criteria will be eval-
(increased frequency, severity, duration), pain that uated: the condition of CABG revascularization con-
lasts 20 minutes or longer and is associated with duits (stenosis, disease, patency); condition of dilated
negative cardiac enzymes, pain that continues within segments; revascularization status (e.g., global left
14 days after infarction, transient ECG changes, and ventricular coronary perfusion and regional perfu-
pain at rest. sion in relation to baseline stenoses of a >50%
4) Myocardial ischemia -exercise treadmill testing reduction in luminal diameter); progression of pre-
will be used to assess ECG evidence of myocardial existing lesions or development of new lesions in
ischemia in randomly assigned patients only. An native vessels that affects revascularization status;
exercise ECG is performed at 4-14 weeks and 1, 3, and adequacy of revascularization, as indicated by
and 5 years. Additional exercise ECGs are required integration of regional coronary distribution with
when myocardial ischemia is a reason for subsequent regional wall motion.
revascularization. All exercise ECGs are evaluated 9) Left ventricular function -left ventricular ejec-
by the CEL for a myocardial ischemic response. tion fraction will be angiographically obtained at
5) Subsequent revascularization-all revascular- baseline and at 5 years. In the event that a contrast
ization procedures subsequent to the initial revascu- left ventriculogram is not obtained or is technically
larization require data collection forms, including inadequate, a radionuclide angiogram will be re-
Repeat Revascularization, PTCA or CABG Proce- quired as a surrogate measurement.
 The BARI Protocol V-9

Concomitant Therapy Guidelines tional Cholesterol Education Program.8 Surgery or a

Introduction recent MI may temporarily lower the serum cholesterol
Patients who are enrolled in BARI must receive value, and this must be taken into account when
careful attention to concomitant medical therapy, evaluating lipid values.
regardless of assignment to PTCA or CABG. It is Additional Concomitant Therapy
important to maintain comparability of concomitant All patients should be indefinitely maintained on
therapy for the two treatment strategies to avoid bias in antiplatelet therapy with one aspirin per day unless
the comparison between the strategies. Although most the drug is contraindicated or not well tolerated. In
patients are followed primarily by their local physicians, the latter case, alternative therapy with other anti-
the BARI investigator must establish an appropriate platelet drugs may be considered. Recurrent myocar-
relationship with the patient and referring physician to dial ischemia should be treated with nitrates, calcium
ensure careful attention to concomitant medical ther-
apy according to the guidelines outlined below.
channel blockers, and /3-blockers according to usual
clinical practice.
Risk Factor Modification Data Analysis
Risk factor modification will be initiated for all Introduction
patients after enrollment in the study. Because all
randomly assigned patients will be seen at the 4-14 BARI compares the safety and efficacy of the PTCA
week follow-up, it is the responsibility of the BARI and CABG treatment strategies rather than the treat-
investigators to initiate risk factor modification if it ments themselves. All primary analyses will be con-
has not already been undertaken at the time of ducted by intention to treat regardless of what treat-
revascularization. ment or treatments patients received. Comparisons of
Behavior modification. Instruction for behavior treatment strategy will be made not only for the entire
modification in the areas of smoking, exercise, and randomly assigned trial population but also within
diet is initiated for all patients after enrollment in the relevant subgroups.
study and is reinforced at each follow-up clinic visit
and telephone contact. End Points
Treatment of specific medical problems. Hyperten- To compare end points, 5-year incidence rates and
sion-treatment may include common medications confidence limits will be calculated for the difference
that are used after PTCA and CABG (calcium in rates between the initial CABG and initial PTCA
channel blockers, /3-blockers) as a first-line therapy strategies. The primary study end point is mortality at
in an attempt to simplify medical therapy. Angioten- 5 years. Other major end points include MI, angina/
sin converting enzyme (ACE) inhibitors, which are chest pain, myocardial ischemic response, subse-
used as a first-line antihypertensive therapy in clini- quent revascularization, resource use, quality of life,
cal practice, may also be used as indicated. The and measures of angiographic characteristics and left
criterion for treatment is a resting systolic blood
pressure of >160 mm Hg or a diastolic blood pres-
ventricular function at 5 years.
sure of >90 mm Hg on two measurements. The goal Composite End Points
of therapy is to maintain a blood pressure of < 140/90
mm Hg. Diabetes this will be treated with diet, oral
-
To simultaneously analyze several outcomes, com-
hypoglycemic agents, and insulin, as clinically indi- posite end points will be created. For example, angina-
cated. When control is poor, consultation with a free and event-free survival are best viewed in a hier-
diabetologist should be obtained. Lipid abnormali- archical manner, which has been used to express
ties -total cholesterol, HDL, LDL, and triglycerides outcome in the National Heart, Lung, and Blood
are obtained at baseline, 4-14 weeks, and annually Institute PTCA Registry. When assessing long-term
thereafter. Baseline diet histories are encouraged symptomatic relief, subsequent procedures and phar-
though not recorded for BARI. Patients are in- macological intervention must be taken into account.
structed in an appropriate diet program based on This will be done in a cross-sectional analysis. When
individual patient profile. Diet therapy remains an given sufficient intervention, it is expected that most
integral part of the approach to lipid disorders and patients will end up either asymptomatic or mildly
must be maintained regardless of drug regimens. symptomatic at follow-up. The comparison of the two
Every reasonable effort should be made to achieve treatments will be based on the number and types of
the following target lipid levels: total cholesterol less procedures and the amount of medication that is
than 180 mg/dl, HDL cholesterol greater than 45 required to achieve asymptomatic status. The distribu-
mg/dl, LDL cholesterol less than 110 mg/dl, triglyc- tion of the most severe form of intervention is expected
erides less than 120 mg/dl, and a total cholesterol/ to be inherently different between the two treatment
HDL ratio of less than 4.5. strategies. The CABG group undergoes a more exten-
Patients who do not sufficiently respond to diet and sive and invasive technique initially, whereas the PTCA
exercise programs should begin lipid-lowering drug strategy group is expected to have a greater number of
therapy, following the algorithm proposed by the Na- subsequent procedures during follow-up.
V-10 Supplement V Circulation Vol 84, No 6 December 1991

Subgroups be separately analyzed for ischemia at follow-up.

As much as possible, we will limit formal analyses to This analysis (in effect two subgroup analyses) would
those subgroups that have been specified a priori. It describe the relative efficacy of the two treatments
should be noted that because of the smaller sample for relief of the specific symptoms that the original
sizes, these subgroup analyses will have limited power. strategy was intended to relieve. Alternatively, the
subgroups can be combined and analyzed for the
Multivariate models that include treatment by covari- recurrence of any symptom.
ate interaction will be used to see whether the treat- Observations at multiple time points will be han-
ments differentially affect outcome in various clinical dled by various approaches. One method will simply
subgroups. Important baseline factors that define sub- be to assess the outcome at final follow-up. Another
groups include the following: symptomatic status (an- approach is to average the severity of response over
gina/ischemia pattern), left ventricular function, extent multiple observations during follow-up. Profile analy-
of myocardial ischemia, and angiographic risk (e.g., the sis, generalized linear models, and survival models
presence of a class III lesion or the amount of myocar- that allow inclusion of time-dependent covariates
dium at risk). Some subgroups might also be created on (e.g., events, medications) may be appropriate to
the basis of exploratory statistical analyses. Analysis of examine differences between treatments over time in
a multitude of such subgroups creates a multiple- an exploratory manner.
comparison problem that will be acknowledged. Further cross-sectional analyses include the follow-
ing examples. Exercise testing - exercise test results will
Methods ofAnalysis be compared at each time point by treatment. The
Methods of analysis can be classified by whether association of exercise test results and subsequent
data are collected cross sectionally (e.g., 5-year ven- revascularization procedures will be examined. Return
triculography), longitudinally (e.g., angina status at to work - employment status before and after treat-
each follow-up point), or the time to event (e.g., time ment will be cross tabulated. The interpretation of such
from baseline to death). data will take into account the fact that a large portion
Analysis of baseline data. Comparability of the two of the BARI population will be men and women near
treatment groups is expected because of the process retirement age who will not remain in the active labor
of random assignment; nevertheless, distributions of force, regardless of the success of the procedure. Cost
baseline characteristics within the two groups will be analysis - a major ancillary study will be devoted to the
examined. The following set of key baseline charac- thorough evaluation of cost data generated from BARI
teristics will be defined: known risk factors for car- by seven clinical centers. Analysis of cost data in the
diac mortality (e.g., history of congestive heart fail- main study will be limited to the comparison of the
ure, previous MI, history of diabetes, and history of number of significant cardiac hospitalizations for each
hypertension); usual coronary risk factors (lipid lev- of the two initial strategies. Follow-up angiographic
els, smoking status, and blood pressure levels); an- analyses-the angiographic follow-up data, collected at
giographic risk factors (severity of coronary artery 1 year at selected centers and at 5 years at all centers -
disease, impaired left ventricular function); duration will provide information in the following areas: graft
and nature of symptoms, functional capacity, medi- patency, patency of dilated segments, and progression!
cation use, and results of baseline ECG; and demo- regression of the atherosclerotic process in both treated
graphic factors (age, sex, race, education, and socio- and untreated segments. The distribution of changes in
economic status). these areas between baseline and follow-up angiogra-
Suitable techniques will be used to test for compara- phy will be compared by treatment group. A myocardial
bility of the distribution of these variables between the jeopardy index, which is computed by the CRL from
two treatment groups. Results of these analyses will be the angiography readings, will allow overall assessment
used to describe the study population and to determine of revascularization results.
any imbalance for which adjustment must be made in
the final analysis of treatment comparisons. Interim Data Analysis
Time-to-event analyses. Standard life table tech- End point data are not revealed to clinical inves-
niques will be used to analyze mortality and MI. tigators during the course of the study but are
All-cause mortality and cardiac mortality will be presented at semiannual meetings of the Safety and
separately assessed. Cumulative event rates for sur- Data Monitoring Board (SDMB). First and foremost,
vival and MI-free survival will be compared between mortality and MI rates by treatment group are pre-
the two treatment groups. sented. These rates are expected to be relatively low
Cross-sectional and longitudinal analyses. Some im- and similar in the two treatment groups. Procedural
portant BARI end points, such as the status of angina complications are reported for each treatment group,
and myocardial ischemic response, will be measured as are rates of subsequent procedures during follow-
and analyzed cross sectionally. up. Should one of the treatment groups experience
Some patients originally present with angina and significantly higher rates of mortality or MI, the
others are asymptomatic with ischemic manifestation SDMB would consider either early termination of the
at entry. Patients with angina can be analyzed for the trial or modification of the protocol to exclude pa-
recurrence of angina, whereas ischemic patients can tients in certain subgroups.
 The BARI Protocol V-11

Registry Analyses treatment can be defined by the management re-

Eligible but not randomly assigned patients. It is as- ceived within 2 weeks of BARI ID assignment, or
sumed that the BARI clinical trial results would hold alternatively within 3 months. The 2-week rule is
true in a larger population of patients whose disease based on the requirement that the assigned treat-
characteristics are similar to those of the BARI pa- ment be delivered within 2 weeks in the trial.
tients. To test this assumption, outcomes from the Quality Control/Reproducibility of Measurements
randomly assigned trial will also be studied in BARI- To establish the reproducibility of key measure-
eligible patients who do not consent to random assign- ments in BARI, replicate blinded assessments will be
ment. If after adjustments for differences between the performed on randomly selecited samples. Measures
randomly assigned and the eligible but not randomly in this category include angiographic baseline char-
assigned groups the long-term outcome by treatment is acteristics, measure of PTCA outcome, occurrence of
similar, then the trial results can be generalized to MI, cause of death, and functional test outcome.
patients who meet BARI criteria. (Details for some of these analyses are described in
In spite of careful statistical adjustments to maxi- the Central Laboratory sections located in Appen-
mize comparability, the investigation of end points dixes 6 and 7).
among eligible but not randomly assigned patients Reliability will be assessed by appropriate statisti-
will only have the status of a good observational study cal techniques that are applicable to the replication
and will not be suitable for providing primary esti- of measurement design.
mates for treatment comparisons. In particular, in
the registry population treatment is recommended Exploratory Methods of End Point Analysis
based on a variety of selection factors, many of which Completeness of revascularization. With respect to
are subtle and cannot be measured. An additional CABG, it has been claimed that the more complete
problem'to be solved in analyses of such data will be the revascularization the better the results. This
the definition and time of treatment assignment. clinically appealing statement has never been ade-
As is typical for observational studies, registry quately tested because no one would randomly assign
analyses will also explore and generate hypotheses patients to complete versus incomplete revasculariza-
regarding the relation of various patient characteris- tion. The BARI trial will indirectly test this hypoth-
tics to outcome (predictors of events). The charac- esis because the PTCA strategy does not necessarily
teristics to be considered will be the key baseline aim for or achieve the completeness of revasculariza-
variables of the randomly assigned study. Periproce- tion that CABG does.
dural outcomes such as mortality, myocardial infarc- Percutaneous transluminal coronary angioplasty as a
tion, and other complications will serve as end points delay for coronary artery bypass graft surgery. An addi-
in these analyses; procedure-related complications tional goal of BARI is to learn the rate and timing of
and outcomes will in turn also serve as risk factors for the crossover from PTCA to CABG. A PTCA patient
long-term events. is a crossover case if he or she undergoes CABG at
Statistical' methods such as linear regression, logis- any time after random assignment either before or
tic regression, and Cox regression will be used to after PTCA. The percentage of patients who cross
obtain adjusted estimates of the effect of prognostic over, as well as the length of delay to CABG, will be
variables on outcome. As in the trial, testing for analyzed.
qualitative interaction between treatments and risk Repeat revascularization. The attribution of events
may be appropriate. in the presence of subsequent revascularization by a
Patients clinically but not angiographically eligible. A treatment other than the one randomly assigned is of
5-10% random sample of patients who meet clinical particular concern in BARI. Although the literature
criteria for the randomly assigned trial but are ex- has described many avenues of analysis in such
cluded on angiographic grounds will be used to assess situations, none of these analyses can be interpreted
the appropriateness of angiographic exclusion for in a classical statistical sense. However, to assess the
BARI. impact of crossover two ancillary methods of analysis
In this sample baseline characteristics as well as are proposed, with the clear understanding that the
in-hospital outcome and major follow-up events will results obtained must be interpreted in the same
be analyzed by the given treatment. Results of these manner as any other observational study results.
analyses will lend insight into the prognostic impor- The first exploratory method is to censor follow-up
tance of those vessel/lesion characteristics that are at the time of the different method of revasculariza-
not otherwise present in BARI-eligible patients. tion. This approach looks at' patients only during the
Determination of time 0. For the randomly assigned continuation of their originally assigned treatment
patients, the clear choice for time 0 is the time of and does not count events after initiation of the other
random assignment. For registry patients, the time of treatment. This method of analysis has less power to
assignment of the BARI identification number is detect differences between the treatment groups
used. Random assignment determines the treatment because many patients may be lost because of a
group in the clinical trial, but a clear-cut treatment subsequent procedure. More importantly, this
assignment does not apply to registry patients. Here, method could also lead to biased results if subse-
V-12 Supplement V Circulation Vol 84, No 6 December 1991

quent procedures occur frequently and in a highly study design, data collection and management, and
selective manner, removing individuals who are ei- analysis of BARI results. The CC staff maintain the
ther at more or less risk than others for the end point protocol and operations manual, design and support
of interest. This possibility must be examined before the data entry and management system, and imple-
such analysis, and great care should be exercised in ment certification and procedures of random assign-
drawing inferences. ment. CC staff are responsible for preparing and
A second auxiliary method is the transitional life distributing regular BARI progress reports and min-
table analysis, in which patients are considered to be utes, preparing reports for the SDMB, monitoring
members of their original treatment group until the end point results, preparing data analyses, and en-
different revascularization procedures and members of suring the quality and accuracy of data collection. CC
the other group thereafter. This approach allows all investigators will ultimately play a key role in the
follow-up data to be used in the analysis, although there preparation of reports for publication.
may again be bias because patients who subsequently Central radiographic laboratory. The BARI CRL is
located at Stanford University, Palo Alto, Calif. The
receive the other treatment may be different from those laboratory is responsible for the receipt, review, and
who remain with their treatment assignment.4 analysis of all BARI radiographic test data, including
These two methods have been useful in previous the assessment of PTCA performance and success,
studies.4 on all randomly assigned patients from each of the
Organizational Structure BARI clinical sites and for transmission of the results
to the CC in a timely fashion. In addition, this
Introduction laboratory is responsible for monitoring the quality,
Investigators in the BARI study collaborate completeness, and timeliness of BARI radiographic
through an organizational structure designed to procedures that are performed at the clinical sites.
maintain the continuity of study operations and to The CRL will also be responsible for reviewing the
facilitate effective communication and cooperation angiograms of the patients who are excluded from
among components. BARI eligibility based on angiography criteria and
who are subsequently selected for the registry. (The
Operational Components functions of the CRL are discussed further in Ap-
Clinical centers. Fourteen primary clinical centers pendix 6.)
with one or more hospitals are participating in BARI. Central ECG laboratory. The BARI CEL is located
All are located at major medical centers in the at St. Louis University. It is responsible for quality
United States and Canada. The clinical centers are control and interpretation of all resting and exercise
responsible for the screening and recruitment of ECGs required by the BARI protocol. (The func-
eligible patients, performance of PTCA and CABG, tions of the CEL are discussed further in Appendix
coordination of patient care and follow-up, and col- 7.)
lection of all clinical information and test data that Coordinating Center for the Study of Economics and
are required by the BARI protocol. Individual clini- Quality of Life. The SEQOL coordinating center is
cal sites may also consider interaction with other currently located at Stanford University, which moved
institutions to enhance recruitment; such relation- from Duke University in January of 1990, and is
ships consist of the following specific categories. responsible for the coordination and quality control of
Coinvestigational sites. Each one of these functions data collection at the seven BARI clinical sites that are
as a satellite of an original BARI clinical center and participating in the SEQOL study (Boston University,
uses resources of the parent site. At these sites the Cleveland Clinic Foundation, Duke University, Mayo
same protocol activities as at the associated parent Clinic, St. Louis University, University of Alabama, and
site are performed; however, computer and data University of Michigan). The ancillary study activities
management resources are shared with the parent are supported by a grant from the Robert Wood
site. Investigators at these sites become voting mem- Johnson Foundation.
bers of the Steering Committee if requirements for
protocol and performance criteria (certification of Administrative Components
surgeons, angioplasty operators, and coordinators) Study Chair. The Study Chair, which is appointed
and minimum recruitment levels (3 patients/mo) are by the National Heart, Lung, and Blood Institute
achieved. director, has major responsibility for the scientific
European parallel study. The Institute for Clinical direction of the BARI trial. The Study Chair has
and Experimental Medicine in Prague, Czechoslova- been appointed to serve for the duration of the study
kia, participates in the BARI study through a United unless other arrangements are made by mutual
States-Czechoslovakian scientific agreement. Ran- agreement between the Chair and the National
dom assignment in this parallel study began in Feb- Heart, Lung, and Blood Institute director. In the
ruary of 1989. All data that are collected will be event that the Study Chair is vacated because of
analyzed as an ancillary study to BARI. death, resignation, or inability to serve because of
Coordinating center. The CC at the University of serious illness, the National Heart, Lung, and Blood
Pittsburgh has primary responsibility for the BARI Institute director will appoint a new Chair.
 The BARI Protocol V-13

National Heart, Lung, and Blood Institute Program The group meets semiannually; in addition, at
Office. The National Heart, Lung, and Blood Institute monthly intervals the CC provides data to the Chair
Program Office is in the Cardiac Diseases Branch of of the SDMB to ensure early identification of poten-
the Division of Heart and Vascular Diseases and is tial adverse outcomes of therapy during the recruit-
responsible for the overall direction and monitoring of ment phase.
BARI. The Program Office participates in the general Morbidity and Mortality Classification Committee.
organizational and scientific guidance of the study and The MMCC is an external National Heart, Lung, and
monitors the study progress for the institute. Statistical Blood Institute-appointed group responsible for the
guidance is provided by the National Heart, Lung, and review and classification of MI and death among
Blood Institute. study patients. Classification is based on case sum-
Steering Committee. The Steering Committee is com- maries provided by the CC. The Committee serves
posed of the Study Chair and the Principal Investiga- independent of BARI and clinical and central labo-
tors from each BARI Clinical Site, the Central Labo- ratory investigators.
ratories, the Coordinating Center, and the National
Heart, Lung, and Blood Institute Program Office. This Data Management/Computing
committee provides the scientific direction for the study Overview
and periodically meets to assess progress. The Steering The CC has implemented a distributed data entry
Committee developed the BARI protocol and is re- system for BARI. Each clinical site has been pro-
sponsible for its execution. vided with an IBM PS/28 (a registered trademark of
The following technical subcommittees have been International Business Machines Corp.) microcom-
appointed: Angiography, PTCA, CABG, Design and puter system that is equipped with the hardware and
Analysis, End Points, Forms, Entry Criteria, Con- software necessary for data collection and manage-
comitant Therapy, Economic, and New Technology. ment at the clinical site. Data are transferred weekly
These subcommittees are charged with overseeing from each site to the CC, where they are maintained
specific areas of the BARI protocol. in an aggregate database on the CC's Digital Equip-
Operations Committee. The Operations Committee ment Corporation VAX 6310 computer system (Dig-
is responsible for the daily conduct of BARI as an ital Equipment Corporation, Maynard, Mass.).
extension of the Steering Committee. Weekly confer-
ence calls are conducted to ensure that important Distributed Data Entry and Management
issues are appropriately addressed and problems The BARI distributed data entry system consists of
resolved. It is composed of the Study Chair, the six phases. At the site level, the forms* are entered
Principal Investigator of the Coordinating Center, into the microcomputer and intraform edits per-
representatives from the National Heart, Lung, and formed. Data are then transmitted to the data center
Blood Institute Program Office, Principal Investiga- by means of a telecommunications link and uploaded
tors of the Central Laboratories, and other BARI into data sets on the VAX computer. At this point
participants on an ad hoc basis as required. interform edits are done, and then finally the data are
Policy and Publication Committee. The Policy and appended to the aggregate BARI data base. These
Publication Committee (PPC) has the responsibility phases are discussed in greater detail below and are
of approving ancillary studies and analyses of the also described in the Data Management Manual.
BARI data base beyond the primary baseline and Form entry. The Epidemiology Data Center at the
outcome papers. Principal Investigators as well as University of Pittsburgh has designed and imple-
associate investigators at all BARI clinical and cen- mented a screen-oriented data entry program for the
tral sites are encouraged to develop protocols and IBM microcomputer called the PoP data entry sys-
publish research papers with BARI study data. The tem. In this system the pages of study forms appear
PPC will ensure that these protocols are methodolog- on the microcomputer screen in much the same way
ically sound and that the publication of high-quality as they appear on paper. Data are transcribed from
manuscripts progresses efficiently. paper forms to the corresponding data fields that are
displayed on the microcomputer. During data entry
Independent Components each field is interactively subjected to data type
Safety and Data Monitoring Board. The SDMB is an verification and range checks. Data are further veri-
external review committee appointed by the National fied by using a double entry system. All data must be
Heart, Lung, and Blood Institute that has the respon- entered twice before subsequent system processing
sibility of protecting the scientific integrity of the BARI can occur. Each discrepancy that requires resolution
trial. The Board reviews interim trial results and ad- during this process is recorded in an audit file that
vises the National Heart, Lung, and Blood Institute on contains a full account of all changes made to the
all policy matters. The SDMB includes senior scientists, study data records, the date of each change, and the
cardiologists, surgeons, a biostatistician, and an ethicist. ID number of the operator who makes the changes.
The Study Chair, the Principal Investigator from the
Coordinating Center, and representatives from the *Data Entry Form sets may be ordered at a cost of $50.00 from
National Heart, Lung, and Blood Institute Program the BARI Data Coordinating Center, University of Pittsburgh, 127
Office also participate as nonvoting members. Parran Hall/130 DeSoto Street, Pittsburgh, PA 15261.
V-14 Supplement V Circulation Vol 84, No 6 December 1991

Intraformt edits. After entry and verification, each Coordinating Center Statistical Computing
data record undergoes defined intraform logic and The aggregate study data base is maintained in the
consistency checks. A report that lists the nature of System 1032 DBMS. Files for analysis that use statisti-
each error is generated for resolution by the clinical cal programs such as SAS, SPSS, BMDP, and MINITAB (SAS
site staff. Corrections to the data are then made Institute Inc., Cary, N.C.; SPSS Inc., Chicago, Ill.;
through the PoP update module. BMDP Statistical Software, Los Angeles; and Minitab,
Data transmission. Data records that are free of Inc., State College, Pa; respectively) are easily gener-
errors are flagged for transmission to the CC. An ated. The 1032 DBMS includes a powerful program-
automated data file transfer system retrieves data ming language that allows new variables to be created
records from each clinical center once a week by from existing information. All study analyses are com-
means of modem and telephone lines. This occurs at pleted on the VAX computer by using standard statis-
night while the centers' microcomputers are not in tical packages whenever possible.
use. The CC host computer dials the telephone
number of the clinical center's modem. As a security Data Security
measure, the center's modem, on recognition of the Clinical site microcomputer. Each clinical site is
incoming call, disconnects the line and then redials responsible for confidentiality by appropriately re-
the CC to proceed with file transfer. The entire stricting physical access to the BARI microcomputer,
transaction is recorded in an audit file that is re- which is placed in a locked room after working hours.
viewed the next morning. Unsuccessful transfers are Data backup is forced daily and can also be
identified and resolved during that day. performed more often if desired. The clinical site
Data that are contained in files transmitted to the data manager is responsible for placing backup dis-
CC can no longer be accessed by the clinical center. kettes in a safe storage area in a location remote from
Any subsequent changes to the data must be made the microcomputer. Damage to the hard disk results
through the submission of data base update requests, in the loss of no more than 1 day's work. All
which is discussed below. microcomputer drives are backed up once each
Load data. At the CC data records are transferred month.
to the VAX 6310 computer and loaded into System To gain access to the PoP data entry system, users
1032 Data Base Management System (DBMS) data must go through a log-in procedure. Each user has a
sets. (System 1032 is a product of CompuServe Data unique identification name. Audit trail files associate
Technologies, Boston.) all PoP data system activities with the user's identi-
Interform edits. At the VAX level each data set is fication name. The clinical site data manager is
processed through extensive interform edits to en- responsible for maintaining a confidential list of
sure consistency and validity across all forms. Error system users and passwords.
reports are generated and sent to the clinical centers Treatment assignment file. The treatment assign-
ment file on each site computer contains the treat-
for resolution. Data changes that are required to ment assignment schedule for that site. When a
resolve errors at this level are made by submitting a patient is randomly assigned, the PoP module selects
data base update record by means of the PoP system. the next treatment assignment in the list and associ-
These records are retrieved during the weekly data ates it with the randomly assigned patient's name and
file transfer and are loaded into the VAX system, BARI identification name number. The BARI iden-
where a program accesses the data base and per- tification and the treatment assigned are then stored
forms the requested change. This system generates in a random assignment file.
an audit trail of all changes made to the BARI data Because of its highly sensitive nature, the random
base at the VAX level. assignment file is encrypted. This ensures that the
Appending data. The data transmitted weekly to the next available random assignment cannot be deter-
CC from the clinical sites are loaded into the aggregate mined by any method other than that of the PoP
data base on the VAX. An append audit trail is random assignment module. The random assignment
generated that shows all records loaded into S1032 data file is protected from accidental deletion and cannot
sets and any records that are rejected. All data analyses be modified.
originate from this final, clean, aggregate data base. In an effort to prevent improper assignment to
randomly assigned treatment, the PoP random as-
Form Inventory signment module requires that patient data that
PoP provides an automatic form inventory system. pertain to exclusion criteria be entered. Treatment
Information from each entered form is automatically will be assigned only after data corresponding to
extracted from the data record and recorded in a appropriate criteria have been entered. Once a pa-
separate inventory data base. The inventory data base tient receives a treatment assignment, there is no way
is used to monitor data management progress at the of removing the patient from the random assignment
site level. In addition, this data base generates sched- file.
uling reports that inform the clinical staff of patients Coordinating Center VAX Computer. The CC en-
who are due for a protocol visit during the next month. sures patient confidentiality by using alphanumeric
 The BARI Protocol V-15

identification names to identify forms. This identifi- surface; or climbing more than one flight of ordinary
cation name is easily linked to the patient name at stairs at a normal pace and under normal conditions.
the site level only. Class III: Marked limitation of ordinary physical
Study data files at the CC are protected against activity, such as walking one to two blocks on a level
inadvertent change or access by nonproject person- surface or climbing one flight of stairs under normal
nel. The VAX computer system provides protection conditions and at a normal pace; comfortable at rest.
at the individual file level by means of passwords and Class IV: Inability to carry on any physical activ-
file protection codes that are controlled by the proj- ity without discomfort. Anginal syndrome may be
ect manager. File protection codes can permit certain present at rest.
users to have "read only" access to data without the
ability to make changes. Clinically Severe Angina or Ischemia
Backup of VAX files is performed daily. In addi- Clinically severe angina or ischemia involves any of
tion, a special tape backup is made weekly for the the following conditions: 1) non-Q wave myocardial
entire VAX computer system. These tapes are stored infarction stabilized 4 hours to 6 weeks; 2) unstable
in a room and location that are separate from the angina stabilized 4 hours to 6 weeks; 3) stable CCSC
daily backup medium. Therefore, a computer system III or IV angina; 4) stable CCSC I or II angina and
failure will at worst result in the loss of 1 day's work. one or more of the following: a) severe ischemia on
In a worst-case disaster (fire, explosion, etc.), no noninvasive testing, b) Q wave MI (stabilized for 24
more than 1 week's work would be lost. hours, revascularization clinically indicated, and abil-
ity to randomly assign patients within 30 days after
Appendix 1: Acronyms MI), c) resting ejection fraction <50%, d) a history of
AAP, Angiographic Assessment Program; ACE, stable class III or IV angina and current intensive
angiotensin converting enzyme; BARI, Bypass An- medical therapy (with exercise testing waived); 5) no
gioplasty Revascularization Investigation; BMDP, Bio- angina but severe ischemia on noninvasive testing and
Medical Data Program; CABG, coronary artery by- prior Q wave MI; 6) no angina within 6 weeks of
pass graft surgery; CAD, coronary artery disease; study entry but a history of prior angina, with current
CASS, Coronary Artery Surgery Study; CC, Coordi- severe ischemia on noninvasive testing.
nating Center; CCSC, Canadian Cardiovascular So-
ciety classification; CEL, Central Electrocardio- Definite Q Wave Myocardial Infarction
graphic Laboratory; CRL, Central Radiographic A definite Q wave MI involves the two-step wors-
Laboratory; DBMS, Data Base Management System; ening of Minnesota code Q waves, as determined by
ECG, electrocardiogram; EF, ejection fraction; E1TT, the CEL.
exercise treadmill testing; IMA, internal mammary
artery; LAO, left anterior oblique; LV, left ventricle; Multivessel Coronary Disease
MC, Minnesota Code; MI, myocardial infarction;
MMCC, Morbidity and Mortality Classification Com- The BARI definition of multivessel coronary dis-
mittee; NHLBI, National Heart, Lung, and Blood ease requires that two or more myocardial territories
Institute; NIH, National Institutes of Health; PI, (anterior, lateral, inferior/posterior) be jeopardized.
Principal Investigator; PTCA, percutaneous translu- Therefore, the following definition could also be
minal coronary angioplasty; RAO, right anterior termed multiterritory disease. To be eligible for
oblique; SAS, Statistical Analysis System; SDMB, BARI a patient must have a 50% or greater diameter
Safety and Data Monitoring Board; SEQOL, Study reduction by caliper measurement in arteries that
of Economics and Quality of Life; spss, Statistical supply at least two of the three major myocardial
Package for the Social Sciences; TIMI, Thrombolysis territories (Table 1, Figure 3). These arteries must be
in Myocardial Infarction. >1.5 mm in diameter adjacent to the site of the
stenoses. Assignment of coronary segments to a
Appendix 2: Bypass Angioplasty Revascularization specific myocardial territory (anterior, lateral, inferi-
Investigation Glossary of Definitions
or/posterior) includes consideration of coronary ar-
Canadian Cardiovascular Society Classification tery dominance (Table 1).
ofAngina
Class I: Ordinary physical activity does not cause
angina, such as walking and climbing stairs. Angina TABLE 1. Coronary Segments*
occurs with strenuous or rapid prolonged exertion at
work or recreation. Myocardial
territory Right dominant Left dominant Balanced
Class II: Slight limitation of ordinary activity, such
as walking or climbing stairs rapidly; walking uphill;
Anterior 12-17, 29 12-17, 29 12-17, 29
walking or stair climbing after meals; the limitation of Lateral 18-22, 28 18-22, 28 18-22, 28
ordinary activity in cold or windy weather, while Inferior/
under emotional stress, or during the few hours after posterior 1-9 23-27 1-5, 23-26
awakening; walking more than two blocks on a level *See Figure 3.
V-16 Supplement V Circulation Vol 84, No 6 December 1991

10

/23

27-_- 26-.- 25-

FIGURE 3. Coronary artery map. Right coronary artery: 1, proximal; 2, middle; 3, distal; 4, posterior descending; 5, right
posteroatrioventricular; 6, first posterolateral; 7, second posterolateral; 8, third posterolateral; 9, inferior septal artery; 10, acute
marginal artery. Left coronary artery: 11, left main; 12, proximal left anterior descending; 13, middle left anterior descending; 14,
distal left anterior descending; 15, first diagonal; 15a, first diagonal branch; 16, second diagonal; 16a, second diagonal branch; 17,
anterior septals; 18, proximal circumflex; 19, middle circumflex; 19a, distal circumflex; 20, 21, 22, first, second, and third obtuse
marginal; 20a, 21a, 22a, first, second, and third obtuse marginal branches; 23, left atrioventricular; 24, 25, 26, first, second, and
third posterolaterals; 27, left posterior descending; 28, ramus; 28a, ramus branch; 29, third diagonal; 29a, third diagonal branch.

Myocardial Ischemic Response (End Point Definition) fixed defect, and one fixed defect and one reversible
A myocardial ischemic response is indicated by any defect remote from the fixed defect. 3) Persantine
of the following exercise-induced ECG responses: 1) 20`T1: multiple reversible defects and one reversible
horizontal or downsloping ST segment depression and one fixed defect. 4) Exercise radionuclide ven-
.1 mm 80 msec after the J-point; 2) a slow upsloping triculography: resting ejection fraction >0.50 and
ST segment depressed .1.5 mm 80 msec after the exercise ejection fraction <0.50, with a decline in
J-point; and/or 3) ST segment elevation .1 mm 80 ejection fraction of at least 5% in the absence of left
msec after the J-point in a non-Q wave lead, each as bundle branch block or significant arrhythmia; and
compared with the rest tracing. resting ejection fraction .0.50, with a severe ST
segment response and work load <450 kpm.
New Technology Devices Severe ST Segment Response
New technology devices include those currently A severe ST segment response may be indicated by
holding Investigational Device Exemption status and any of the following exercise-induced ECG re-
any one deemed "new technology" by the New sponses: 1) horizontal or downsloping ST segment
Technology Committee. Some examples are stents, depression >1.5 mm 80 msec after the J-point; 2)
atherectomy, and laser catheters. (See pp V-3-V-6, slow upsloping ST segment depression .2.0 mm 80
Angioplasty Guidelines.) msec after the J-point; and/or 3) ST segment eleva-
tion >1 mm 80 msec after the J-point in a non-Q
Severe Ischemia on Noninvasive Testing (Entry wave lead, each as compared with the rest tracing.
Criteria Definition)
Severe ischemia on noninvasive testing is defined Suspected Myocardial Infarction
as the results of testing as listed herein: 1) Exercise MI is suspected if any episode of chest pain lasts 20
treadmill testing: exercise-limiting definite angina, minutes or longer or if there is the occurrence of new
with final exercise stage less than Bruce stage 3; and Q waves in the ECG or if there is elevation of cardiac
exercise-induced severe ST segment response, with a enzymes, as well as any other reason to suspect an
final exercise stage less than Bruce stage 3. In addi- MI.
tion, patients who do exercise in Bruce stage 3 may Flow Classification (Thrombolysis in Myocardial
also be considered for random assignment if the Infarction): A Randomized Clinical Trial of
ischemic changes occur within the first 6 minutes and Tissue- Type Plasminogen Activator (t-PA) in Patients
the patient is judged to be a candidate for revascu- With Acute Myocardial Infarction
larization. 2) Exercise 201T1: multiple reversible de- Four specific grades of coronary flow rate, as
fects, increased lung uptake and a single reversible or developed in the TIMI study, are given below.
 The BARI Protocol V-17
Grade 3 flow. Antegrade flow into the terminal other words, the sample size was chosen in such a
coronary artery segment distal to a stenosis occurs as way that if the two strategies were in fact identical in
promptly as anterograde flow into a terminal coro- terms of mortality, the study would have a high
nary artery segment proximal to the stenosis. The probability of ruling out excess mortality of 2.5% or
clearance of contrast material from the distal coro- more in the PTCA group relative to the CABG
nary artery terminal segment is as rapid as clearance group.
from an uninvolved, more proximally located coro- Review of the data from the National Heart, Lung,
nary artery segment. and Blood Institute PTCA Registry and the CASS
Grade 2 flow. Contrast material opacifies the ter- Trial and Registry indicates that for patients with
minal coronary artery segment distal to the stenosis. multivessel CAD, a 5-year mortality rate of 5% after
However, the entry of contrast material into the either therapy was a reasonable estimate. For the
terminal coronary artery segment is perceptibly purpose of calculating sample size, we assume this to
slower than it is into a more proximally located be the true common mortality for both PTCA and
terminal coronary artery segment. Alternatively, the
clearance of contrast material from a terminal coro- CABG patients in BARI. Under this assumption, a
nary artery segment distal to a stenosis is noticeably sample size of 2,400 patients (1,200 assigned to each
slower than that from a comparable segment that is treatment) yields a probability of 88% that the upper
not preceded by a significant stenosis. 95% confidence limit for the true difference between
Grades 1 and 0 flow. There is no obvious flow of the 5-year mortality rates of PTCA and CABG will
contrast material beyond the stenosis in question. A not exceed 2.5%. Note that if the upper confidence
small amount of contrast material may penetrate the limit of the true difference is 2.5%, then the observed
occlusion (grade 1); however, the contrast material difference must be appreciably less than 2.5%. For
fails to fully opacify the coronary artery beyond the example, if the observed 5-year mortality rates were
occlusion. 5.9% for PTCA and 5% for CABG, the observed
Flow of contrast material not assessable. In this difference would be 0.9% and the upper 95% confi-
situation there may be competitive flow to a terminal dence limit for the true difference would be 2.4%.
coronary artery segment. For instance, total occlu- Appendix 4: Random Assignment
sion in the proximal right coronary artery may be
associated with both anterograde (ipsilateral) and Stratification
retrograde (contralateral) collateral flow to the distal The sequence of random assignment is stratified by
right coronary artery bed. There may be relatively clinical center. This allows for approximately equal
faint visualization of the distal right coronary bed by allocation of the two treatments within each center as
the anterograde collateral flow and also by retro- well as preservation of the overall balance between
grade collateral flow. In this situation it is not possi- the two treatment groups.
ble to accurately assess the flow rates of contrast Because of the large number of patients antici-
material in the terminal right coronary artery bed pated (1,200 for each treatment), the probability of
because of the competitive nature of flow. It may also serious imbalance in the distribution of risk factors is
be impossible to judge relative flow in a proximal unlikely. In case of such an occurrence, adjustment
total occlusion in which collateral flow to the termi- can be accomplished during the analysis. This can be
nal coronary segments is poor. If the terminal coro- done without a significant loss of efficiency, and it is
nary segments in this situation are poorly visualized, preferable to the administratively more complex pro-
the correct judgment concerning perfusion may be cess of stratifying during random assignment. There-
unclear, and this choice should be indicated. fore, the sequence of random assignment is stratified
by clinical center only, and other clinically important
Appendix 3: Sample Size factors will be adjusted for (if necessary) in analyses
Mortality was used as the basis for selecting the after random assignment.
sample size for the trial. For PTCA (a procedure less
invasive than CABG) to be recommended as an Design of Random Assignment
initial revascularization strategy, it is of primary The BARI clinical trial requires a scheme of
importance to know that the long-term results with random assignment that will provide a balance in the
PTCA are not appreciably worse than with CABG in total number of patients assigned to each treatment
terms of mortality. With this in mind, the basis of the group both early and late in the trial. Early balance is
BARI sample size calculation was to choose a sample necessary to maximize the power to detect differ-
size large enough so that, with a high probability, the ences for safety-monitoring purposes and to adjust
upper confidence limit for the absolute difference for any early learning curve effects in the trial. To
between PTCA and CABG mortality rates would not achieve this balance, a blocked scheme of random
exceed a specified value.5 This specified value repre- assignment has been implemented, with generation
sents the maximum acceptable absolute difference of a separate sequence for each clinical center as
between the CABG and PTCA event rates. In BARI described below.
the specified value was taken to be 2.5%, in light of Block lengths were selected at random. After this
the estimated 5% mortality rate discussed below. In number was chosen, a permutation of this number of
V-18 Supplement V Circulation Vol 84, No 6 December 1991

treatment assignments, with half being PTCA and Cardiac Tamponade

half CABG, were selected at random from all possi- Hemodynamic compromise, with hypotension and
ble permutations. This permutation then became low cardiac output secondary to increased pericardial
part of the assignment sequence. The process was pressure.
repeated until the assignment sequence was of a
length of 500 or more. (Each center is expected to
recruit an average of 175 patients for a trial size of Cerebrovascular Accident (CVA)
2,400 patients, but assignment schemes are longer to A focal neurological deficit that is still at least
allow some clinics to recruit more patients in case partially evident more than 24 hours after its onset.
other clinics fall behind the projected recruitment
rates or have to stop recruitment.) Chest Tubes for More Than 5 Days After Surgery
(Self-Explanatory)
Treatment Assignments
By using the blocked random assignment scheme Coma
indicated above, the CC has generated the treatment Profound depression in the level of conscious-
assignment schemes for all centers. The file with the
random assignment sequence for each center is en- ness, reflected by the loss of contact with the
coded and stored on the same microcomputer that is environment and loss of spontaneous movement.
used for data entry. Because a high level of security is Brain stem activity (respiration and response to
essential for this file, it is encrypted on each micro- deep pain) may or may not be preserved.
computer to ensure that there is no unauthorized
access to these blinded data. Security issues are Congestive Heart Failure (CHF)
discussed further on p V-13 (Data Management). Manifested by one or more of the following fea-
When a patient is to be randomly assigned, the tures: dyspnea on exertion, edema, fatigue, orthop-
clinic coordinator uses the software developed for nea, and/or paroxysmal nocturnal dyspnea. Other
random assignment. The coordinator is prompted findings that support the clinical diagnosis include
with a list of questions that reflect BARI trial inclu- the presence of an S3 gallop, elevated jugular venous
sion and exclusion criteria. Answers to these ques- pressure, and radiographic evidence of pulmonary
tions must match the entry criteria (i.e., BARI eligi-
bility must be verified) before the program will congestion. Verification by a physician's statement in
indicate that the patient is indeed eligible and pro- the medical record is required.
vide the patient's treatment assignment. The CC
closely monitors the use of the random assignment Deep Wound Infection
program. Positive culture from the mediastinal tissues be-
A contract is maintained at all centers to guarantee neath the sternum.
service within 4 hours of notification of computer
breakdown. The CC is available to assign treatments by Dementia
telephone in the event that computer repair is delayed.
Copies of each random assignment sequence are stored Broad-based loss in higher intellectual function
at the CC in secured computer files. (including cognitive, perceptual, calculational, and/or
recall functions) that is evident to family members
Appendix 5: Complications and close associates or demonstrated on serial func-
Abrupt Closure (Regardless of Previous Dilatation) tional testing.
There is clinical evidence of one or more coronary
arteries abruptly closing. Hemorrhage
Severe bleeding sufficient to require transfusion of
Arterial Embolus of Extremity or Permanent Loss packed red blood cells.
of Pulse
Acute occlusion of a main or distal arterial trunk Hypersensitivity Reaction
supply in a limb assessed by the lack of detectable An allergic reaction to iodine-containing radio-
distal arterial pulsations (by pulsation or Doppler
examination) that had previously been observable. graphic contrast media or protamine, which is
The loss of pulse may or may not be associated with marked by the development of urticaria, wheezing,
ischemia of the affected limb. prolonged hypotension, or laryngospasm.
Cardiogenic Shock Hypotension
Hypotension (systolic blood pressure of less than Reduction in systolic blood pressure to <90 mm Hg
80 mm Hg) that is associated with reduced urine or reduction by .30 mm Hg compared with the base-
output, decreased mental acuity, and compensatory line value, which persists for more than 1 minute and
vasoconstriction. requires treatment.
 The BARI Protocol V-19

Nonfatal Cardiac Arrest That Requires CPR Wound Dehiscence

or Countershock (Self-Explanatory) The splitting or bursting open of a procedural
wound.
Postthoracotomy Syndrome
Clinical syndrome of pleuro/pericarditis mani- Appendix 6: Angiography and Functions of the
fested by chest pain and fever. Central Radiographic Laboratory
Purpose of the Central Radiographic Laboratory
Pulmonary Edema (Cardiac) The CRL is responsible for processing all BARI
Acute congestive heart failure resulting in the data that relate to angiograms. Through the develop-
accumulation of pulmonary interstitial and alveolar ment and implementation of the AAP, the CRL
fluid, manifested by profound dyspnea, orthopnea, ensures that angiographic data regarding coronary
rales, and evidence of pulmonary congestion on chest anatomy and lesion morphology are collected in a
X-ray. standard format across all sites. Angiographic data
for all BARI patients are entered into the AAP at
Pulmonary Embolus each site. In addition, the CRL performs central
Acute occlusion of one or more branches of the angiographic readings for all randomly assigned pa-
pulmonary arteries with thrombotic material, usually
tients. The central readings performed by the CRL
originating from the iliofemoral or pelvic veins, mani- will be used in the primary analysis of BARI.
fested by the abrupt onset of pleuritic chest pain, Angiographic Procedural Requirements
worsened gas exchange, increased pulmonary artery
pressure, or frank hemodynamic collapse.
Nitroglycerin is usually administered immediately
before coronary arteriography. Notations of catheter
Renal Failure size, nitroglycerin administration, and projections
used are documented on a Procedure Report Form.
Progressively deteriorating renal function requir- Multiple standard views are obtained by using caudal
ing dialysis. RAO and cranial LAO projections of the left coro-
nary artery and paired LAO/RAO projections of the
Reoperation for Bleeding right coronary artery when possible. Six-inch or 7-in.
Reoperation performed to remedy excessive bleed- image-intensifier modes are used for standard views.
ing after initial surgery. Before PTCA two views are obtained of each lesion,
and these views are replicated after PTCA. A left
Respiratory Failure That Includes Noncardiac ventricular angiogram in which a 300 RAO view is
Pulmonary Edema and Adult Respiratory Distress used is required, and a cranially angulated steep
Syndrome (ARDS) LAO view is also obtained when possible.
Inability of the patient to maintain adequate gas
exchange during spontaneous ventilation, even with Required Angiograms
the assistance of supplemental oxygen. In cases All angiograms that are performed on BARI patients
where a patient is receiving mechanical ventilatory must be entered into the AAP at the clinical site.
assistance after surgery, respiratory failure shall be Angiograms are assigned to the following categories.
considered the inability to wean the patient from Baseline. The angiogram that is used to document
mechanical ventilation within 48 hours of completion angiographic eligibility for BARI is considered the
of the surgical procedure. baseline angiogram. This angiogram must be entered
In noncardiac pulmonary edema the same clinical into the AAP within 5 working days of random
and radiographic appearances that are in cardiac assignment.
pulmonary edema are present. However, the pulmo-
Before and after initial PTCA. For patients that are
randomly assigned to PTCA, data must be entered
nary capillary wedge pressure is 15 mm Hg.
within 5 working days of completion of the procedure.
ARDS is the increased interstitial and alveolar lung Intercurrent angiograms. All angiograms that are
water resulting from lung injury from any of a variety of performed during follow-up, whether they are part of
causes in the absence of a cardiac etiology. a subsequent revascularization procedure or not, are
considered intercurrent angiograms. These must be
Superficial Wound Infection entered within 5 working days of the angiogram for
Positive culture from the surgical or PTCA wound randomly assigned patients.
site.
Angiographic Data Entry, Management,
Transient Ischemic Attack (TL4) and Interpretation
A focal neurological defect (usually correspond- The acquisition of angiographic data is performed
ing to a single vascular territory) that spontaneously by using the AAP, a microcomputer and software
resolves so that no residual evidence exists within system that was developed specifically for use in
24 hours. BARI. The AAP provides a uniform format for the
V-20 Supplement V Circulation Vol 84, No 6 December 1991

entry and storage of coronary and left ventricular evidence of QRS, ST, and T wave items by using the
angiographic data at the clinical sites and provides standard Minnesota code and supplemental criteria.7
tools (calipers, tables, reports) for clinicians and Exercise ECGs are evaluated for ischemia based on
angiographers at each institution. Electronic calipers the BARI definitions for severe ischemia on nonin-
are used to measure catheter diameter (as a calibra- vasive testing (entry criteria definition) and myocar-
tion), lesion diameter, and the diameter of normal dial ischemic response (end point definition).
reference segments of the vessel proximal and distal
to the lesion. The microcomputer system facilitates Resting ECG Acquisition by Using Marquette
the electronic transfer of data from the clinical sites MAC-12 Units
to the CRL. The same AAP system is used at the Through an educational grant from Marquette
CRL for angiographic interpretation and data analy- Electronics (Milwaukee, Wis.), the CEL has pro-
sis. The microcomputer hardware was provided to vided Marquette MAC-12 units to the original 14
the original 14 clinical centers and to the CRL by a
gift from Hewlett-Packard Co., Palo Alto, Calif. clinical centers that did not have Marquette equip-
All baseline, PTCA, and intercurrent films for ment. Baseline, preprocedure, postprocedure, and
randomly assigned patients are sent to the CRL by scheduled follow-up ECGs are acquired on Mar-
commercial delivery service as soon as possible after quette equipment. This standard recording appara-
patient discharge. Films are read at the CRL and tus facilitates ECG transmission to the CEL and the
returned within 20 working days whenever possible. system provides a repository for digitized ECGs,
Site angiographic readings that entered into the AAP allowing easy retrieval for subsequent coding.
are transmitted to the CRL by way of telephone
connections on a weekly basis. ECG Coding
At the CRL evaluation of baseline and intercur- Rest ECGs are evaluated by using the Minnesota
rent angiograms is done by a senior angiographer, Code. To more accurately describe the patient popula-
who reviews coronary vessel distribution, branching, tion in BARI, additional data are collected as part of an
and myocardial territory supplied. Lesions are as- extension of the Minnesota Code. The actual height or
sessed for location, severity (calipers), and morphol- depth of the ST segment shift is recorded, and codes
ogy. LV angiography is evaluated by using quantita- have been added to describe the depth and extent of T
tive methods for the measurement of ejection wave inversion in the anterior precordial leads. Finally,
fraction and regional wall motion. These evaluations Q wave codes have been added to identify characteris-
are compared with the initial site reading, and if tics that, along with ST segment and T wave changes,
there are no major differences the CRL reading is are associated with a high probability of myocardial
final. In the case of major differences, the film is read infarction.
by a second CRL reader and adjudication is per- ECGs are received and logged in at the CEL. When
formed. Evaluation of PTCA films is performed by a
senior angioplasty operator, who reviews the studies multiple ECGs are sent to document the postproce-
for lesion severity and morphology before and after dure period or evaluate a suspected myocardial infarc-
PTCA. The same adjudication procedures that were tion, the ECGs that have the best representation of Q
outlined above are performed during this evaluation. waves and ST segment changes are chosen for analysis.
ECGs are first evaluated for technical quality, including
Quality Control the presence of missing leads, excess artifacts, excess
Angiograms. The CRL reviews all randomly as- baseline wandering, switched leads, and invalid calibra-
signed studies for image quality and protocol compli- tion. ECGs are logged into an inventory system, and a
ance. Feedback is provided to the site, the angiogra- coding clerk measures Q wave depth and width, R wave
pher, and the Operations Committee. amplitude, ST segment displacement, and T wave
PTCA procedures. Films that document PTCA pro- amplitude by using an eight-power magnified loupe
cedures are reviewed to assess PTCA outcome. Films calibrated in increments of 0.1 mm. These measure-
are graded with respect to protocol compliance as ments are verified by the coding supervisor and then
well as angiographic findings. entered into a MicroVAX computer. Logical edits are
CRL angiographic readings. The CC randomly des- then performed to verify the codes, and any discrepan-
ignates angiographic studies to be reread at the CRL. cies are resolved. The codes are then reviewed by a
This rereading is performed in a blind manner to physician, who has the option to make changes to the
compare the initial CRL reading with the rereading data. All digitized ECGs are transmitted to Dalhousie
of the same angiogram to determine the amount of University, where electronic determination of the Min-
variability in the CRL interpretation of specific an- nesota Code is performed. Any discrepancies between
giographic variables. the Dalhousie and CEL readings are adjudicated, with
Appendix 7: Electrocardiography and Functions of the CEL making the final code determination. The final
the Central Electrocardiographic Laboratory data are used to create summary files that are trans-
mitted weekly to the CC.
Purpose of the Central Electrocardiographic
Laboratory (CEL) Quality Control of ECG Coding
The CEL analyzes all rest and exercise ECGs that Coding clerks and physician electrocardiographers
are collected in BARI. Rest ECGs are evaluated for must pass a comprehensive coding test before the
 The BARI Protocol V-21

actual coding of project tracings. Within the CEL, exercise test and the ECG or test that preceded it. Both
ECGs are periodically recirculated to detect coding ECGs or sets of ECGs are available to the laboratory
deficiencies. A staff member who is identified as supervisor and physician at the time of overreading.
having inadequate performance undergoes remedial The standard Minnesota and supplemental codes are
training until performance improves. Kappa statistics used to determine significant change in Q, ST, and T
are calculated biannually for the physicians and wave items.
laboratory supervisors.
In addition, the CC is recirculating a blinded sample Appendix 8
of 200 baseline ECGs for Q wave interpretation. A Form A
25% sample of ECGs that indicate Q waves during
follow-up and an additional 0.5% of ECGs without Q Consent for Participation in the Research Study
waves are also being recirculated. This serves to docu- Entitled Bypass Angioplasty Revascularization
ment the reliability of the central diagnosis of this Investigation (BARI), a Randomized Trial Comparing
major BARI end point. Interreader variability of binary Coronary Bypass Surgery to Coronary Angioplasty
data will be assessed by using the Kappa statistics. Study Investigators:
Exercise Test Acquisition Office of Human Research:
Exercise tests are performed by using a motor- General Description and Purpose of Research
driven treadmill and following the Bruce protocol8 You are invited to participate in a trial sponsored
with no or at least one half of a warmup stage if
necessary. A physician is in attendance throughout by the National Heart, Lung, and Blood Institute
the test, and the patient and ECG tracings are which involves comparison of two treatments for
carefully observed. All patients undergo a pretest patients with severe myocardial ischemia and coro-
evaluation, during which ongoing drug therapy and nary artery disease. You have coronary artery dis-
indications for stopping the exercise are ascertained. ease involving at least two of the major coronary
The Mason-Likar 12-lead torso ECG is the lead arteries that supply blood to your heart muscle.
set used. Four specific tracings are required: upright Your doctors have determined that you are a
immediately before exercise; peak exercise; immedi- candidate for either percutaneous transluminal cor-
ate postexercise; recovery period (3-5 minutes after onary angioplasty (PTCA), a procedure in which a
exercise); or an ECG showing maximum postexercise balloon catheter will be passed across the narrowing
change. in your coronary artery to open the narrowing, or
coronary artery bypass graft (CABG) surgery, an
Exercise Test Digitization operation in which the narrowing in the coronary
Exercise ECGs are first examined for the presence of artery [Link] bypassed by using a vein from your leg
excess artifacts, baseline wandering, and missing leads or an artery from your chest. One of the potential
or ECGs. Each test is reviewed by the laboratory complications of using the balloon catheter (PTCA)
supervisor or associate to determine the presence of is that the narrowing may return at the site of
exclusion codes such as left bundle branch block, left dilation. This renarrowing has been found in as
ventricular hypertrophy, etc. The maximum depth of many as 30% of patients undergoing this procedure
ST segment depression, maximum height of ST eleva- at various medical centers. Narrowing generally
tion in a non-Q wave lead, maximum number of occurs within the first 6 months, and a second
abnormal leads, and maximum height of ST elevation balloon angioplasty procedure usually can be per-
in a Q wave lead are calculated. A Q wave lead is one formed. When the narrowing does not recur within
in which the Q wave width exceeds 0.03 seconds. This is the initial 6 months, it is not likely to occur in the
determined from the resting ECG that undergoes subsequent 3-5 years.
Minnesota Coding. These exercise ECG data are trans- I understand that the risks of CABG and PTCA
mitted weekly to the CC. are considered to be similar and that it is possible for
death to occur in approximately 1-2% of patients
Quality Control of Exercise ECG Data undergoing these procedures. The likelihood of sus-
A 2% sample of scheduled and unscheduled exer- taining a heart attack (myocardial infarction) during
cise ECGs are submitted for reproducibility studies. either of these procedures is approximately 5%.
The sample is enriched with 25% of all exercise tests During the study, 2,400 patients will be assigned by
that indicate a myocardial ischemic response. Varia- chance to either form of therapy during a period of 2
bles to be analyzed for reproducibility include the years in 14 hospitals.
maximum number of abnormalities in any lead If I agree to participate in the randomized clinical
group, the total number of abnormal leads, and the research study, one of these two revascularization
maximum depth of ST segment depression. procedures will be chosen randomly (by chance) as
the method to correct the narrowings of my coronary
Serial Comparison arteries. If I am selected to receive CABG, this
For both rest and exercise ECGs, serial comparisons procedure will be carried out within 2 weeks after my
are performed by using the most recent rest ECG or coronary angiogram was made, at a time that is
V-22 Supplement V Circulation Vol 84, No 6 December 1991

mutually agreeable to me and the physicians caring participate will involve no penalty or compromise in
for me. my medical care or loss of benefits to which I am
If I am selected to receive PTCA, I agree to otherwise entitled. I may discontinue participation at
undergo a repeat heart catheterization and to have any time without penalty or loss of the benefits to
balloon angioplasty performed on one or more of the which I am entitled.
narrowings in my coronary arteries. The PTCA pro-
cedure is successfully performed in approximately Alternative Treatments
80% of patients. Should this procedure fail because Possible alternatives to my participation in this
of complications resulting from the procedure, I study include medical therapy with standard ap-
understand that it may be necessary to proceed proved drug regimens, coronary angioplasty, or cor-
immediately with CABG. onary bypass surgery. Each of these three options is
After the treatment, whether it is PTCA or CABG, standard in the management of patients with coro-
I will be carefully followed for 5 years and will be
seen, or contacted by telephone, throughout my nary artery disease.
hospital stay, at 4-12 weeks after hospital discharge, Investigational Sponsorship and Cost Considerations
at 6 and 12 months after the procedure, and every
year thereafter. An exercise test will be performed at The physician investigators listed on this form;
4-14 weeks and at 1, 3, and 5 years after the research personnel associated with this study at
procedure to determine maximal exercise capacity; ; and the National Heart, Lung, and
this test may provide diagnostic information concern- Blood Institute (the sponsor of the study) may
ing my state of health. I understand that certain inspect and copy my medical records relating to this
complications may occur during or after an exercise study. The results of the study will be reported to a
test, such as abnormal blood pressure response, coordinating center selected for data analysis. Con-
dizziness, irregular heart beat, or, very rarely, a heart fidentiality of my medical record will be maintained
attack or death. However, the risk of a complication by the use of a numerical or alphabetical code. In
during exercise testing is infrequent, and the risk is the event of any publications regarding this study,
further decreased by the presence of appropriate my identity will not be disclosed. I will be informed
medical facilities and an experienced medical team. of any significant new findings developed during the
At 5 years, I will be admitted to the hospital for a course of the research which may relate to my
repeat coronary angiogram. Coronary angiography willingness to continue in the study. These findings
defines the presence, location, and extent of coronary may also be published in the medical literature.
disease and associated heart muscle function. The The choice of PTCA or CABG may affect the
test will assess the results of the PTCA or CABG. number of days I will be hospitalized. In general,
Complications from a coronary angiogram are infre- CABG requires a longer hospitalization than PTCA
quent (less than 1%) and include a blood clot where does. However, PTCA may require repeat admis-
the catheter is introduced and the possibility of a sions in the initial year after the procedure because
stroke, heart attack, or death (less than 5%). The risk of recurrent narrowings, which may require a second
is no greater than that incurred during the initial angioplasty procedure. One of the purposes of this
coronary angiography. study is to determine whether there are differences in
the total number of days of hospitalization, over the
Risks and Benefits long term, depending on the initial revascularization
I understand that there are possible risks to me if I procedure chosen. My physician has recommended
agree to participate in this study. The risks are those of that I would best be treated by having PTCA or
the PTCA procedure, CABG, exercise tests, and coro- CABG. Therefore, the costs of the PTCA and of the
nary angiography as stated above. I understand that my CABG, the direct hospital expenses, and the direct
doctor has recommended that I have a revasculariza- physician fees that are considered clinically indicated
tion procedure (PTCA or CABG) and that the test for my care will not be paid for by the sponsor of this
procedures described above are routinely used in the study. The use of exercise studies with or without a
assessment and follow-up of patients with coronary radionuclide is considered routine clinical care in the
artery disease. With participation in this study there is follow-up of patients with coronary artery disease
a risk that I may receive the less effective of the and will not be paid for by the sponsor. The cost of
treatments being studied. However, the potential ben- the coronary angiogram performed 5 years after
efits include the possibility that I may receive the more CABG will be paid for by the sponsor of the study if
effective therapy. The best medical knowledge at this this test is not required as part of usual care.
time does not permit a scientific recommendation to (Medical Center Title), in fulfilling its public re-
me as to which option is better. sponsibility, accepts professional liability and respon-
sibility for physical injury if it is caused by negligence
Voluntary Consent, Right to Withdraw of the Center and its employees or agents. No person
If I agree to participate in this study, I understand shall have any authority, orally or in writing, to
that my care will in no way be compromised. My change the terms of the foregoing. Any questions that
participation in this study is voluntary. My refusal to I have concerning the research study or my partici-
 The BARI Protocol V-23

pation in it, before or after my consent, will be to participate in this Registry. If you decide to
answered by participate, you will be interviewed each year for the
In the event I believe that I have suffered any next 5 years or more by mail or telephone. Should
injury as a result of participation in the research you move from your present address, a private agency
project, I am to contact , who will be able may be used to determine your new home or place of
to refer me to an individual who will review the work. These interviews will take approximately one-
matter with me, identify other resources that may be half hour of your time. The questions asked will be
available to me, and provide information concerning about your symptoms of angina or other heart prob-
additional inquiries. If I am a woman, I am not lems. You will also be asked whether or not you have
pregnant, to the best of my knowledge. I am not undergone any hospitalizations because of your heart
participating in any other medical research study. I problems or whether you have undergone a repeat
have read the above statement and have been able to PTCA or CABG or have been hospitalized for treat-
express concerns which have been satisfactorily re- ment of a heart attack. The questions are not con-
sponded to by the investigator. sidered to have psychological stress. No standard
I believe I understand the purpose of this study, as treatment will be required by your treating physicians
well as the potential benefits and risks that are nor will any standard treatment be withheld from
involved. I hereby give my informed and free consent you. You will receive no payment or other compen-
to be a participant in this study. sation for participating in this Registry. There may be
no immediate benefit to you, but the combined
information from this Registry may help your physi-
Patient's Signature Date cian and other physicians in the future to better
understand the effects of the treatment you have
received on ischemic heart disease and its course.
Witness's Signature Date I understand that, upon my request, a BARI
research investigator will answer questions about the
I certify that I have explained to the above indi- study. I may refuse to participate or discontinue
vidual the nature, purpose, potential benefits, and participating in the study at any time without penalty
possible risks associated with participating in this or loss of benefits available to me as a patient at this
research study. I have answered any questions that medical center.
have been raised and have witnessed the above I understand that no commitment is made to pro-
signature. These elements of informed consent con- vide complimentary medical care or compensation for
form to the assurance given by Depart- any adverse results of my participation in this study.
ment of Health and Human Services, to protect the Further information concerning institutional policies
rights of human subjects. I provided the patient a in this regard or information about the conduct of this
copy of this signed consent document. study or the rights of research subjects may be ob-
tained from
Confidentiality of information concerning partici-
Investigator's Signature Date pants will be maintained. Names of participants or
material identifying participants will not be released
Appendix 8 without written permission, except as such release is
Form B required by law. Medical records related to this study
may be made available to the Food and Drug Ad-
Patient Consent to Participate in the Bypass ministration, as provided for in federal regulations.
Angioplasty Revascularization Investigation Registry
Principal Investigator Date Signature of Participant
Approved by Institutional Review Board
Date Signature of BARI Investigator
Date_ Obtaining Consent
Some patients at the Medical Center Appendix 9
undergoing percutaneous transluminal coronary an- BARI
gioplasty (PTCA) or coronary artery bypass graft List of Participants
(CABG) surgery are being enrolled in a Registry of Clinical Centers
scientific data as part of a national collaborative
clinical study comparing the outcome of these proce- University of Alabama
dures. The purpose of this Registry is to provide Principal Investigator: William J. Rogers, MD
scientific information about the merits of PTCA PTCA Operators: William A. Baxley, MD
Larry S. Dean, MD
compared with CABG, another procedure for restor- Gary S. Roubin, MD
ing the flow of blood to heart muscle. You are invited Surgeons: James K. Kirklin, MD
V-24 Supplement V Circulation Vol 84, No 6 December 1991

John W. Kirklin, MD Former Participants: John Frierson, MD

Albert Pacifico, MD Jay Hollman, MD
George L. Zorn, MD
Associate Investigators: Edgar Charles, PhD Duke University
Thomas D. Paine, MD Principal Investigator: Robert M. Califf, MD
Coordinators: Larry E. Maske, RN PTCA Operators: Robert P. Bauman, MD
Terri E. Morgan, RN Victor S. Behar, MD
Leah C. Carr (SEQOL) Yihong Kong, MD
Data Manager: John A. Trobaugh, RA Mitchell W. Krucoff, MD
Staff: Karen W. Anderson Kenneth G. Morris, MD
Fredericka Harris Robert H. Peter, MD
Former Participants: Thomas Bulle, MD Harry R. Phillips, MD
J. Bradley Cavender, MD Richard Stack, MD
Paul J. Garrahy, MD James E. Tcheng, MD
Surgeons: Robert H. Jones, MD
Brown University: Rhode Island Hospital H. Newland Oldham, MD
Principal Investigator: David 0. Williams Peter Van Trigt, MD
PTCA Operators: Thomas M. Drew, MD Associate Investigators: Thomas Bashore, MD
David 0. Williams Donald F. Fortin, MD
Surgeon: Arun K. Singh, MD David J. Frid, MD
Associate Investigators: George N. Cooper, MD Kerry Lee, PhD
Barry L. Sharaf, MD Michael J. Miller, MD
Coordinators: Mark Macedo, RN E. Magnus Ohman, MD
Janice L. Wheeler, RN David B. Pryor, MD
Former Participants: John Moran, MD Alan N. Tenaglia, MD
Edward S. Thomas, MD Coordinators: Mary Ann Sellers, RN
Harvey White, MD Ellen T. Hampton, RN
Data Manager: Terri Daniels
Bellevue Hospital (Satellite to Brown)
Staff: Leonard Santoro
Former Participants: Sandra G. Burks, RN
Principal Investigator: Frederick Feit, MD Stephanie Caminiti, RN
PTCA Operators: Michael J. Attubato, MD Peter J. Quigley, MD
Frederick Feit, MD J. Scott Rankin, MD
Surgeons: Stephen B. Colvin, MD Joan Richard, RN
Aubrey C. Galloway, MD
Associate Investigator: Peter F. Pasternack, MD Harvard University: Beth Israel Hospital
Coordinator: Sonja Shapiro, RN Principal Investigator: Donald S. Baim, MD
Coprincipal Investigator: Robert Safian, MD
Boston University PTCA Operators: Julian Aroesty, MD
Principal Investigator: David P. Faxon, MD Donald S. Baim, MD
PTCA Operators: John E. Brush, MD Daniel Diver, MD
David P. Faxon, MD Beverly Lorell, MD
Gary R. Garber, MD Surgeons: Robert Johnson, MD
Alice K. Jacobs, MD Robert Thurer, MD
Nicholas A. Ruocco, MD Ronald Weintraub, MD
Surgeons: James D. Fonger, MD Coordinator: Mary Cunnion
Richard S. Shemin, MD Former Participants: Raymond McKay, MD
Associate Investigators: Michael A. Bettmann, MD Tia DeFeo-Fraulini, MS
Jesse W. Currier, MD Carolyn McCabe
James A. Rothendler, MD
Thomas J. Ryan, MD Maine Medical Center (Satellite to Harvard)
Donald A. Weiner, MD Principal Investigator: Mirle A. Kellett Jr., MD
Coordinators: Beth R. Hankin, RN PTCA Operators: Warren D. Alpern, MD
Mary E. Mazur, RN Richard A. Anderson, MD
Former Participants: Roger M. Mills, MD D. Joshua Cutler, MD
Gaetano Paone, MD Mirle A. Kellett Jr., MD
Paul W. Sweeney, MD
Cleveland Clinic Foundation Surgeons: Desmond J. Donegan, MD
Principal Investigator: Patrick L. Whitlow, MD Saul Katz, MD
PTCA Operators: Irving Franco, MD Robert S. Kramer, MD
Patrick L. Whitlow, MD Chris A. Lutes, MD
Surgeons: Delos Cosgrove, MD Jeremy R. Morton, MD
Floyd Loop, MD Edward R. Nowicki, MD
Bruce Lytle, MD Joan F. Tryzelaar, MD
Robert Stewart, MD Richard L. White, MD
Paul C. Taylor, MD Associate Investigator: Costas T. Lambrew, MD
Associate Investigators: Alexios P. Dimas, MD Coordinators: Pamela Birmingham, RN
William Proudfit, MD Jane Conner Kane, RN
Benjamin Robalino, MD Nancy Tooker, RN
Eric Topol, MD
Kevin Vaska, MD University of Massachusetts
George Williams, PhD Principal Investigator: Bonnie H. Weiner, MD
Coordinator: Amy Rogers, RN PTCA Operator: Bonnie H. Weiner, MD
Data Manager: Sharon Senick Surgeons: Dani Bitran, MD
Staff: Kathy Comella, RN John Moran, MD
Marsha Lowrie, RN Okike N. Okike, MD
Joyce Tedrick, RN Thomas Pezzella, MD
 The BARI Protocol V-25

Thomas J. VanderSalm, MD Yves Leclerc, MD

Coordinators: Marie Borbone, RN Conrad Pelletier, MD
Paul Wanta, RN Associate Investigators: Andr6 Arseneault, MD
Data Manager: Theresa Wisnewski Gilles Hudon, MD
Former Participants: Joseph Benotti, MD Jacques Lesp6rance, MD
James Dalen, MD David D. Waters, MD
John Gaca, MD Coordinator: Johanne Trudel, RN
Jeffrey Leppo, MD Data Manager: Claudette Faille
Michael Pasque, MD Staff: Huguette Flageol
Marilyn Shay, RN Lucette Whittom, RN
Mayo Clinic Toronto Hospital (Satellite to Montreal)
Principal Investigator: Michael Mock, MD Principal Investigator: Leonard Schwartz, MD
PTCA Operators: John Bresnahan, MD PTCA Operators: Harold Aldridge, MD
David Holmes, MD Leonard Schwartz, MD
Guy S. Reeder, MD David Uden, MD
Surgeons: Charles Mullany, MD Surgeons: Tirone David, MD
Thomas A. Orszulak, MD Chris Feindel, MD
Hartzell Schaff, MD Bernard Goldman, MD
Associate Investigators: Peter B. Berger, MD Irving Lipton, MD
Bernard Gersh, MD Lynda Mickleborough, MD
Raymond Gibbons, MD Richard Weisel, MD
Stephen L. Kopecky, MD Associate Investigators: David Almond, MD
Fred Nobrega, MD Michael McLoughlin, MD
Robert S. Schwartz, MD Leon Zelovitsky, MD
Hugh C. Smith, MD Coordinator: Linda Ganassin, RN
Coordinator: Sylvia Matheson, RN Staff: Karen Mackie, RN
Data Managers: Lisa Kelly
Mary Peterson New York Medical College
Lou Ann Pierre, RN Principal Investigator: Michael V. Herman, MD
Former Participants: Dennis Bresnahan, MD PTCA Operator: Melvin B. Weiss, MD
Ronald Vlietstra, MD Surgeons: Richard Moggio, MD
Richard Pooley, MD
Medical College of Virginia George Reed, MD
Principal Investigator: Michael J. Cowley, MD Mohan Sarabu, MD
PTCA Operators: Michael J. Cowley, MD Coordinator: Doris Efstathakis, RN, BSN
Germano DiSciascio, MD Data Manager: Yonina Sait, PA
George Vetrovec, MD Former Participants: Peter Praeger, MD
Surgeons: Albert Guerraty, MD Kathleen Ryman, MD
David Salter, MD Eric Somberg, MD
Andrew Wechsler, MD Jonathan H. Stein, MD
Associate Investigator: Chauncey W. Crandall, MD
Coordinators: David Debottis, RN St. Louis University
Ann Maziarz, RN Principal Investigator: Bernard R. Chaitman, MD
Former Participants: Kim Kelley, RN PTCA Operator: Morton J. Kern, MD
R. R. Lower, MD Surgeons: George Kaiser, MD
Amar Nath, MD Vallee Willman, MD
Bonnie Sechrist, RN Associate Investigator: Robert Wiens, MD
S. Szentpetery, MD Coordinator: Katherine Galan, RN
Staff: Jane Fehl, LPN
University of Michigan Barbara Poole
Principal Investigator: Bertram Pitt, MD Former Participants: Hendrick Barner, MD
PTCA Operators: Eric Bates, MD Ubeydullah Deligonul, MD
Stephen Ellis, MD Michel G. Vandormael, MD
Linda Lee, MD
Eric J. Topol, MD Jewish Hospital (Satellite to St. Louis)
Joseph Walton, MD Principal Investigator: Ronald J. Krone, MD
Surgeons: Steven Bolling, MD Coprincipal Investigator: Nicholas Kouchoukos, MD
Michael Deeb, MD PTCA Operators: Ronald J. Krone, MD
Marvin Kirsh, MD Ali Salimi, MD
Coordinator: Karen Burek, RN Surgeons: Nicholas Kouchoukos, MD
Staff: Linda Belzowski, RN Thomas H. Wareing, MD
Markus Schwaiger, MD Associate Investigators: Patricia Cole, MD
Hui-lee Shu Robert Kleiger, MD
Former Participants: Diane Scarpace, RN Sandor Kovacs, MD
Mack Stirling, MD Michael Rich, MD
Peter Thomasma Anil Shah, MD
Eric J. Topol, MD Coordinators: Mary Caruso, RN
Steve Werns, MD Jane Humphrey, RN
Lisa Spinner, RN
Montreal Heart Institute Staff: Cynthia Benson
Principal Investigator: Martial G. Bourassa, MD Gail Eisenkramer
PTCA Operators: Raoul Bonan, MD Thelma Jones
Gilles Cdt6, MD Ethel Kelly
Jacques Cr6peau, MD Jean Moore
Pierre DeGuise, MD Rose Umstead
Surgeons: Yves Castonguay, MD Juanita Weaver
V-26 Supplement V Circulation Vol 84, No 6 December 1991

BARI/Parallel Study Coordinating Center

Institute of Clinical and Experimental Medicine
Prague, Czechoslovakia University of Pittsburgh, Pittsburgh, Pennsylvania
Principal Investigator: Vladimir Stanek, MD Principal Investigators: Katherine M. Detre, MD,
PTCA Operators: Alfred Belin, MD DrPH
Josef Kovac, MD Sheryl F. Kelsey, PhD
Surgeons: Pavel Firt, MD Coordinator: Kim Sutton-Tyrrell, RN, DrPH
Jan Pirk, MD Assistant Coordinator: Barbara L. Naydeck
Associate Investigators: Vladimir Kocandrle, MD Statisticians: Huiman X. Barnhart, MA
Michael Zelhzko, MD Joyce F. Killinger, MS
Coordinator: Ruzena Jandova, MD Joseph E. Melvin
Data Manager: Erhard Tchernoster, Ing. Allan D. Rosen, MS
Ann Rodewald Steenkiste, MS
Former Site Data Coordinator: Sharon Weber Crow
Data Managers: Regina M. Hardison
Georgetown University Meg Cooper, MEd
Principal Investigator: Kenneth Kent, MD Gail Harger, MSIS
PTCA Operator: Kenneth Kent, MD System Programmers: Dave W. Burry
Surgeons: Nevin M. Katz, MD Jeffrey P. Martin
Robert Wallace, MD Emil A. Maurer
Associate Investigators: Larry Elliott, MD System Support: William P. Amoroso, MPH
Curtis Green, MD Larry Kamons, MSIS
James Lavelle, MD Timothy E. Kuntz, MSIS
Charles Rackely, MD Nancy H. Remaley, MSIS
Coordinator: Beverly Shriver, RN Consultants: Joel Greenhouse, PhD
Robert L. Hardesty, MD
Central Electrocardiographic Laboratory Barry F. Uretsky, MD
Former Participants: Lynette M. Anderson
St. Louis University Medical Center Donald Borrebach, MS
Mary Ann Carr, MS
Principal Investigator: Bernard R. Chaitman, MD Richard Holubkov, MS
Associate Investigators: Robert D. Wiens, MD Diane F. Hursh
Preben Bjerregaard, MD Jennifer A. Metzler
Leonardo Maitas, MD Angela Spadaro
Bonpei Takase, MD John Wilson, PhD
Liwa T. Younis, MD, PhD
Coordinators: Leslee Shaw, MA
Karen Stocke Pirogrnam Office
Programmers: Li-Yung Lui, MS Cardiaac Diiseases Branch
Kim Russell Division of HeDart aand Vascular Diseases
Staff: Steve Cannon National Heartt, Lu]ng, and Blood Institute
Teresa Cutts
Debbie Kargl National Ins ;titutes of Health
Willie Mathis Beti hesdaa, Maryland
Mark Muehl, MA Program Administrator: George Sopko, MD, MPH
Barbara Poole Coinvestigators: Michael Horan, MD
Laura Schaefer Rachel Solomon, M.H.S.
Art Stelken, MA David Zucker, PhD
Jesse Williams Former Participants: Michael J. Domanski, MD
Former Participants: Brian Bilgere Louis Offen, MD
Mary Ellen Hauesen Thomas Robertson, MD
Jeanine Sabatino Joel Verter, PhD
Central Radiographic Laboratory Safety and Data Monitoring Board
Stanford University Medical Center Chair: J. David Bristow, MD, Oregon
Health Sciences University
Principal Investigators: Edwin L. Alderman, MD Members: James Childress, PhD,
Michael Stadius, MD University of Virginia
Associate Investigators: Byron Brown, PhD Timothy Gardner, MD, Johns
William Sanders, MSEE Hopkins Hospital
Lewis Wexler, MD Herman Gold, MD,
Coordinator: Brooke Hollak, RN Massachusetts General
Staff: Terri Beam Hospital
Gao Shao-Zhou, MD J. Ward Kennedy, MD,
Former Participant: Robert Moore, MD University of Washington
Genell Knatterud, PhD,
kncillary Study Maryland Medical Research
Institute
SEQOL: Study of Economics and Quality of Life Robert Roberts, MD, Baylor
Stanford University School of Medicine College of Medicine
John Waldhausen, MD,
Principal Investigator: Mark Hlatky, M.D. Hershey Medical Center
Coordinator: Kathryn Cavanaugh, MPH Carl White, MD, University of
Former Participant: N. Clapp-Channing, MPH, RN Minnesota Hospital
 The BARI Protocol V-27

Morbidity and Mortality Classification Committee National Hospital Discharge Survey Data. Circulation 1989;
Chair: Ronald Prineas, MD, PhD, 79(suppl I):I-13-1-18
University of Miami 2. Harrison DC: Cost containment in mediums: Why cardiology?
Members: Charles Fisch, MD, Krannert Am J Cardiol 1985;56:1OC-15C
In,titute of Cardiology 3. The Expert Panel: Report of the National Cholesterol Educa-
Leoni Greene, MD, tion Program Expert Panel on Detection, Evaluation, and
Harborview Hospital Treatment of High Blood Cholesterol in Adults. Arch Intern
Robert B. Karp, MD, Med 1988;148:36-69
University of Chicago
Medical Center 4. Peduzzi P, Detre KD, Whittes J, Holford T: The intent-to-treat
Spencer B. King III, MD, analysis and the problem of crossovers: An example from the
Emory University Hospital VA Coronary Bypass Surgery Study. J Thorac Cardiovasc Surg
Jay Mason, MD, University of 1991;101:481-487
Utah Medical Center 5. Makuch R, Simon R: Sample size requirements for evaluating
Jack L. Titus, MD, PhD, a conservative therapy. Cancer Treat Reports 1978;62:1037-1040
United Hospital 6. Grizzle J-E: A note on stratifying versus complete random
assignment in clinical trials. Controlled Clin Trials 1982;3:
Office of Study Chair 365-368
Mayo Clinic Foundation 7. Prineas RJ, Crow RS, Blackburn H: The Minnesota Code
Rochester, Minnesota manual of electrocardiographic findings. Boston, John Wright
Robert Frye, MD PGS Inc, 1982
Professor of Medicine 8. Bruce ZA: Exercise testing of patients with coronary artery
Chairman of Internal Medicine disease. Ann Clin Res 1971;3:323-332

References
1. Feinlieb M, Havlik RJ, Zillim RF, Pokras R, McCarthy E, KEY WORDS: * PTCA * multivessel CAD * CABG
Moien M: Coronary heart disease and related procedures: revascularization