vate health plans’ coverage and reimbursement decisions.
The NETT was conducted and analyzed in accordance with standard practice
for clinical trials. Despite, or perhaps because of, the rigorous design, the main
outcome of the study was mixed, and many felt that the results did not show con-
vincing proof of meaningful benefit. The economic analysis also suggested that the
procedure had relatively poor cost-effectiveness in the short run. Nevertheless, a
few months after the release of the trial results, Medicare agreed to cover LVRS,
and designated surgery centers have begun marketing the procedure to the public.1
In this paper we describe the history of LVRS and the NETT, and how this his-
tory and the results of the trial influenced the CMS’s coverage decision for LVRS.
We place this decision in the context of the new CMS initiative in evidence-based
medicine and other recent coverage decisions for costly procedures. We conclude
by describing the implications of the NETT study for future studies of expensive
surgical procedures that will primarily affect Medicare beneficiaries.
History Of Lung Volume Reduction Surgery
The history of LVRS prior to the NETT is an excellent example of a convergence
of clinical, economic, and legal factors that foster rapid diffusion of a technology in
advance of evidence of its efficacy and cost-effectiveness.
For tens of thousands of patients with severe emphysema, quality of life is often
poorer than for patients with cancer, coronary artery disease, and AIDS.2
members often must care for emphysema patients’ basic needs, including cooking,
cleaning, bathing, and dressing. Standard medical treatment offers only modest
symptom relief, and no intervention except smoking cessation early in the course
of the condition has been shown to substantively alter the course of disease. As
with other serious medical conditions, this combination of poor quality of life
coupled with lack of effective therapies creates an intense demand among patients
and their families for new technologies that promise symptom relief, improved
survival, or both, regardless of the cost.
Originally developed in the 1950s, LVRS is a surgical approach to reducing the
volume of the hyperinflated lung parenchyma that is characteristic of severe em-
physema. Although 75 percent of patients reported clinical improvement in an
early case series, the lack of objective documentation of benefit and an operative
mortality of 18 percent prevented widespread acceptance of the procedure.3
1995 Joel Cooper and colleagues reported a modification of the original volume re-
duction operation in which diseased lung tissue was removed from both lungs via
a surgical procedure called a median sternotomy. After favorable results from a
small, uncontrolled study of patients with severe emphysema who underwent
LVRS were reported in 1994, it was not surprising that the pressures to adopt this
technology were strong.4
Patient advocates posted testimonials on a thoracic sur-
gery Web site, and favorable reports appeared in the popular press.5
When LVRS was gaining popularity, it had no specific billing code. Physicians
5 6 J a n u a r y / F e b r u a r y 2 0 0 5
C a s e S t u d i e s
and hospitals thus obtained reimbursement by submitting preexisting but non-
specific billing codes for multiple resections of lung tissue. LVRS is neither techni-
cally difficult nor costly, nor does it require the purchase of costly capital equip-
ment. Because Medicare’s reimbursement for the nonspecific lung procedures was
fairly generous, there were few financial disincentives to performing the operation.
Given a prevalent, severely disabling disease; a technically simple, low-cost,
and potentially effective procedure; and a relatively generous reimbursement cli-
mate, it is perhaps not surprising that it took little time for several institutions
throughout the country to develop LVRS programs. In less than eighteen months
from the initial published reports, more than 1,200 LVRS procedures had been per-
formed on Medicare beneficiaries nationwide.6
These procedures were also likely
performed on patients in private health plans during the same time period.
n Regulatory response: the National Emphysema Treatment Trial. After
identifying sharp increases in heretofore little used claims for lung resection in
1994–95, the CMS asked AHRQ to conduct an evidence review of LVRS. In addition,
the CMS learned from an internal claims review that mortality at six months for pa-
tients undergoing LVRS approached 17 percent.7
Based on this finding and the
AHRQ review, which found no compelling evidence of benefit, the CMS announced
in December 1995 that Medicare would not pay for LVRS, on the grounds that insuf-
ficient safety and efficacy evidence was available to support reimbursement.8
To what degree did economic considerations figure into the CMS’s decision to
suspend reimbursement for LVRS? Certainly, agency officials were concerned that
hundreds of millions of dollars would be spent on a procedure whose clinical
value was unknown and that might be ineffective or inferior to medical therapy.9
On the other hand, the political cost of denying patients access to a potentially
beneficial procedure was likely to be great. With many uncertainties in play, regu-
latory officials made an unprecedented decision: When the CMS announced that
it would suspend payment for LVRS, it also announced that it would collaborate
with the National Institutes of Health (NIH), which would sponsor a multicenter
randomized controlled trial (RCT) to evaluate the efficacy of LVRS in conjunction
with the NHLBI. Furthermore, the CMS agreed to pay for the costs of the proce-
dure, an exception to its usual policy of paying only for routine medical care that is
part of clinical trials.10
Apparently, the CMS did not require additional regulatory
or statutory authority to provide coverage for the procedure during the trial.
n Science meets politics in a clinical trial. After a competitive, peer-reviewed
bidding process sponsored by the NHLBI, eighteen centers around the United
States were ultimately invited to participate, including Washington University in
St. Louis, home of the rejuvenator of the procedure, Joel Cooper. Given the obvious
L u n g S u r g e r y
H E A LT H A F F A I R S ~ Vo l u m e 2 4 , N u m b e r 1 5 7
“The political ramifications of the CMS decision became apparent
even before the trial had enrolled its first patient.”
economic implications of LVRS, several centers proposed substudies to conduct
cost-effectiveness analyses alongside the trial. AHRQ responded by providing fund-
ing for such an analysis. The CMS supported the cost-effectiveness study by provid-
ing claims and reimbursement records for trial participants.
The political ramifications of the CMS decision became apparent even before
the trial had enrolled its first patient. In the fall of 1996 Congress passed a law re-
quiring the CMS to provide a report that included an analysis of all available data
for LVRS and recommendations as to the appropriateness and conditions of Medi-
care coverage of the procedure. On 17 April 1997, at hearings conducted in part to
provide a forum for those concerned with Medicare’s decision to cover LVRS only
under the NETT, Rep. Jim Ramstad (R-MN) expressed “concern that the decision
not to extend Medicare coverage for the procedure [LVRS] could negatively affect
the lives of thousands of older Americans who suffer from this disease.”11
same hearing, Cooper also criticized the CMS’s decision to deny payment for
LVRS outside of the trial.
Just prior to the start of the trial, Cooper announced that Washington Univer-
sity would withdraw from the multicenter study, on the grounds that its institu-
tional review board (IRB) felt that a controlled study was unethical, given what
they saw as obvious benefits of the procedure.12
Cooper continued to lobby vigor-
ously for the procedure (and against the trial) during the course of the study, lik-
ening the controlled trial to “hiding behind a scientific study to ration the proce-
Nevertheless, despite the political difficulties, in October 1997 seventeen
centers began enrolling patients.
Trials And Tribulations
The Statistical Coordinating Center at the Johns Hopkins University and par-
ticipating centers were responsible for ensuring the quality of the trial design and
conduct of the study.14
The problem of crossover—patients randomized to the
medical care arm who later obtain surgery outside of the trial—was a major con-
cern at the start of the trial because of the level of publicity surrounding the proce-
dure. Crossover ultimately proved inconsequential. Loss of patients at follow-up,
another problem that can plague large clinical trials, was minimal: At the time of
publication of the main outcome paper, 99 percent of all surviving patients con-
tinued to complete the telephone or clinic visits.
For all of its success, the NETT was not without challenges. Patient recruit-
ment proved difficult. The original enrollment goal of 4,500 patients was later re-
duced to 2,500 when it became clear that recruitment was going to be far below
When interim analysis identified a subset of patients with excess mor-
tality and no apparent benefit from surgery, many cited this outcome as evidence
that the procedure was unsafe, further adding controversy to the trial and reduc-
ing the number who presented for screening.16
Ultimately, 1,218 patients were ran-
domized. Interestingly, slow enrollment halted a planned clinical trial in Canada
5 8 J a n u a r y / F e b r u a r y 2 0 0 5
C a s e S t u d i e s
to be conducted roughly within the same time frame as the NETT. The Canadian
researchers also planned a cost-effectiveness analysis that would ostensibly in-
form coverage policy.
In May 2003 the NETT research group published its main findings, which
showed no improvement in survival for surgical patients but differential improve-
ments in exercise capacity and respiratory-specific quality of life for a percentage
of the surgical patients.17
A secondary analysis suggested differential relative bene-
fits for patient subgroups defined by combinations of regional predominance in
emphysema distribution (by computed tomography [CT] scan); and (2) exercise
capacity following pulmonary rehabilitation. One subgroup showed statistically
significant improvements in survival for the surgical cohort. Surgical patients in
three of the four subgroups had improvements in exercise capacity and quality of
life compared with medical therapy patients.
The cost-effectiveness of LVRS compared with medical therapy during the
three years following initiation of treatment was judged to be unfavorable, at
$189,000 per quality-adjusted life year (QALY) gained, compared with the often-
cited conventional level of $50,000 per QALY.18
In the subgroups, the cost-
effectiveness analysis was more favorable for patients with upper-lobe emphy-
sema and low exercise capacity but less favorable in the two remaining subgroups.
Both the clinical and cost-effectiveness results were published simultaneously
in the New England Journal of Medicine, an unprecedented occurrence that under-
scored the interest in both the clinical and economic outcomes for the NETT.
Given the relatively modest clinical benefits for the main analysis, uncertain sig-
nificance of the subgroup analysis, and poor short-term cost-effectiveness, the ac-
companying editorial was relatively circumspect in its assessment of the role of
the procedure in clinical practice.19
The CMS faced a difficult decision about cov-
erage, heightened by uncertainty about the rate at which LVRS would be adopted
after the trial.
The CMS’s Coverage Decision
At the same time that the trial results were published, the CMS released a na-
tional coverage memorandum for LVRS for comment. LVRS would be covered for
patients with upper- or lower-lobe predominant emphysema with low exercise
capacity, or upper-lobe emphysema with high exercise capacity. These groups
constituted three of the four subgroups from the trial for whom benefits were
identified. The CMS originally specified that LVRS is reasonable and necessary
only when performed at facilities that were identified by the NHLBI as meeting
the thresholds for participation in the NETT, although it noted that it was particu-
larly interested in comments on the “potential for restricting LVRS to certain fa-
cilities that meet predefined criteria.” The final rule was amended to include sites
that have been approved by Medicare as lung transplant facilities.20
The CMS im-
plemented its final coverage policy 5 January 2004.
L u n g S u r g e r y
H E A LT H A F F A I R S ~ Vo l u m e 2 4 , N u m b e r 1 5 9
n Concerns about the CMS’s policy. Several analysts have noted that the
CMS’s relatively generous coverage decision for LVRS seems disproportionate to
both the weight of the evidence and incremental benefit for LVRS patients shown in
Findings from subgroup analyses of a single clinical trial are considered
relatively weak evidence, particularly in situations where the trial fails to show a hy-
pothesized benefit for the experimental group. Certainly, the U.S. Food and Drug
Administration (FDA) would not approve drugs in such instances; rather, the
agency would request that the manufacturer conduct a separate trial that consid-
ered only the subgroup(s) in question. Because there was no prior hypothesis of or
biological rationale for differential benefit based on distribution of emphysema and
presurgical exercise capacity, the post hoc subgroup analysis is open to criticisms
that the investigators undertook a “data dredging” exercise to identify groups with
the findings they sought: improvement in survival, exercise capacity, and quality of
life. As a practical concern, the CMS’s coverage policy excludes only patients with
high exercise capacity prior to surgery. Given that exercise ergometry is technically
difficult to standardize and relatively easy to manipulate (patients wanting LVRS
could simply not exercise as hard), it may be difficult to limit LVRS based on exer-
cise capacity. In effect, the CMS may have designed a coverage policy that includes
all patients meeting the original NETT eligibility criteria.
Limiting LVRS reimbursement to NETT centers and Medicare-designated lung
transplant facilities could also be criticized as being inconsistent with evidence
and the CMS’s own rationale for limiting high-risk procedures to designated cen-
ters of experience. Studies have shown that patients undergoing high-risk surgi-
cal procedures at institutions and by surgeons with the most experience have
better outcomes than patients at lower-volume facilities.22
The NETT surgical
centers were probably, with one exception, the only centers in the United States
that had substantive experience with the procedure.23
When the CMS suspended
coverage for LVRS, the number of procedures performed outside of the NETT fell
to a trickle, as most private insurers suspended coverage while awaiting results
from the trial.24
Expanding coverage to include Medicare-designated lung trans-
plant centers (that were not also NETT facilities) thus is potentially problematic,
since few surgeons at those centers had recent experience with the procedure.
n What shaped the CMS’s coverage policy? The issues noted above might
lead one to conclude that in suspending its evidence policy for LVRS, the CMS was
bowing to political, advocacy, professional, and other pressures. Although there is
substance to this line of reasoning, it is also important to consider pragmatic issues
and the state of the evidence.
First, patients in the surgical arm of the NETT did demonstrate statistically sig-
nificant improvements in pulmonary functioning, exercise capacity, and respira-
tory-specific quality of life compared with those in the medical therapy arm.
Although one might argue that the clinical significance was modest, such distinc-
tions would not necessarily be valid points of contention unless the sponsors had
6 0 J a n u a r y / F e b r u a r y 2 0 0 5
C a s e S t u d i e s
decided a minimal clinical difference beforehand. Interestingly, neither the CMS
nor the trial sponsors defined clinical significance thresholds for functional or
quality-of-life endpoints at the onset of the trial.
Second, one of the early stated goals of the NETT was to refine the eligibility
criteria for the procedure—that is, to determine whether it was possible to iden-
tify subpopulations with differential benefit. Although those subpopulations did
not turn out to be the ones that were initially postulated, it was clear at the onset
that there was much to learn about this procedure.
Third, the NETT is certainly going to be the only large RCT of LVRS with
long-term endpoints. The trial cost more than $125 million, including procedural,
patient care, and research costs.25
Given such expense, and the well-recognized
recruitment difficulties, there is little interest in mounting a second controlled
study to validate the findings. It is difficult to expect patients, clinicians, or health
insurers to wait for more evidence on subgroups, since no trial of the magnitude
and rigor of the NETT will be conducted hence.
There are also arguments in favor of expanding coverage to more surgical cen-
ters. During the trial, it became clear that many patients and their families had
great difficulty traveling many miles to receive care. The evaluation process to de-
termine eligibility often took days, and for those enrolled, six to ten weeks of pul-
monary rehabilitation were required prior to surgery. The seventeen NETT cen-
ters leave many geographic “holes” without LVRS centers for hundreds of miles.
Expanding coverage to the forty-four lung transplant facilities in the United
States will greatly improve access both to the evaluation and rehabilitation proc-
ess and to the procedure itself. There are short-term concerns about experience,
but these are mitigated by the fact that lung transplant centers are already skilled
at caring for patients with severe respiratory disease.
Did political considerations ultimately influence the CMS’s coverage policy? If
the CMS deemed it appropriate to base coverage on subgroup analyses, it could
have restricted coverage to those with upper-lobe disease and low exercise capac-
ity, since this subgroup showed both a mortality gain and a much greater benefit
of exercise and quality of life compared with the other subgroups. Because less
than 25 percent of all patients in the trial met these criteria, such a restriction
would have greatly reduced potential procedure volume nationwide.
Second, expanding coverage to include lung transplant centers allowed the
CMS to permit LVRS at the home of its most vocal proponent, Joel Cooper.
Cooper originally proposed expanding coverage to CMS-designated lung trans-
plant centers (Washington University in St. Louis is a designated center).
n Did cost-effectiveness play a role in the decision? Many experts agreed be-
forehand that the cost-effectiveness of LVRS should be considered relative to out-
comes and costs for other medical and surgical procedures.26
Although the CMS did
support the cost-effectiveness study that was conducted alongside the trial, there is
little evidence that the results of the cost-effectiveness analysis played an important
L u n g S u r g e r y
H E A LT H A F F A I R S ~ Vo l u m e 2 4 , N u m b e r 1 6 1
role in its coverage decision. Calculated over the observation period for the trial, the
cost-effectiveness of LVRS was unfavorable compared with that of other common
and reimbursed thoracic surgical procedures. The projected long-term cost-effec-
tiveness (derived through a statistical modeling approach) is somewhat more favor-
able but also quite uncertain, largely because of clinical uncertainty about the dura-
tion of any survival or quality-of-life benefit. If these considerations were taken into
account, use of LVRS would likely be greatly restricted.
CMS officials have stated that costs do not play a role in their coverage deci-
Although there is no specific regulatory language prohibiting the CMS
from considering economic evidence in its decisions, past attempts to incorporate
economics in the regulatory process have failed, largely because of political con-
Indeed, as has been noted, adopting economic criteria implies that the
CMS will deny a clinical benefit because of cost, a task that is much more difficult
politically than even the current sensitive task of weighing evidence of clinical
risks and benefits.29
An important precedent of the NETT was the CMS’s willingness to cosponsor
the cost-effectiveness component of study. This may reflect the agency’s interest
in developing a process of conducting economic analyses alongside clinical trials
that have great importance to their covered population.
Implications For Future Studies
Surgical procedures are essentially unregulated in the United States; as a result,
many new procedures enter clinical practice with little evidence that allows
meaningful assessment of their clinical benefits and risks. LVRS demonstrated
that when a new surgical procedure is designed to treat a highly prevalent, se-
verely disabling condition with few therapeutic alternatives, patient and physi-
cian demand can “drive” utilization well ahead of evidence of benefit, and serious
risks become known only after many patients have been treated. Because of these
issues, the CMS’s decision to suspend payment and sponsor a clinical trial of LVRS
is a critical policy development. Will the NETT experience become a paradigm for
future surgical trials?30
We think that this will depend on a number of lessons that
can be drawn from the trial.
n Agency collaboration. First, collaboration between the CMS and other fed-
eral agencies that evaluate new medical technologies is important. A regulatory
“Bermuda Triangle” often exists between the CMS, NIH, and FDA, where important
devices and procedures are implicitly or explicitly covered, then enter widespread
use prior to the availability of high-quality evidence supporting their clinical effec-
tiveness and safety.31
The FDA and NIH are in the best position to alert the CMS
when important technological innovations are moving from research settings to
clinical use, particularly technologies that are licensed and regulated by the FDA or
invented with NIH funding. Paradoxically, it was the CMS’s own internal audits
that found sizable mortality concerns for patients undergoing LVRS. The three
6 2 J a n u a r y / F e b r u a r y 2 0 0 5
C a s e S t u d i e s
agencies would benefit from periodic, coordinated reviews of the medical technol-
ogy landscape so that proactive evaluations can take place for important emerging
procedures. Part of the review procedure may be developing criteria that would trig-
ger a recommendation for conducting jointly sponsored, large-scale RCTs.
n Consensus about acceptable quality of evidence. Second, in cases where it
appears that a clinical trial is desirable, the CMS should work early on with external
investigators and the sponsoring federal research agency (such as the NHLBI) to de-
velop consensus about the need for high-quality evidence. This is important both for
gaining acceptance from the clinical community and for elucidating all relevant is-
sues pertaining to the procedure. Prior to the CMS’s decision to suspend payment
for LVRS, it cosponsored a workshop with the NHLBI where experts from many
disciplines agreed that the state of knowledge regarding LVRS was such that a
multicenter RCT was needed.
n External peer review. Third, high-quality external peer review is necessary to
ensure the integrity of the process of selecting centers to participate in a trial. As the
NETT also demonstrated, this does not ensure that those who are successful appli-
cants will be neutral in their appraisal of the state of the technology. Indeed, those
with the most experience in experimental procedures (and thus most qualified to
participate in a trial) are often the most vocal proponents. As occurred with Cooper
and the NETT, such proponents often state that the evidence is such that controlled
trials are unethical. It is important to counter such arguments forcefully during the
design phase of the study, as they are part of a longstanding double standard that ex-
ists for evaluating surgical (versus medical and pharmaceutical) treatments. Publi-
cation of a consensus statement from the aforementioned NHLBI/CMS workshop
on the need for a randomized trial would have been helpful to counter criticism from
those who felt that a trial was unwarranted, and would have served as a reference for
IRBs that might otherwise be influenced by advocates for adopting new technolo-
gies prior to thorough evaluation. The arguments against controlled surgical trials
have been refuted on both ethical and scientific grounds.32
A few of the many threats
to study design that occur in discussions of surgical trials include differential eligi-
bility criteria for cases and controls, use of inappropriate or sham treatments as con-
trols, and allowing patients who are randomized to the nonexperimental arm to
cross over to the treatment arm. Indeed, such design problems were apparent in a
privately funded study that was initiated to evaluate LVRS (the study failed to
achieve recruitment goals and was cancelled).33
In the case of LVRS, an argument
could be made that blinding (that is, a sham surgery arm) would have improved the
rigor of the study by removing the placebo effect of surgery. Such sham-controlled
trials have often yielded surprising results and called into question the therapeutic
value of the procedure.34
Because the NETT was not a blinded study, it is open to
criticism that quality-of-life differences between treatment and control can be ac-
counted for in part by the placebo effect.
n Predefined “significant clinical benefit.” We also recommend that clinical
L u n g S u r g e r y
H E A LT H A F F A I R S ~ Vo l u m e 2 4 , N u m b e r 1 6 3
and regulatory officials work together to prespecify levels of clinical benefit that
would be considered clinically as well as statistically significant. Trials would then
be powered to test hypotheses against these levels of difference. Applying a defini-
tion of clinical significance to the treatment and patient population in question
would be a valid interpretation of Medicare’s regulatory language specifying that
“reasonable and necessary” services should be covered.35
When multiple clinical
endpoints are being considered, the clinical study should be designed to have
enough patients to test the primary and secondary hypotheses as well as any
planned subgroup analyses. In the case of the NETT, gathering opinions about a
clinically meaningful difference in exercise ergometry beforehand would have re-
duced problems with interpreting the findings later. Of note, economic evaluation
could also be used to inform the level of benefit that is considered reasonable value
for money. For example, economic modeling could have been used to translate exer-
cise thresholds and projected changes in life-expectancy gains into QALYs, the most
widely used metric of benefit in cost-effectiveness analysis. Expected intervention
costs could be assessed relative to benefits compared to standard therapy. Such a
process would avoid cost-effectiveness analysis thresholds that are anathema to many.
n Costs of experimental treatment. Another issue is whether the CMS should
pay the costs of the experimental treatment. There are several reasons why the CMS
should consider making such payment a rule rather than an exception for studies of
surgical procedures. Unlike drugs and devices, surgical procedures typically do not
have sponsors that can cover the costs of treatment, as drug and device manufactur-
ers do. By paying for procedures in trial settings, the CMS lowers an important fi-
nancial barrier to participation. Patient recruitment can be increased, and surgeons
do not have to bear the expense (both real and in terms of opportunity costs) of do-
nating their time during the evaluation phase. Larger trials with power to test multi-
ple endpoints would become more feasible. In addition, surgeons who have less con-
flict of interest could be recruited to participate. Of note, Cooper is a major
stakeholder in the company that manufactured the strips used to seal the lungs of
LVRS patients. Such obvious financial conflicts can run afoul of rules regarding fed-
erally sponsored clinical trials.
n Long-term follow-up. Clinical trials are by design of limited duration, but tri-
als conducted in collaboration with the CMS have the advantage of permitting
long-term follow-up through claims and Medicare vital status files. The opportu-
nity for post-trial evaluation using observational epidemiology methods and claims
data analysis should be strongly considered in cases where the treatment is ex-
pected to influence patient outcomes long after the study is completed or there is a
need to monitor use and outcomes for patients after the coverage decision has been
made. Such evaluations would require that the CMS adopt an intra- or extramural
research program or work with AHRQ or the NIH to undertake evaluations of cov-
6 4 J a n u a r y / F e b r u a r y 2 0 0 5
C a s e S t u d i e s
1. G. Kolata, “Can a Unique Surgical Procedure Transform the Treatment of Emphysema?” New York Times
Magazine, 16 May 2004, 21.
2. R.M. Kaplan, C.J. Atkins, and R. Timms, “Validity of a Quality of Well-Being Scale as an Outcome Measure
in Chronic Obstructive Pulmonary Disease,” JournalofChronicDisease 37, no. 2 (1984): 85–95; R.M. Kaplan et
al., “Validity of the Quality of Well-Being Scale for Persons with Human Immunodeficiency Virus Infec-
tion: HNRC Group, HIV Neurobehavioral Center,” Psychosomatic Medicine 57, no. 2 (1995): 138–147; T.G.
Ganiats, L.A. Palinkas, and R.M. Kaplan, “Comparison of Quality of Well-Being Scale and Functional Sta-
tus Index in Patients with Atrial Fibrillation,” Medical Care 30, no. 10 (1992): 958–964; and R.M. Kaplan,
“Quality of Life Assessment for Cost/Utility Studies in Cancer,” CancerTreatmentReview 19, Supp. A (1993):
3. O. Brantigan and E. Mueller, “Surgical Treatment of Pulmonary Emphysema,” American Surgery 23, no. 9
4. E. Trulock and J.D. Cooper, “Reduction Pneumoplasty for COPD” (abstract), Chest 106, no. 3 (1994): 52S.
5. See American College of Surgeons, “Advocacy and Health Policy,” 2 June 2004, www.facs.org/ahp/in-
dex.html (14 October 2004); J. Marquis, “Pioneered Surgery Gives Science Pause, Patients Relief,” Los An-
geles Times, 28 August 1994; and C. Groman, “Are Surgeons Too Creative?” Time, 4 September 1995.
6. H.F. Huizenga, S.D. Ramsey, and R.K. Albert, “Estimated Growth of Lung Volume Reduction Surgery
among Medicare Enrollees: 1994 to 1996,” Chest 114, no. 6 (1998): 1583–1587.
7. Agency for Healthcare Research and Quality, Lung Volume Reduction Surgery for End-Stage Chronic Obstructive
Pulmonary Disease, Technology Assessment Report, Pub. no. 96-0062 (Rockville, Md.: AHRQ, 1996).
8. Centers for Medicare and Medicaid Services, CoverageIssuesManual:MedicalProcedures, section 35-93, 16 Sep-
tember 2004, www.cms.hhs.gov/manuals/06_cim/ci35.asp#_1_97 (14 October 2004); and T.V. Holohan and
H. Handelsman, “Lung-Volume Reduction Surgery for End-Stage Chronic Obstructive Pulmonary Dis-
ease,” Health Technology Assessment, no. 10, AHCPR Pub. no. 96-0062, September 1996, www.ncbi.nlm.nih
.gov/books/bv.fcgi?rid=hstat6.chapter.41412 (12 November 2004).
9. C. Gentry, “Second Opinion: Why Medicare Covers a New Lung Surgery for Just a Few Patients,” Wall
Street Journal, 29 June 1998.
10. CMS, Coverage Issues Manual: Clinical Trials, section 30-1, 16 September 2004, www.cms.hhs.gov/manuals/
06_cim/ci30.asp#_1_1 (14 October 2004).
11. Rep. Jim Ramstad, “Issues Relating to Medicare’s Coverage Policy,” Prepared Statement for House Com-
mittee on Ways and Means, 17 April 1997.
12. J.D. Cooper, “Paying the Piper: The NETT Strikes a Sour Note—National Emphysema Treatment Trial,”
Annals of Thoracic Surgery 72, no. 2 (2001): 330–333.
13. Gentry, “Second Opinion.”
14. National Emphysema Treatment Trial Research Group, “Rationale and Design of the National Emphysema
Treatment Trial (NETT): A Prospective Randomized Trial of Lung Volume Reduction Surgery,” Journal of
ThoracicCardiovascularSurgery 118, no. 3 (1999): 518–528; and S.D. Ramsey et al., “Economic Analysis of Lung
Volume Reduction Surgery as Part of the National Emphysema Treatment Trial: NETT Research Group,”
Annals of Thoracic Surgery 71, no. 3 (2001): 995–1002.
15. The reasons why enrollment fell far below expectations were never clearly identified. Several factors were
hypothesized: patients’ unwillingness to undergo randomization; an extensive screening process that of-
ten took several days; cost to patients of undergoing screening (subsidies were later provided to defray
travel and lodging costs for those who had to travel great distances to NETT centers); and the high num-
ber of exclusion criteria.
16. NETT Research Group, “Patients at High Risk of Death after Lung-Volume-Reduction Surgery,” New Eng-
land Journal of Medicine 345, no. 15 (2001): 1075–1083.
17. A. Fishman et al., “A Randomized Trial Comparing Lung-Volume-Reduction Surgery with Medical Ther-
apy for Severe Emphysema,” New England Journal of Medicine 348, no. 21 (2003): 2059–2073.
18. NETT Research Group, “Cost Effectiveness of Lung-Volume-Reduction Surgery for Patients with Severe
Emphysema,” New England Journal of Medicine 348, no. 21 (2003): 2092–2102.
L u n g S u r g e r y
H E A LT H A F F A I R S ~ Vo l u m e 2 4 , N u m b e r 1 6 5
19. J.H. Ware, “The National Emphysema Treatment Trial—How Strong Is the Evidence?” NewEnglandJournal
of Medicine 348, no. 21 (2003): 2055–2056.
20. CMS, “Facilities Eligible for Reimbursement for Lung Volume Reduction Surgery,” www.cms.hhs.gov/
coverage/lvrsfacility.pdf (14 October 2004).
21. M.R. Gillick, “Medicare Coverage for Technological Innovations—Time for New Criteria?” New England
Journal of Medicine 350, no. 21 (2004): 2199–2203.
22. P.D. McGrath et al., “Relation between Operator and Hospital Volume and Outcomes following Percuta-
neous Coronary Interventions in the Era of the Coronary Stent,” Journal of the American Medical Association
284, no. 24 (2000): 3139–3144; S.D. Ramsey et al., “Clinical and Economic Effects of Pulmonary Artery
Catheterization in Nonemergent Coronary Artery Bypass Graft Surgery,” Journal of Cardiothoracic Vascular
Anesthesiology 14, no. 2 (2000): 113–118; and J.A. Showstack et al., “Association of Volume with Outcome of
Coronary Artery Bypass Graft Surgery: Scheduled versus Nonscheduled Operations,” JournaloftheAmerican
Medical Association 257, no. 6 (1987): 785–789.
23. The one notable exception is Washington University in St. Louis, where Joel Cooper performed nearly
dozens of procedures on Medicare-eligible patients during the years of the NETT. B.F. Meyers et al., “Re-
sults of Lung Volume Reduction Surgery in Patients Meeting a National Emphysema Treatment Trial
High-Risk Criterion,” JournalofThoracicCardiovascularSurgery 127, no. 3 (2004): 829–835. Cooper was able to
obtain reimbursement from the CMS for many Medicare patients on the basis of medical necessity. This
exception requires the ruling of a local adjudicator.
24. A small trial of LVRS sponsored by a private health insurer started at approximately the same time as the
NETT was suspended after failing to meet recruitment goals. R.L. Berger et al., “Limitations of Random-
ized Clinical Trials for Evaluating Emerging Operations: The Case of Lung Volume Reduction Surgery,”
Annals of Thoracic Surgery 72, no. 2 (2001): 649–657. A few procedures were also performed at Veterans Af-
fairs (VA) Medical Centers during this time period.
25. T. Carino, S. Sheingold, and S. Tunis, “Using Clinical Trials as a Condition of Coverage: Lessons from the
National Emphysema Treatment Trial,” Clinical Trials 1, no. 1 (2004): 108–121.
26. AHRQ, Lung Volume Reduction Surgery; J.M. Drazen, “Surgery for Emphysema—Not for Everyone,” New Eng-
land Journal of Medicine 345, no. 15 (2001): 1126–1127; and A.M. Fein, “Lung Volume Reduction Surgery: An-
swering the Crucial Questions,” Chest 113, no. 4 Supp. (1998): 277S–282S.
27. G. Kolata, “Newest Treatments Create a Quandary on Medicare Costs,” New York Times, 17 August 2003.
28. Gillick, “Medicare Coverage for Technological Innovations.”
29. S.R. Tunis, “Why Medicare Has Not Established Criteria for Coverage Decisions,” New England Journal of
Medicine 350, no. 21 (2004): 2196–2198.
30. D.E. Wood and M.M. DeCamp, “The National Emphysema Treatment Trial: A Paradigm for Future Surgi-
cal Trials,” Annals of Thoracic Surgery 72, no. 2 (2001): 327–329.
31. L. Kessler et al., “Clinical Use of Medical Devices in the ‘Bermuda Triangle,’” Health Affairs 23, no. 1 (2004):
32. M.R. Tonelli, J.O. Benditt, and R.K. Albert, “Clinical Experimentation: Lessons from Lung Volume Reduc-
tion Surgery,” Chest 110, no. 1 (1996): 230–238.
33. Berger et al., “Limitations of Randomized Clinical Trials”; and G.L. Snider, “Health-Care Technology As-
sessment of Surgical Procedures: The Case of Reduction Pneumoplasty for Emphysema,” AmericanJournalof
Respiratory Critical Care Medicine 153, no. 4, Part 1 (1996): 1208–1213.
34. J.B. Moseley et al., “A Controlled Trial of Arthroscopic Surgery for Osteoarthritis of the Knee,” NewEngland
JournalofMedicine 347, no. 2 (2002): 81–88; L.A. Cobb et al., “An Evaluation of Internal-Mammary-Artery Li-
gation by a Double-Blind Technic,” NewEnglandJournalofMedicine 260, no. 22 (1959): 1115–1118; E.G. Dimond,
C.F. Kittle, and J.E. Crockett, “Comparison of Internal Mammary Artery Ligation and Sham Operation for
Angina Pectoris,” American Journal of Cardiology 5 (1960): 483–486; and M. Saririan and M.J. Eisenberg,
“Myocardial Laser Revascularization for the Treatment of End-Stage Coronary Artery Disease,” Journal for
the American College of Cardiology 41, no. 2 (2003): 173–183.
35. U.S. Government Social Security Act Amendments, Sec. 1862(a)(1)(A), enacted 30 July 1965.
6 6 J a n u a r y / F e b r u a r y 2 0 0 5
C a s e S t u d i e s