Last January, 13 of the nation's foremost cancer centers formed
tional Comprehensive Cancer Network (NCCN), an alliance that spans
the nation geographically, with the goal of providing high-quality
cost-effective medical services to cancer patients across the
The means to achieve that goal include the development by each
member institution of regional networks so that patients can be
treated where they live, the formation of practice guidelines
to standardize care, and the collection of outcomes data to assess
the effectiveness of treatment.
ONCOLOGY NEWS INTERNATIONAL interviewed physicians involved in
the NCCN for a perspective on practice guidelines and an update
on how the network is proceeding in meeting its goals.
"As oncology care begins to consolidate and become more uniform,
there are advantages to having expert centers come together to
develop mutually acceptable ways of approaching cancer patients,"
said Rodger Winn, MD, of M.D. Anderson Cancer Center.
Steven Rosen, MD, director, Northwestern University Lurie Cancer
Center, and member of the NCCN executive committee, said that
by working together to develop guidelines and collect outcomes
data, the alliance can help insurers and third-party payers get
a handle on the cost of care for specific cancer diagnoses.
"In addition to establishing the most appropriate way to
treat a disease, guidelines give you for the first time an opportunity
to reasonably assess what it costs to provide care for patients,"
Dr. Rosen said.
He noted that the endeavor represents a departure for institutions
with traditional academic ties such as the Lurie Cancer Center.
"External influences are making us revisit how we do business.
To attract patients to our institutions, we must make sure that
patients have the support of their insurance carriers, and to
get that support we must provide cost-effective treatment,"
Dr. Winn emphasized that the guidelines under development allow
physicians flexibility in their treatment approaches. "Any
guideline is just that, a set of recommendations that probably
apply to the majority of patients," he said.
For example, in a certain disease, there may be general agreement
among NCCN members to use multimodality therapy (chemotherapy,
surgery, and radiation), but the precise way of putting together
the three elements or the exact drug regimens used may vary from
institution to institution.
The guidelines panels are multidisci-plinary, including surgeons,
radiation oncologists, and medical oncologists from various NCCN
sites who are recognized experts in the field. Each panel consists
of five to eight members.
The NCCN staff prepares a basic outline or flowchart as a starting
point for the panel to consider at its meeting (an evening and
all-day conference). The panel's goal is to revise the initial
document to reflect what they consider to be the general NCCN
approach. From there, the document goes to each of the 13 member
institutions for revisions and comments.
The revisions from each institution are returned to the expert
panel for collation, and if a large disparity is found between
the panel's recommendations and members' practice, the panel reconvenes
for more discussion.
Why would any busy physician want to sit on yet another committee?
According to Dr. Winn, "most panel members say that it's
fun, and there's very little fun left in medicine today. They
enjoy coming together with their colleagues for a day when the
phone doesn't ring just to discuss and debate issues in their
Assessing the Guidelines
Four categories of acceptance are provided for each element of
the guidelines, to allow physicians to assess how they are to
be used, based on the strength of the recommendation.
- Category 1 elements are areas of treatment where there is
general agreement among most oncologists.
- Category 2 elements represent the NCCN consensus while recognizing
that alternative treatments are available.
- Category 3 elements are temporary guidelines concerning an
area of treatment where the data are not clear and more intensive
analysis is needed.
- Category 4 elements represent areas in which there is disagreement
between NCCN institutions about the preferred treatment. Physicians
must use their own clinical judgment as to treatment; enrollment
in clinical trials is encouraged.
"The guidelines are built on the premise that it is always
appropriate to put a patient on a sound clinical trial,"
Dr. Winn said. "They never preempt clinical trials, but the
only time this is actually mentioned is in areas where there does
not appear to be any other real option."
He added that for categories 2, 3, and 4, the guidelines will
be annotated so that physicians will know the thinking behind
Level of Complexity
Because panel members are expert subspecialists, the NCCN guidelines
are reaching a level of complexity beyond that of most such efforts,
Dr. Winn said.
He used small-cell lung cancer as an example. In their preliminary
deliberations the panel working on the guidelines for this disease
agreed that radiation therapy should be considered after chemotherapy,
but they also went into such details as to when to irradiate and
whether to irradiate to a field the size of the original tumor
or to narrow the field to encompass only the remaining tumor.
He noted that guidelines for cancer prevention will come later,
after the treatment guidelines are completed.
Catherine Harvey, Dr.PH, chief operating officer for the NCCN,
told ONCOLOGY NEWS INTERNATIONAL that currently six clinical guidelines
for the common cancers have been through expert panel and another
four are expected to be finished in the next few months. By early
next year, she expects that the first NCCN guidelines will be
in testing at all member institutions.
She noted that guidelines will be updated on a regular schedule,
but will also be revised any time new clinical data become available.
Dr. Harvey referred to the NCCN clinical guidelines as "full
course of care guidelines that go from diagnosis to death or cure.
It's a way for our member institutions to assure that the cancer
patients who are not eligible for clinical trials receive the
best possible standard therapy available."
Outcomes Data Collection
Appropriate outcome measures are needed to allow the quality of
care of NCCN members to be assessed and compared "among ourselves
and to any other national benchmarks," Dr. Winn said.
Such measures, currently under development by NCCN committees,
will include "process measures," he said. "For
example, what percentage of our stage II breast cancer patients
receive adjuvant chemotherapy? We know from national figures it
may be as low as 65% or 70%. We need to know if our own figures
within the NCCN are higher." Another example would be the
percentage of breast cancer patients in the network who receive
The outcomes study that is furthest along, Dr. Winn said, asks
what percentage of NCCN's T3 and T4 larynx cancer patients undergo