NEW ORLEANS--Integrated health care systems, such as HMOs, offer clear
advantages for epidemiologic research, Edward Wagner, MD, MPH, said at
the American Society of Preventive Oncology (ASPO) annual meeting. Dr.
Wagner is director of the Center for Health Studies of Group Health Cooperative,
a Seattle-based HMO.
HMOs have a defined population of enrollees, he said, so that intervention
studies can be done without the use of volunteers, who are more likely
to show a good outcome and skew the results.
Two Group Health intervention studies, one with volunteer smokers and
one with nonvolunteer smokers, showed that smokers who volunteered for
the study were much more likely to quit smoking than nonvolunteer smokers.
Said Dr. Wagner: "There is much to be learned about differences in
volunteers and nonvolunteers in research programs."
HMOs also offer a defined population of providers. The gatekeeper system
improves disease prevention, he said, because intervention is done by an
accountable provider. Computerized data can be linked to outcomes measurements,
and specific providers can be evaluated by comparing their outcomes with
evidence-based practice-based guidelines.
Dr. Wagner illustrated the advantages of defined populations of patients
and accountable providers by citing a low back pain study showing that
physicians providing less pain medication and encouraging return to activity
had better results than physicians who treated low back pain with more
medication and bed rest.
Complete Health Data Bases
HMOs maintain increasingly complete health and health care utilization
data on their enrollees, he said. Because HMOs offer comprehensive services,
these computerized data bases contain information about the full range
of health care received by each patient. "This facilitates studies
relying exclusively or heavily on administrative data bases, which are
much less bothersome to patients and less expensive," he said.
The availability of computerized pharmacy data at Group Health facilitated
the development of a case mix measure--the chronic disease score. "Use
of this measure allows researchers to adjust for the severity of illness,
an important advantage in epidemiologic studies," Dr. Wagner said.
In addition, HMO computerized data bases can be linked with other important
data bases, such as death and disease or regional cancer registries, to
provide further epidemiologic information on each enrollee.
"The research and prevention opportunities of HMOs depend on having
a large, representative population that is stable, receives all services
from the plan, has accountable primary care providers, and makes data accessible,"
he said. "We exploit this at Group Health."
A Different Philosophy
Managed care, with its emerging focus on outcomes measurements, offers
opportunities to provide care that is most beneficial to most of the population,
said another speaker at the ASPO meeting, Robert M. Kaplan, PhD, of the
Department of Family and Preventive Medicine, University of California,
This philosophy differs from that of the traditional biomedical model,
he said. The traditional model emphasizes the diagnosis and treatment of
disease, while the outcomes model focuses on extension of life and quality
of life. "Appropriate diagnosis and treatment should extend life expectancy
and improve quality of life," he said, "but quality of life outcomes
are rarely measured and are sometimes neglected in treatment decisions."
Dr. Kaplan said that there are some circumstances in which aggressive
screening may lead to treatment decisions that reduce quality of life with
no clear evidence that life expectancy is improved. For example, the traditional
model for prostate cancer favors screening older men, while the outcomes
model favors informing men of the complexities and involving them in decision
Properly run managed care is the right setting for pursuing the outcomes
model, since it seeks to maximize benefit to an entire population, Dr.
Kaplan said. He added, however, that this model also emphasizes patient
decision-making and quality of life.