“People with diabetes can ‘get by,’ living with their
disease for years on end until they go blind in an eye or lose sensation
in a foot,” said David McCulloch, M.D., professor of medicine at
the University of Washington. McCulloch and others presented their findings
at the American Association for the Advancement of Science’s annual
conference over the weekend.
McCulloch also works at the Group Health Cooperative of Puget Sound, where
he has introduced a computerized system to aggressively monitor patients
and issue reminders to physicians.
“We don’t have health care systems, we have illness treatment
systems,” McCulloch said. Such systems may be fine for diagnosing
and treating acute conditions, he said, but they don’t work well
for long-term management of chronic diseases like diabetes.
“Diabetic patients show up at the wrong time,” he said, “… when
they have a problem.”
Regardless of when they sought treatment, McCulloch found, patients saw
primary care physicians instead of specialists 90 percent of the time.
But when McCulloch surveyed doctors in the managed care organization where
he works, he found that the doctors underestimated the number of patients
who had diabetes by 50 percent and overestimated the care they provided
to those patients by 100 percent.
McCulloch designed a system to give doctors easy access
to individualized patient information and treatment guidelines. The
system included reminders
that popped up on the doctor’s computer screen before patients
arrived for appointments, drawing attention to problematic test results
and prompting
the doctor to conduct routine screenings.
Despite some complaints that having structured guidelines
in place might lead to “cookbook” medicine, McCulloch argued that access to
the computerized information allows doctors to spend more time concentrating
on a patient’s specific needs.
“These patients get not 12 minutes with a doctor, but two or three
hours together in a setting which is not as intimidating as the traditional
one-to-one doctor-patient meeting,” McCulloch said. “But
benefits are greater than efficiency. They go beyond objective results,
as patients
start talking and support each other.”
In McCulloch’s observations, the increased attention
paid off in improved patient health and better tracking of potential
diabetes
complications.
And despite adding lab tests and other measures involving physician and
nurse practitioner time, costs actually decreased, he said. Inpatient
days declined by 26 percent while specialist visits dropped by 23 percent.
Pharmaceutical
costs rose by 16 percent or $11, but total costs went down by 11 percent.
The system saved $50 million while patient outcomes improved.
While clinical visits are important, the vast majority
of diabetes care takes place in the ordinary course of the patient’s life. And because
so much of that care occurs at patients’ homes, they must be more
closely involved in their own care, said Harold Goldberg, M.D., professor
of medicine at the University of Washington.
“The office-based system isn’t working,” Goldberg said, “so
how can we extend care beyond the office?”
His answer: Web sites and e-mails can be used to maintain
a flow of information between patient and physician. For example, patients
can
access the doctor’s
notes and see reminders for screening tests, or e-mail questions to a nurse
practitioner monitoring their case. They can even upload their self-administered
blood glucose readings twice a day, which are then analyzed by the doctor’s
computer. The computer can then suggest to the patient a change in diet
or medication, as needed.
This approach is undoubtedly the wave of the future,
Goldberg said. “In
2030, we’ll have 170 million people who will be chronically ill
and wired.”
Goldberg cited the example of one patient whose glucose
tests were normal each morning, so the patient decided he didn’t
need to test more than once a day. But a nurse practitioner eventually
persuaded him to
start monitoring his glucose after dinner as well, which revealed sharply
higher
levels.
The nurse prescribed medication and the patient’s glucose level
dropped from 8 percent to 6 percent. That decrease of two percentage points
cut the patient’s risk of blindness in half.
