ROANOKE TIMES 
                      Copyright (c) 1996, Roanoke Times

DATE: Saturday, August 10, 1996              TAG: 9608120037
SECTION: NATIONAL/INTERNATIONAL   PAGE: A-3  EDITION: METRO 
DATELINE: TORONTO
SOURCE: Associated Press 


COMPUTER `CHATS' IMPROVE HEALTH

EXPERIMENTS show that regular phone talks with a computer helped patients behave more healthily.

Patients who had regular telephone consultations with a computer lowered their cholesterol, exercised more and did better at taking their medicine, says a researcher who predicts the approach will become a major way of delivering health care.

The results came from three experiments in which patients called a talking computer that assessed their conditions and progress, and offered custom-made guidance through recorded messages.

Patients like the approach, and it's only a dollar or two per call, said Dr. Robert Friedman, chief of the Medical Information Systems Unit at the Boston University Medical Center.

It might be able to save money when used to monitor people with chronic diseases such as diabetes, by keeping them out of the hospital, he said.

Computer-controlled phone conversations ``will be a major means of delivering health care in the 21st century,'' he said.

Friedman spoke in an interview Friday after describing the experiments at the annual meeting of the American Psychological Association.

Participants talked to the computer by pressing numbers on a push-button phone.

One study involved 68 sedentary elderly people in the Boston area. To see if the computer chats could get them to walk for exercise, they were randomly assigned to have either a weekly call along with their usual care from their doctors, or just their usual care.

In all three experiments, each participant was expected to make the call. But if he missed the arranged time, the computer called instead.

In the conversations, the computer helped patients overcome obstacles to exercise, such as not knowing where to walk or how to make it fun. For example, it suggested getting a friend to come along.

Once the person began walking, the computer tracked his or her progress toward the goal of walking 60 minutes a week and offered encouragement.

After three months, those who had the computer conversations walked an average of 34 minutes a day, vs. 21 minutes for those in the comparison group.

In a different study, 65 adults with high cholesterol were randomly assigned to call the computer twice a week along with getting their usual medical care, or to get only their usual care.

The computer quizzed them on what they ate and how much. Then they got tailored guidance on improving their diets.

At the end of three months, those who had chatted with the computer showed an average 17-point drop in cholesterol levels, compared with three points for the other participants.

The third experiment involved 267 elderly people who were taking medication for high blood pressure. The goal was to help them take their pills as prescribed.

The computer asked how often each participant had missed a dose, explored the reasons and offered tips for overcoming the problem. In later conversations, it asked whether the tips worked.

Overall, compliance improved by 18 percent in the computer group and 12 percent in the others. The computer users also showed a bigger drop in blood pressure.


LENGTH: Medium:   67 lines




























































by CNB