Virtually all the hospitals that use machines to monitor fetal hearteats are doing it wrong, according to a new study by Dr. John Patrick of the University of Western Ontario.

Furthermore, Patrick says, the fetal monitoring machines were in widespread use in hospitals eight years before the first study of their value was even begun. "No one should have installed these machines until they knew whether they were of any value -- because there are risks involved. And also they should know what the measurements on the machines mean before using them -- and doctors didn't than and still don't now."

There are between 30 and 50 percent false positive results in the use of fetal monitors, Patrick said, and that means that many babies are delivered under emergency procedures, such as Cesarean operations, when no such action is needed.

Though most of these babies and mothers are not hurt as a result of the mistakes, a small percentage of such emergency procedures results in injury or death.

The fetal monitoring machines are used before and during labor to monitor heartbeat and fetal movement; if either (or both) is abnormal, emergency measures may be taken to remove the fetus from the womb or change its position within the womb.

In current practice, if the fetus's heartbeat is lower than 120 beats per minute or higher than 160, it is considered abnormal. Also, if in 20 minutes of monitoring there are fewer than three movements by the fetus, it is considered abnormal.

"This test is widely accepted in America and Western Europe," Patrick said. Over the past five years, Patrick and his colleagues have made measurements of fetal and maternal heartbeat and activity rates, and found the currently accepted rules wrong.

A normal fetal heart rate, according to Patrick's studies, has a daily low and a daily peak. Both the peak and the low can easily be outside what is considered normal. Normal fetuses, especially during the early morning hours, can have a heartbeat 20 beats lower than what is now called "normal."

Patrick has also found that fetuses also have regular rest and activity cycles that are often longer than 20 minutes. A healthy fetus may have rest and activity cycles that average about 50 minutes.

He suggests that fetal monitoring should be carried on for 80 to 120 minutes before deciding there is an abnormal lack of movement.

"But I think my ideas should be further tested before being put into practice, otherwise I would be acting as badly as the ones who put these machines in and used them without tests," Parick said.

Patrick has also begun research on a new method of monitoring fetal heartbeat that would be "noninvasive" -- that is, that would not be attached directly to the fetus's skull as it often is now.

There are now machines that mmonitor heartbeat without attaching an electrode to the fetus, but Patrick said that they are notoriously inaccurate because the fetus moves and garbles the readings.

His device would be a belt with electronic sensors that is placed around the woman's abdomen. An electronic sensor would read the breathing and heartbeat pattern, and as the signal became weaker during movement, the device would electronically follow its position with other sensors spaced around the belt.

"I would hope that this device and making use of the knowledge of what is normal activity and heart rate for a fetus might allow us to eventually eliminate virtually all of the false positives. We should be able to catch the sick basies much more effectively," Patrick said.