The systems -- which check patients' blood pressure, weight and other health-related measurements daily and send information to their medical team -- are one strategy to catch problems early to prevent hospitalizations and help control health spending as the population ages.

The new findings don't mean so-called telemonitoring will never work, researchers said, but they suggest that the home check-ins might not be useful, or cost-effective, for everyone.

"These new technologies have tremendous promise to extend the ability of health care professionals to monitor patients outside the office," said Dr. Harlan Krumholz, a cardiologist from the Yale School of Medicine in New Haven who has studied telemonitoring but wasn't involved in the new research.

"Sometimes patients get into trouble long before they see us. The hope is that with knowledge that they're starting to get into trouble... we can intercept before they get to the point where they have to hospitalized," he told Reuters Health.

But all that promise, he said, "is failing to achieve the results that we've hoped."

MORE DEATHS AMONG THOSE MONITORED

Some studies of patients with heart failure in particular have suggested that telemonitoring can be helpful in avoiding hospitalizations -- but other trials haven't been successful.

The new study involved 205 elderly patients from Minnesota who were deemed to be at high risk for hospitalization because of their age and current medical conditions.

That included patients with heart failure, lung disease, diabetes and kidney disease.

Half of the patients were randomly chosen to get the at-home monitoring systems, produced by Intel. Those participants had five to ten minutes of measurements taken daily and also talked with nurses over the phone and could videoconference with them if they needed help.

The other half of study participants kept getting their usual treatment, with routine primary care and specialist visits.

Dr. Paul Takahashi from the Mayo Clinic in Rochester, Minnesota and his colleagues found that a year after getting their monitoring systems, patients in the telemonitoring group were still going to the ER and being hospitalized just as often as those without the systems. Sixty-four percent of telemonitoring patients visited the ER or were hospitalized during the study, compared to 57 percent of the usual-care comparison group.

And during that year, 15 percent in the telemonitoring group died, compared to just four percent of usual-care patients.

The researchers couldn't explain the increase in deaths with more monitoring, noting that there was no difference in quality of life between the patient groups or in their number of recent hospitalizations at the start of the study.

It's possible that the patients were different in other unmeasured ways, such as in the support they got from caregivers and their access to transportation, the researchers wrote Monday in the Archives of Internal Medicine.

Or, monitoring patients too much could actually be doing harm, Krumholz said -- for example if doctors overreacted to small changes in patients' health and treated them for problems that would have resolved on their own.

The monitoring systems "seem benign," he said. "But they generate information that people act on, and the question is, are we acting in ways that are beneficial?"

Takahashi agreed that there's a need for more research to tell doctors how to most effectively work with the constant data they get from telemonitors.

"I think you have to have a plan for when you get information. Once you have an effective plan, after you get it, I think you could very well intervene and make things a lot better," he told Reuters Health.

"Computers by themselves don't necessarily help the world -- it's what you do with the information."

Krumholz said that for now, use of telemonitoring should be restricted to studies and the systems shouldn't be put into general practice until researchers have figured out in what patients they may be useful.

The study was funded by the Mayo Foundation and the National Institutes of Health, and Intel provided the telemonitoring systems.

Researchers said it was hard to estimate how much the at-home devices would cost. In a previous study telemonitoring ran at between $1,000 and $2,000 per patient each year.

SOURCE: bit.ly/fO01ME Archives of Internal Medicine, online April 16, 2012.

• A Permanent Economic Emergency