It makes intuitive sense: More doctors' eyes and ears on a patient should make for better care. For this reason, the potential of telecommunications to improve quality in intensive care units has been hailed since at least the 1970s, according to Jeremy M. Kahn, MD, associate professor of critical care and health policy at the University of Pittsburgh.
However, Dr. Kahn told attendees at the American Thoracic Society's annual meeting in New Orleans in May, experimentation with telemedicine in ICUs has not quite lived up to expectations. “Results are pretty mixed,” said Dr. Kahn. “The early results were pretty positive, but some of the later studies showed no change [in outcomes].”
Depending on how it is applied, the technology could even have negative effects, wasting scarce resources or worsening care, he warned. During his lecture, Dr. Kahn described some potential pitfalls and offered ideas on how telemedicine actually could be used to improve care.
The rich get richer
One problem with many existing telemedicine programs is their method of implementation. “You can think of ICU telemedicine almost like a pulmonary artery catheter. There's nothing about that catheter sitting in the pulmonary artery that's going to improve outcomes. It's how we use it,” Dr. Kahn said.
Most trials of the technology—six of which Dr. Kahn reviewed during his talk—have used it for continuous monitoring during the night. An intensivist at another hospital watches ICU patients (sometimes at multiple hospitals) through a camera and monitors.
So far, the trials haven't shown clear benefits to patients. “Some preliminary evidence indicates improvement in the processes of care, but there's not yet strong evidence linking the processes to outcomes,” said Dr. Kahn. The improved processes include use of thrombolytics for acute stroke, shorter wait time for evaluation by a neuro-intensivist, implementation of a daily sedation holiday and consultation by a pharmacist.
But those processes don't usually happen in the middle of the night, which is when the ICUs are typically using telemedicine. Most of the studied hospitals already had on-site critical care specialists during the day, so their telemedicine programs were only adding nighttime coverage.
“Is this model, which is how telemedicine has primarily been applied, really going to impact outcomes?” asked Dr. Kahn. “How much more benefit are you going to get by adding intensivist coverage at night?”
Also, a survey of hospital administrators, published by Health Affairs last August, found that ICUs that used telemedicine were almost universally compliant with Leapfrog standards (such as computerized order entry and intensivist staffing) already, Dr. Kahn reported. “The rich get richer,” he said.
This allocation of telemedicine is a problem not just because it's unequal, explained Dr. Kahn. Because the country's supply of intensivists is limited, daytime care could potentially be affected by the focus on night coverage. “Think about the workforce as a zero-sum game. If the intensivist is in the box at night, that intensivist can't be in a rural community providing care,” he said.
However, if telemedicine were applied differently—emphasizing consultations during the day in hospitals without intensivists, instead of as night vigilance—it could potentially improve outcomes and alleviate physician shortages, Dr. Kahn suggested.
“The study that needs to be done is a trial of small community hospitals that have no access to intensivists,” he said. Theoretically, at least, the consulting intensivist could increase the chances that processes of care—like sedation holidays or head of bed elevation—are implemented in hospitals where these actions aren't already standard.
That model of telemedicine hasn't been studied much, perhaps because, as the Health Affairs study found, administrators at these non-Leapfrog-compliant hospitals have reservations about teleICUs. They worry that the benefits of the programs don't outweigh the costs and that they might shift resources around rather than expanding access or improving care, among other issues.
“Buy-in is a problem,” said Dr. Kahn. “We need to research the mechanisms of buy-in.”
Thus far, the research tends to support the administrators' reluctance. A study of ICU telemedicine published in December 2009 in the Journal of the American Medical Association found that, overall, telemedicine had no effect on mortality or length of stay in ICU patients.
When the patients in the study were divided up by severity of illness, the findings got worse—the less severely ill patients actually had higher mortality with telemedicine. “It was really helping people who are sick. It was hurting people who are not [very] sick,” said Dr. Kahn.
He offered a possible explanation. “It was potentially a vigilance issue. If a nurse knows that the ICU telemedicine camera is in the room, perhaps he or she is not going to spend the same time with that patient.”
Like most studies of telemedicine completed to date, this one had a methodological limitation in that it compared outcomes before and after implementation of telemedicine, so any other changes that occurred in the interim could have confounded the results.
Research methodology is only one of many issues that need to be tackled before telemedicine can be ideally implemented, according to Dr. Kahn. A reimbursement system for physician payment is needed, as well as resolution of the legal and regulatory issues surrounding practice across state lines. It would also help if the price of telemedicine systems came down, he noted.
Given the current limitations, he's not ready to give an absolute yes or no on whether telemedicine is a viable solution to intensive care workforce shortages. “If we harness the distance capability and we truly provide care where it's needed most—in rural areas—and we emphasize consultation rather than vigilance, then the answer is maybe,” said Dr. Kahn.