Researchers call for EHR triaging in ICU

A central part of American hospitals, the intensive care unit (ICU) could benefit greatly from evolving health information technologies, offering clinical decision support, quality analysis and a new foundation for a critical care system changing with an aging patient population. Yet ICU triaging — prioritizing of the most severe conditions — has been underutilized and under-encouraged by the federal government, a team of medical researchers argue in a New England Journal of Medicine commentary.

In the early 1960s, many hospitals opened ICUs planning to care “for the sickest patients, using the newest technology,” write University of Michigan researchers, including Lena Chen, MD, an internist and clinical lecturer, and Timothy Hofer, MD, an internist and researcher at the Veterans Affairs health system in Ann Arbor. Today, U.S. ICU care costs $80 billion annually, and about one in five Americans die during a hospitalization with time spent in an ICU. “With an aging population and ever-growing demand for critical care, some observers worry that the number of staffed ICU beds will become increasingly inadequate,” the team writes.
 
In response to the “shortage of intensivists,” several new professional strategies and care models have been proposed and tested, including critical care certification and remote ICU telemonitoring. But as critical care becomes more important amid aging demographics, the researchers say the ICU potential of health IT has been under-studied. Specifically, Chen and Hofer wanted to find ways that information technologies can be used to triage decisions, to calculate patient priorities, the severity of conditions and risk of death or decline.
 
Chen, Hofer and two other co-authors say integrated care networks, such as the Veterans Affairs (VA) Healthcare System and Kaiser Permanente, already have electronic health records (EHRs) that could generate reliable 30-day risk of death estimates for every patient on admission. “Yet these calculations of risk” — combining real-time data on laboratory results, demographics, coexisting conditions and vital signs — “are not being used to inform decisions about admission to the ICU.”
 
The researchers examined a cohort of 101,912 patients admitted for reasons other than surgery to 121 VA acute care hospitals in the 2009 fiscal year, finding the ICU decisions varying widely among hospitals, “which suggests that there is at least some misallocation of resources.”
 
Although the results are still being finalized, the researchers say their preliminary findings add to evidence from the United Kingdom and the U.S. “challenging the notion that scarce and costly critical care is reserved for the sickest patients,” with one U.S. study finding as many as 40 percent of U.S. ICU patients receiving no ICU-level interventions and having a low risk of dying.
 
Among several hypotheses to explain the high use of ICU by relatively medically-stable patients, the researchers suggest that patients with cardiac diseases may have needs not captured by ICU severity codes, although they say that “the VA's ICU severity score is an excellent predictor of the 30-day risk of death.” It's also possible that patients with cardiac illness are overrepresented in the ICU because they often need automated biometric monitoring, were admitted to the ICU for protocol adherence or were awaiting transfer to another facility.
 
“Doctors may perceive patients with cardiac illness as being at higher risk for sudden, unexpected deterioration," the authors wrote. "In combination with concern about medicolegal repercussions and available staffing and clinical expertise on non-ICU wards, this perception may propel physicians toward the routine admission to the ICU of relatively well patients with cardiac illness.”
 
Arguing that “taking care of the sickest patients is the only role that has been explicitly endorsed for the ICU,” the researchers wrote. that overuse of the ICU — whether for “providers' convenience or peace of mind” or “as a temporizing measure for staffing problems” — is “unlikely to be an efficient use of valuable resources.”
 
Prioritizing care with triaging can “frame the subsequent course of care for all hospitalized patients,” the researchers wrote. “Reliable, individualized, EHR-based predictions of risk have the potential to improve our ability to triage — and hence care for — patients.” As the federal government and the greater healthcare community continue to craft incentives for EHRs, Chen and Hofer are encouraging more robust use of specific clinical analytics functions, saying that what currently “constitutes ‘meaningful’ clinical decision support is loosely defined.”
 
Although meaningful use has rightly given hospitals flexibility, the researchers note, “just as the explicit target of reaching the moon catalyzed space exploration, a few specific goals could accelerate HIT-related research in areas that are clinically pressing.”