Jan 012012

Each individual episode of critical illness produces a mushroom cloud of data, most of which dissipates without being recorded at all (think realtime infusion rates of vasopressors and continuous ECG monitoring). A few large databases capture outcomes data from multiple participating hospitals (like the National Inpatient Sample), and the new MIMIC-II integrated data system can collect and record high volumes of realtime patient data, which opens the door to more intensely granular analyses of events that occur during ICU stays. Despite these instances, from an epidemiologic standpoint (i.e. considering more than one ICU at a time), it's always been darn near impossible to see the forest for the trees in critical care.

One unexpected side benefit of telemedicine may be the aggregation, storage and organization of this kind of granular data, making new kinds of analyses available for researchers.

Craig Lilly et al report on a huge mound of data collected from more than 240,000 ICU admissions from 271 ICUs nationwide participating in the eICU Research Institute network. The most striking thing was that the patients weren't that sick:

  • Half were younger than 65.
  • The most common diagnosis was myocardial infarction.
  • About half the patients were admitted from the emergency department.
  • Most got antibiotics.
  • Only 27% received mechanical ventilation.
  • 24% received vasoactive medications (pressors, vasodilators or inotropes)
  • One in five were transfused blood products.
  • 40% did not have any intensive treatment documented.


About half the hospitals were non-teaching or "community" hospitals; only 17% were affiliated with a university.

Besides providing a potential new trove of data for researchers, this new eICU data feed provides an interesting snapshot of how ICU beds are likely used in the real world, outside tertiary centers. My impression is that in addition to the truly critically ill, a large proportion of the ICU admissions at community hospitals were "safety admits" or observation-only: watch them overnight or for a few days in the ICU to make sure their disease stabilizes. Nothing wrong with that, if you've got the beds.

Lilly CM et al. Benchmark Data From More Than 240,000 Adults That Reflect the Current Practice of Critical Care in the United States. CHEST 2011;140(5):1232-1242.

Liked this post? Get a weekly email update, and explore our library of clinical guidelines, practice updatesreview articles. and board review questions.

PulmCCM is an independent publication not affiliated with or endorsed by any other organization, society or journal referenced on the website. (Terms of Use | Privacy Policy)

Authors: contribute your work in a guest post.


Most people admitted to U.S. ICUs are not that sick (Review, CHEST)