PeraHealth Insights

Optimizing Medical Intensive Care Units



Many patients owe their lives to the care they receive while in an intensive care unit (ICU). However, ICU beds are expensive to operate, and having a limited number of them means that demand often outstrips supply, preventing some patients from being admitted to the ICU in a timely fashion, if at all. Determining the appropriateness of ICU admission is a multi-faceted problem involving objective data and clinical judgment.

Patients who have declined past a certain point may not reap the benefits of the care provided in the ICU. Similarly, patients who aren’t gravely ill don’t necessarily need the level of care available on the unit. Recognizing this, a Society of Critical Care Medicine (SCCM) task force has suggested guidelines to help prioritize patients for ICU admission based on how much they are likely to benefit according to whether the patient is:

  1. Critically ill – needing intensive treatment and monitoring that cannot be provided outside of ICUs
  2. Not critically ill but requiring close monitoring and potentially immediate intervention
  3. Critically ill but reduced likelihood of recovery because of underlying diseases or severity of acute illness
  4. Not appropriate for ICU – equivalent outcomes achievable with non-ICU care owing to low risk of clinical deterioration, presence of irreversible illness, or imminent death

As sensible and straightforward as these guidelines appear, a study from JAMA found that over 50% of patients admitted to the ICU at an academic public hospital were of questionable priority – they were either too well (priority 2) or too sick (priority 3) or could have received equally beneficial care outside the ICU (priority 4).

While ICU admission and discharge decisions are often complex, this study’s findings highlight an opportunity to improve decisions about who should receive ICU treatment and when.

What if the decision to admit a patient to the ICU took account of a broader range of patient data, drawing on hundreds or thousands of clinical measures from the patients’ medical record? What if the care team could visualize a patient’s condition to see when they start to decline hours or days before they code? Or better yet, what if they could visualize not only the trajectory of the patient’s current visit, but trends of prior visits as well to get a truly longitudinal view of the patient’s progression?

Tools such as the Rothman Index can enable this. The Rothman Index can assist clinicians in making more timely decisions, for example escalating patient care to the ICU before the patient is in crisis, and can provide better context for decisions by illustrating the patient’s physiologic stability over time.

The last twenty years have seen a revolution in the technology available to patients once they are inside the ICU. The next major gain in ICU care won’t come from how we treat the patient in the bed, but from the technology that supports getting the right patient in that bed at the right time.