Intensive Care Units (ICUs) Don't Save Lives Says Study!

Do intensive care units save lives?  That's a great question to ask in today's rapidly changing hospital payment environment.  We must continue to be diligent in asking ourselves if the care we are providing is the least resource intensive cheapest care possible without sacrificing outcomes.   If you get admitted to a hospital, most facilities will determine which floor or unit you go to based on your presenting diagnosis as well as how  intensive your nursing needs will be.  Are you a dialysis patient?  You may go to a dialysis floor.  Are you a heart failure patient?  You may go to the cardiac floor.  Are you old and weak?  You will probably go to the hallway in the basement holding area near the hospital's centralized utility system near the social worker's break room.  We don't want these low paying social admits taking up higher margin surgical beds.  Those doctors make a lot of money for hospitals.

We can't make them wait.   Intensive care units can even be subspecialized into medical units, surgical units and even subspecialty surgical units (like cardiac or transplant ICUs).  ICU nurses often train along side other ICU nurses for many months before they are free to practice on their own.  I met Mrs Happy when she was training as an ICU nurse.  Her final test?  She had to win at  ICU Bingo before she was free to see patients on her own.  And of course, the nurses climb all over each other to try and  get paid that whole extra dollar an hour to be the preceptor of the day.

How are decisions made for patients to be  transferred to the intensive care unit?  Hospitalists and other physicians often write the order to admit a patient to the ICU as a direct admission from an outside ER or other inpatient hospital stay or they may transfer a patient to the ICU if their clinical condition deteriorates during their hospital stay.  Why would a physician transfer a patient to the intensive care unit?  For many hospitals, the ICU is the only place that allows certain monitoring, procedures or interventions to be performed.  From a resource utilization standpoint, use of  ventilators, arterial lines, central venous pressure monitoring, titration of continuous medication drips and frequent nursing vital signs are all excellent reasons for patients to be admitted or transferred to the intensive care unit.  

Physicians will also admit or transfer patients to the intensive care unit who are experiencing a rapid deterioration in their clinical condition.  Transfers are made for conditions such as hypotension, acute respiratory failure, rapid gastrointestinal blood loss, severe sepsis or septic shock  and any other number of conditions that require intensive and continuous technological monitoring and bedside nursing care and that carry a high probability of necessary interventions that require ICU status.

Then there are hospital policies and procedures which state that certain objective data points require the patient to be ICU status.  Things like committee defined critical potassium levels of 6.0 (or 6.1 or 6.2 or 6.3) or blood sugars greater than 500 (or 600 or 700 or 800) or sodiums less than 115 (or 110 or 105 or 100) often lead to ICU admissions or transfers by default, independent of the lack of clinical  decompensation   and often without any allowable  input of the attending physician.  These are hospital defined parameters that committees have determined to be in the patient's best interest, even though no data exists that such policies are medically necessary.  

These are the three main ways patients end up in the intensive care unit.  Intensive care units are expensive for patients, insurance companies and hospitals. So I ask the question again, "Do ICUs save lives?"   Do we overutilize intensive care units?  Have they become really expensive nursing homes and end of life hospice houses at many institutions?  Those are questions  Paul Levy (ex Boston hospital CEO) eludes to in reference to his post  last month in his blog Not Running a Hospital.  This discussion was in reference to an Archives of Internal Medicine  titled:

Intensive Care Unit Bed Availability and Outcomes for Hospitalized Patients With Sudden Clinical Deterioration.


I read this intriguing study.  From 2007 to the end of 2009, consecutive hospitalized patients (excluding cardiac surgery and coronary care units) in Calgary, Alberta, Canada were selected based on their sudden clinical deterioration triggering medical emergency team activation . The intensive care units in this study were closed units staffed by intensivists.  

A total of 3494 patients were selected based on the above criteria.  The attending physician requested the ICU bed without routine notification of whether ICU beds were available or not (no beds, one bed, two beds, more than two ICU beds).  The primary outcome was ICU admission within 2 hours of emergency team activation. Four secondary outcomes were also evaluated:
  1. Change in patient goals of care(resuscitative, medical, or comfort) within 24 hours of emergency activation
  2. Hospital mortality
  3. Health care resource use
  4. ICU admission during the remainder of the hospitalization
Results of the study concluded that patients were less likely to be transferred to an ICU within 2 hours of a medical emergency team activation if ICU bed availability was reduced. In fact, when no ICU beds were available, patients were 33% less likely to be to be admitted to the ICU. And when no ICU beds were available, patients were nearly 90% more likely to change the goal of care from resuscitative to medical or comfort care.    Interestingly, the odds ratios for hospital mortality did not change across all four groups based on ICU bed availability (approximately 33% mortality for all groups)

What does this data suggest?   How available our ICU beds are can determine processes of care in the hospital but it does not appear to change mortality.  In other words, if we build it, they will come. And if we set up archaic hospital policies to transfer patients to the ICU, we can then capture these charges  because that's the policy we have in place. But it won't make them better or save their lives

If we have big ICUs with lots of  beds, we are more likely to fill those beds with patients that don't necessarily require ICU monitoring and who's care in the ICU will not change their outcomes.  And given that most physicians have no formal training to  determine which patient is appropriate for the ICU and which isn't, and making the decision to transfer a patient to the ICU is a highly individualized process that is decided on a case by case issue between doctors and nurses as well as outdated hospital policies and procedures.    This is not to say that ICUs are not an important component that can provide necessary care.  Just that we probably use them a lot more often than we need to because we lack a formal process to define their role.  Now the question becomes, how do we start using intensive care units in a more cost effect way?

This sounds like a perfect opportunity to create another evidence based ICU checklist  process for hospitals to  define who is appropriate for the ICU and who isn't. Perhaps the Society of Hospital Medicine could tackle this process, if they haven't already.    This is also a good opportunity for hospitals to reevaluate their old school voodoo medicine ICU criteria that hasn't been evaluated and probably never  validated and to rid hospital policies and procedures filled with honky tonk  rules that do nothing for patient care but increase cost and inconvenience for patients and create an unnecessary paper trail the size of Texas for nurses stuck in the middle.  

Far too often the ICU is a place of convenience for doctors, nurses and patients who falsely believe that ICU care is better than non ICU care.  This data may be telling us that is not the case.   As I suspect to be true  in so  much of what we do, less is more.

Alternatively, it might suggest ICUs don't save lives because we are making up our own data, as this  original Happy Hospitalist ICU ecard so eloquently explains:

"I'm proud to announce my ICU had 47 cases of ventilator associated pneumonia last year, but we called bull on all of them and reported zero instead."

I'm proud to announce my ICU had 47 cases of ventilator associated pneumonia last year, but we called bull on all of them and reported zero instead doctor ecard humor photo.


Some of this post is for entertainment purposes only and likely contains humor only understood by those in a healthcare profession. Read at your own risk.

Print Friendly and PDF