Intensive Care Unit

How Many ICU Beds Does A Hospital Need?

When a new hospital is built or an existing hospital plans to expand, a key question is “How many ICU beds will we need?“. There is not a one-size-fits-all answer to that question but there are some general principles that guide the number of ICU beds that a hospital requires.

Data from the American Hospital Association’s 2020 publication on U.S. hospital resources reveals that there are 6,146 hospitals in the U.S. Excluding the psychiatric hospitals and federal hospitals, there are 5,198 community hospitals in the United States, of which, 51.4% have intensive care units. Overall, there are 792,417 community hospital beds in the U.S., of which 13.5% are ICU beds. The overall make-up of U.S. community hospital beds is:

  • 792,417 total hospital beds
  • 55,663 medical/surgical ICU beds
  • 15,160 cardiac care unit beds
  • 7,419 other ICU beds
  • 22,721 neonatal ICU beds
  • 5,115 pediatric ICU beds
  • 25,157 step down unit beds

Location, Location, Location

Intensive care units are not uniformly distributed in the United States. Many smaller hospitals lack ICUs and consequently, more than 50% of U.S. counties do not have any ICU beds. Most ICU beds are found in metropolitan areas (defined as > 50,000 population) or micro metropolitan areas (defined as 10,000 – 49,999 population:

  • 94% of ICU beds are in metropolitan areas
  • 5% of ICU beds are in micro metropolitan areas
  • 1% of ICU beds are in rural areas

There are a number of reasons for the paucity of ICU beds in rural areas but perhaps the most important reason is that an ICU is more than just a bed and a ventilator, an ICU requires critical care-trained nurses, advanced pharmacy support, 24-hour respiratory therapy, and physicians with critical care skills. Smaller hospitals in rural areas generally cannot support all of these specialized personnel to provide care for a relatively small number of ICU beds.

Although the overall average percentage of hospital beds in the U.S. that are ICU beds is 13.5%, because many rural hospitals lack ICUs, the percentage of ICU beds in metropolitan hospitals is necessarily higher than 13.5%. This is particularly true for academic medical centers and pediatric hospitals that function as tertiary care facilities with the result that these hospitals admit more complex patients who more often require ICU services. So, for example, in Columbus, Ohio, between the 3 major hospital systems plus the children’s hospital, there are 3,873 hospital beds. Of these, 572 (15%) are ICU beds.

Another way of analyzing ICU bed use is by expressing the number of ICU beds per 10,000 population. This can be misleading, however, because many rural areas will send most of their ICU-level patients to nearby metropolitan areas with the result that the larger metropolitan areas will have more ICU beds per capita. Nevertheless, in an analysis by the Washington Post, the major metropolitan areas in Ohio varied significantly in ICU beds per 10,000 population:

  • 6.3 Toledo
  • 5.3 Cleveland
  • 5.0 Cincinnati
  • 5.0 Dayton
  • 4.7 Akron
  • 4.4 Canton
  • 3.6 Columbus

What About Utilization?

The Society of Critical Care Medicine recently analyzed ICU occupancy. Overall in the United States, the ICU occupancy rate is 66.6% for adult ICUs, 61.6% for pediatric ICUs, and 67.7% for neonatal ICUs. One of reasons that these percentages seem low is that there can be wide swings in occupancy and a hospital needs to have sufficient resources to accommodate high-census times. Furthermore, ICU occupancy is generally based on the midnight census in a hospital, that is, the number of patients in a bed at midnight. Because patients usually do not get transferred out of ICUs until early afternoon (after the physicians make morning rounds) but patients get transferred into ICUs continuously throughout the day, the 66.6% occupancy rate at midnight for adult ICUs underestimates the peak occupancy for ICUs at noon which is considerably higher.

When a hospital’s ICU occupancy rate is low, the ICU tends to harbor less acute patients. Conversely, when a hospital’s ICU occupancy rate is high, higher acuity patients often get admitted to non-ICU locations such as step-down units. Consequently, a patient with a COPD exacerbation requiring non-invasive ventilation with BiPAP but not requiring intubation with mechanical ventilation would be admitted to an ICU bed when there is adequate ICU capacity but might be admitted to a step-down bed when there is insufficient ICU capacity.

So, how many ICU beds does a hospital need?

If we start with the U.S. average, then a hospital needs 13.5% of its beds to be ICU beds. Hospitals in larger cities need a higher percentage whereas hospitals in small towns need a lower percentage. Within larger cities there will also be variation: tertiary care hospitals and children’s hospitals will require a higher percentage than other hospitals in metropolitan areas.

In deciding whether to expand ICUs, a hospital should also look at its ICU occupancy rate. If the average occupancy rate is < 66% then the hospital likely does not need additional ICU beds. However, if the occupancy rate is > 66%, then ICU bed expansion may be warranted.

Lastly, the hospital should examine the acuity of patients in the ICU when determining whether ICU beds should be increased. A 2013 study of ICU occupancy and ventilator use in the U.S. found that the mean percentage of ICU patients on a ventilator at any given time was 40%. If a hospital’s ICU ventilated patient percentage averages less than this, then it may not need additional ICU beds. However, if the percentage of ICU patients on a ventilator is > 40%, then more ICU beds may be needed.

A hospital needs to have the correct number of intensive care unit beds to support its operating rooms and general nursing units. There also needs to be sufficient ICU beds regionally to support the community’s need to care for the sickest patients. ICUs are expensive and the specialized staff it takes to care for patients in ICUs are even more expensive. However, the DRG reimbursement for these patients is high and so ICUs can be financially lucrative for a hospital. When hospitals plan bed expansion, it must be done with the right balance of ICU to non-ICU beds in mind.

October 10, 2020

Intensive Care Unit

The Pandemic Of Dogma Within The Pandemic Of COVID-19

Our generation of physicians prides itself on the practice of evidence-based medicine. Ideally, this means making medical decisions based on peer-reviewed clinical studies and randomized, controlled clinical trials. It means getting away from the “this-is-the-way-you-do-it-because-this-is-the-way-we’ve-always-done-it” approach to medicine to ensure that patients get the best known treatment for any given medical condition.

But what happens when you face a disease and there are no peer-reviewed publications and randomized controlled clinical trials? In that situation, physicians’ definition of what constitutes evidence can vary considerably. Thus enters dogma and nowhere in recent memory has there been a greater pandemic of dogma than in our intensive care units managing patients with COVID-19. As critical care physicians, we hold our convictions about how to best treat patients with COVID-19 respiratory failure more tightly than we hold religious convictions or political convictions. And given that many of these convictions are diametrically opposed, we can’t all be right. Forty years ago, in my first week of medical school, one of the professors told me that 50% of everything I was about to learn was going to ultimately be proven to be wrong. I think those words could be just as easily applied to our approach to managing patients with COVID-19 in the ICU today.

To Ventilate or Not To Ventilate?

That  is the question… or is it? For decades, we have defined patients as having ARDS (acute respiratory distress syndrome) if they have acute onset of diffuse pulmonary infiltrates with severe hypoxemia in the absence of heart failure and in the presence of something known to cause non-cardiogenic pulmonary edema, such as infection. And ever since 1967 when surgeon David Ashbaugh and pulmonologist Tom Petty first described the ARDS in the medical literature, it has been well-accepted that mechanical ventilation with PEEP (positive end-expiratory pressure) is the first line treatment.

But in the era of COVID-19, we read on-line news articles about a hospital in New York that reported 88% of their patients placed on mechanical ventilators died. And then all of a sudden, some critical care physicians are having second thoughts about intubating COVID-19 patients in respiratory failure and instead letting them be hypoxic. On the other hand, we read a blog post from a physician in Europe that he observed hospitalized COVID-19 patients develop sudden severe hypoxemia and go from needing 4 L oxygen by nasal cannula to having respiratory arrest despite 100% oxygen by face mask in just 30 minutes. And all of a sudden, some critical care physicians are intubating every COVID-19 patients who needs 4 L oxygen by nasal cannula.

Maybe COVID-19 really is different from ARDS from any other infection. But until someone proves that, I know that mechanical ventilation can bridge patients through life-threatening ARDS until time heals the lungs, that PEEP helps, and that low tidal volume ventilation is better than high tidal volume ventilation. We should not throw out everything that we’ve learned about the management of ARDS over the past 53 years because of a blog post.

Steroids Yes or Steroids No?

Over the past 35 years, the steroid pendulum has swung back and forth several times with respect to treating ARDS. First, studies showed steroids were beneficial, then studies showed they were not beneficial, and now studies again suggest they might be beneficial again. Similarly, you can find studies that show steroids improve the mortality rate of other coronaviruses and influenza; you can also find that steroids have no effect on the mortality rate of viral pneumonia. Some critical care physicians believe that steroids are the cure to the “cytokine storm” attendant to COVID-19 respiratory failure. Other critical care physicians believe that steroids paralyze the body’s immune defenses against COVID-19 resulting in increased viral replication. Our resident and fellow trainees are often caught in the middle, hearing that “You’re going to kill your patients if you don’t give then steroids” from one critical care attending physician on Monday and then hearing “You’re going to kill your patients if you give them steroids” from another critical care attending physician on Tuesday.


A non-randomized, non-placebo-controlled study from France suggested that 20 COVID-19 patients who got anti-malaria drug hydroxycholorquine had lower levels of detectable virus than patients previously published in the literature. This made immediate news in the lay press and the U.S. President called the drug a “game changer”. Within 3 days, pharmacies all across the country were sold out of hydroxychlorquine and I had patients calling in and asking me to prescribe it for them to prevent getting COVID-19 infection. Physicians throughout the world began prescribing it for any of their patients sick enough to be admitted to the intensive care unit. But then other studies showed that patients who received hydroxychloroquine actually did worse than those did not receive it because of potential fatal heart rhythm disturbances brought on by hydroxychloroquine. Once again, you’ll find critical care physicians who think it is the standard of care and others who think that it is nonsense.


Patients with COIVD-19 have high levels of the cytokine, IL-6. This occurs during the “cytokine storm” that these patients can get when their macrophages and monocytes produce enormous quantities of pro-inflammatory cytokines. This is also called the “macrophage activation syndrome”. Tocilizumab is an inhibitor of IL-6 and so some physicians believe that by inhibiting IL-6, the cytokine storm can be attenuated. It is one of those “makes sense, no data” treatments that might make patients better, might not do anything at all, or might actually make them worse. But in the absence of randomized, placebo-controlled clinical trials, you can find critical care physicians who are staunch proponents and others who are staunch opponents.

And Everything Else?

Across the United States, there are some critical care physicians who believe that because D-dimer levels are high, that anticoagulation helps by preventing clotting; other critical care physicians thing that empiric anticoagulation just makes patients bleed more. Some physicians believe that inhaled vasodilators such as nitric oxide or epoprostenol improve oxygenation in COVID-19 patients by redirecting blood flow to less affected parts of the lungs; other physicians believe that these drugs can cause patients to become hypotensive and develop cardiac arrest. Other treatments that might or might not work include transfusion of plasma from patients who recover from COVID-19 infection, the anti-viral drug lopinavir/ritonavir, another anti-viral drug remdesivir, and the complement inhibitor eculizumab.

As humans, for thousands of years we have sought ways to control nature. And we base a lot of our attempts at control on anecdotal experience that leads to superstition. For example, a child falls into a volcano and the next day it rains so the village starts throwing lots of children into the volcano the next year when there is a drought. As physicians, we are no different. We see or hear about a patient who got one treatment or another and got better and then that one patient or small group of patients becomes the evidence that we base our practice on when there is a vacuum of randomized, placebo-controlled clinical trials. COVID-19 has overtaken the world suddenly, too fast for science to give us direction about how to best treat patients and so we fall back on medical superstition. Some of those superstitions will ultimately be proven to be right and others will ultimately be proven to be wrong.

So, all of a sudden, what constitutes evidence in evidence-based medicine today is a lot different than what constituted evidence last year.

April 25, 2020

Emergency Department Inpatient Practice Intensive Care Unit

With COVID-19, Hope For The Best But Prepare For The Worst

In August 2004, my family was vacationing on the North Carolina Outer Banks. I had been following Tropical Storm Alex as it came north from the Caribbean toward the island that we were staying on. On August 3rd, it was looking like the storm was going to head out over the Atlantic the next afternoon and miss Cape Hatteras. Not wanting to take any chances, I decided to get up early the next morning, pack up the kids, and head inland for the day, just to be sure. When I woke up at 5 AM, the first thing I heard on TV was that overnight, the storm had picked up wind speed, was moving across the ocean faster than expected, and had turned inland – directly toward our rental house in the town in Salvo. The second thing that I heard was that there was that storms overnight had caused sand and water to block the only road on the island leading to the bridge to mainland. The news announcer said to all of the people now stuck on Hatteras Island “Hope for the best but prepare for the worst.”

Having 4 children, my wife and I were used to buying in bulk and since this was at the beginning of our planned 2-week vacation, we were already pretty well stocked with food and supplies. We filled up all of the bathtubs with water for bathing and filled up as many bottles as we could find with drinking water.

By the time the storm hit us, Alex was now a level 2 hurricane. The eye wall passed over our rental house and as the wind changed direction with the passage of the eye, we moved all of the kids from a bedroom in one corner of the house to bedrooms in other corners. As the power went out, the wind sounded like a freight train and I watched as siding and parts of roofs were torn off of houses around us. A 2×4 board flew through the air like a missile across the street. Picnic tables, bicycles, and and lawn furniture were flung a hundred yards like toys. The roads all turned into rivers. Meanwhile, we played games with the kids and fed them Cheerios to keep them distracted.

It seemed like the end of the world and I wanted to be almost anywhere other than where we were.

But by afternoon, the wind died down, the clouds cleared, and the sun came out. All of a sudden, it was just another beautiful day on the Outer Banks. Over the next 3 days, the power returned, the flood waters subsided, and the sand was cleared from the roads. The bridge re-opened and the people staying in Salvo came out and cheered when one of the first vehicles that crossed the bridge to the island was a Budweiser truck.

COVID-19 is a lot like Hurricane Alex. The patient surge is coming and we can’t just wish it away. Just as the news announcer said on TV in the morning of August 4, 2004, we should hope for the best but prepare for the worst. But also like Hurricane Alex, the COVID-19 surge is going to pass; the clouds and pandemic storm is going to eventually subside; and life will be back to normal once more.

April 1, 2020

Inpatient Practice Intensive Care Unit

Reducing Hospital Employee Exposures To COVID-19 Patients

Having patients with COVID-19 in the hospital can be disturbing to the doctors, nurses, and respiratory therapists who take care of them. The good news is that isolation procedures work and proper use of personal protective equipment can dramatically reduce the chance of getting healthcare workers infected. Even though that risk is low, there are certain simple steps you can take that will reduce the risk even further. By taking these steps, you not only reduce healthcare worker exposures but you can also conserve personal protective equipment (masks, gowns, gloves). Here are a few:

  1. Use the right personal protective equipment (PPE) and be sure that it is used correctly.
  2. Minimize blood draws. If you don’t need daily labs, don’t send the nurses in to draw them. When you do get labs, try to cluster all of the lab tests that you need in a single phlebotomy.
  3. To anticoagulant a patient, use oral apixaban, oral rivaroxaban, or subcutaneous enoxaparin instead of a heparin drip. The problem with heparin drips is that you have to do frequent PTT blood tests. Other anticoagulants do not require testing.
  4. Use a sliding scale of subcutaneous insulin rather than an insulin drip. Insulin drips require the nurse to check the patient’s blood glucose every 1-2 hours whereas the SQ insulin sliding scale may only need to be done every 6 hours.
  5. Synchronize medications. Ordering a Q6 hour medication plus a Q8 hour medication means that a nurse has to go into a patient room 7 times a day. If that Q6 hour medication can be stretched out to be given Q8 hours, then a nurse only has to enter a patient’s room 3 times a day. Even better, use medications that only have to be given once a day whenever possible. This is particularly true of empiric antibiotics where there may be multiple equally appropriate antibiotic choices – some that have to be given 3 or 4 times a day and some that only have to be administered once a day.
  6. Use meter dose inhalers instead of nebulizer treatments. Nebulizers can result in aerolsolization of viral particles, at least in theory. Meter dose inhalers for bronchodilator treatments reduce the amount of time that a respiratory therapist has to be in a room to deliver a bronchodilator treatment.
  7. Have patients self-administer meter dose inhalers (or nebulizer treatments). The respiratory therapist can often observe the patient from a door window or a video monitor to ensure that the patient uses proper technique.
  8. Minimize the rounding team. If bedside rounds normally consist of the attending physician, a nurse, a resident, and a physician assistant, then reduce that to just the attending physician and just once a day.
  9. Don’t use physical and occupational therapy if you don’t need it. Frequently, admission order sets will include PT and OT for nearly every admission. Only order it if you really need it.
  10. Don’t order tests that you don’t need. “Routine” daily chest x-rays are usually unnecessary.
  11. Don’t order tests that can be done later. If a chest x-ray shows a suspicious pulmonary nodule and a chest CT is recommended for confirmation, that CT can wait a few weeks.
  12. Empiric treatment is OK. If a patient has epigastric pain, rather than ordering an endoscopy right away, give the patient some empiric omeprazole to minimize procedures.
  13. Utilize inpatient telemedicine for consults. There are two ways to do this, by a regular telemedicine visit or by an eVisit.
    1. CPT 99451 is for an eVisit and reimburses at 1.04 RVUs. There has to be an order for the consult and the consultant has to put a note in the medical record. The consultant must document his/her time and it must be > 5 minutes. This is a way to get reimbursed for the so-called “curbside consult”. An example would be “What follow up should occur for the incidental 5 mm pulmonary nodule that was seen on my patient’s CT scan?”
    2. CPT G0425 (30 minutes ), G0426 (50 minutes), and G0427 (70 minutes) are for initial inpatient telemedicine consults. For follow up inpatient consult visits, use CPT G0406 (15 minutes ), Go407 (25 minutes), and G0408 (35 minutes). These codes are based on the amount of time communicating with the patient
  14. Can you run your pumps outside of the patient’s door? continuous infusion pumps are forever alarming or needing infusion rates to be frequently adjusted. If the infusion pumps can be placed outside of a door with the tubing running under the door then the pumps can be adjusted without the nurse having to enter the room.
  15. Eliminate visitors. Visitors can bring COVID with them and many visitors have often had close contact with COVID patients before they were admitted, making them especially high risk. By eliminating visitors, there are fewer members of the public in patient care areas who can infect hospital staff. Furthermore, there are fewer times that the patient’s door is opened and no additional personal protective equipment consumed by the visitors.
  16. Be sure that the healthcare personnel are getting enough rest. When a nurse, RT, or physician works too long of a shift or too many shifts, fatigue can set in and with fatigue brings mistakes. Mistakes with isolation procedures can create infection risks.

March 31, 2020

Epidemiology Intensive Care Unit

Re-Using N-95 Masks In The Time Of COVID-19

The geniuses at Battelle have done it again. This time, they have invented a process for sterilizing and re-using N-95 masks using vaporized hydrogen peroxide. Battelle Memorial Institute is a non-profit scientific research and development institute here in Columbus, Ohio that is located about 3 blocks from the OSU hospital. Researchers at Battelle were the ones who invented the photocopier (and then launched Xerox Corporation), the cruise control for automobiles, the first nuclear fuel for nuclear-powered submarines, and the reusable insulin pen for injecting insulin for diabetics. In full disclosure, about 25 years ago, I had a grant from Battelle to assist with development of inhaled chemotherapy for lung cancer and that led to my receipt of the endowed Battelle Professorship in Inhalational Therapeutics that I held until assuming my position as the medical director of our hospital.

N-95 refers to a mask that can filter 95% of airborne particles. In medicine, we use N-95 masks when we care for patients with infectious diseases that are transmitted by airborne routes, such as tuberculosis, disseminated varicella, and measles virus. The virus that causes COVID-19 is the SARS-CoV-2 virus and this is believed to be transmitted by droplet spread rather than by airborne spread. Normally, viruses spread by droplets do not require the use of N-95 masks; a simple surgical mask with a plastic face shield will suffice. However, certain medical procedures, such as endotracheal intubation, can result in aerosolization of droplets containing viral particles and that is when the N-95 masks are needed.

For an N-95 mask to work properly, a healthcare worker must be fit tested to determine which specific type of N-95 mask fits tightly against the face. If a type of mask does not pass the fit test, then it will not filter out 95% of airborne particles and is no better than a regular surgical mask. Everyone’s face is shaped a little differently so different people will need different N-95 mask types. All healthcare workers who use these masks are required to get fit tested once a year to ensure that the mask that they are wearing actually does what it is supposed to do. Recently, OSHA declared that men who wear beards should not be fit tested because beards can interfere with a tight fit of the masks. For many years, I always passed my fit test with a specific type of N-95 mask despite my beard but because of OSHA’s rules, I was not able to be fit tested last year. Two weeks ago, our hospital required all men who could be involved in the care of a COVID-19 patient to shave their beards (so that they can be fit tested for N-95 masks) and thus, I shaved for the first time in 37 years!.

N-95 masks have come to the forefront of public consciousness recently because the COVID-19 outbreak is causing many hospitals to run low on N-95 masks. A misconception has arisen that N-95 masks are safer than regular surgical masks plus a face shield. For day-to-day care of patients with COVID-19, this really is not true because unless you are performing a procedure such as endotracheal intubation, an N-95 mask is unnecessary. Overuse of N-95 masks in situations when they are not necessary now will result in inadequate supples of these masks in situations when they are necessary in the future. In addition, the over emphasis on N-95 masks could lead the public to overlook the single most important way to prevent the spread of viruses spread by droplets, namely washing one’s hands after they touch various surfaces that those droplets land on (such as door handles and elevator buttons).

With supplies dwindling, Battelle invented a process for sterilizing N-95 masks so that they can be reused up to 20 times. They built the equipment to process 160,000 masks per day and this would greatly improve the nation’s N-95 mask inventory. However, medical equipment is overseen by the Food and Drug Administration. The FDA would only grant Battelle’s mask sterilization equipment limited approval, meaning that they are only permitted to sterilize 10,000 masks per day and only here in Central Ohio. That’s good news for our hospital because now we can count on a steady supply of masks in the upcoming weeks of the COVID-19 surge. But it is bad news for every other hospital in the United States.

Desperate times call for desperate measures. This may be a time for the FDA to take the desperate measure of cutting through bureaucracy.

March 29, 2020


Intensive Care Unit

Preparing For ICU Surge Capacity In The Time Of COVID

The COVID-19 pandemic has created enormous demand on the world’s intensive care units. As of today, Central Ohio is still in the very early stages of the outbreak whereas countries such as Italy, Spain, China, and Iran have had large numbers of patients. About 10% of those infected eventually need admission to an intensive care unit so it is the ICUs that get the highest volume of hospitalized patients. When the infection peaks in your community, the hospital has to be prepared for the possibility that the demand for ICU beds could exceed the supply of ICU beds. Here are some of the things to consider in preparation for the peak demand:

Alternate Sites for Intensive Care

The physical characteristics of a hospital room is the first consideration. Not all rooms are as equally adapted to ICU rooms as others. The first consideration is whether there is monitoring capability – those rooms that already have monitors can more easily become ICUs. The second consideration is whether the room has a medical gas supply built into the wall. Many rooms will have oxygen supplies but most ventilators need both oxygen and compressed air supplies in order to blend to a specific oxygen concentration that is delivered to the patient. Wall suction is also necessary. Because COVID-19 patients require droplet isolation, the room should have a door (as opposed to just a curtain).

As you plan for alternative areas for ICU surge care, make up a table of various patient care areas with these various characteristics in mind. Each hospital will be a bit different depending on the availability of monitors, doors, and medical gas supplies in different areas. Some locations may be able to fully meet all specifications for an ICU to care for COVID-19 ICU patients and others may only meet specifications for non-COVID-19 ICU patients. In general, these are the areas that may be considered as ICU expansion areas:

  1. Existing step-down units
  2. Cardiac care units
  3. Other med-surg nursing units
  4. Surgical pre/post-op recovery rooms
  5. Endoscopy pre/post-op recovery rooms
  6. Cardiac cath lab pre/post-op recovery rooms
  7. Operating rooms

Alternate Nursing and Respiratory Therapy Staff

Just having physical beds does not complete an intensive care unit. You have to also have nurses and respiratory therapists. In times of crisis, many hospital areas will not be active so recruiting operating room nurses, endoscopy nurses, and outpatient clinic nurses should be considered. Not all of these will be adept at caring for critically ill patients with COVID-19 ARDS so alternative staffing models need to be considered: for example, one critical care nurse could be supervising 2-3 recovery room nurses. Respiratory therapists may be more of a limiting factor and may need to be augmented with other health care workers (nurses, NPs, PAs, etc.) who are tangentially familiar with respiratory therapy duties. Also consider identifying nurses and respiratory therapists who have recently retired. EMTs may be another potential resource.


Even if you have enough beds, nurses, and respiratory therapists, if you don’t have ventilators, you cannot treat COVID-19 patients with ARDS.  So where do you find ventilators when you run out? There are several possibilities:

  1. BiPAP machines. These are not ideal but can be adapted to function similarly to a regular ventilator
  2. Children’s hospitals. COVID-19 primarily affects adults; the older the person, the sicker they tend to get. Children generally do not get as sick. Consequently, there may be extra ventilators at children’s hospitals.
  3. Home respiratory therapy companies. They may have extra ventilator inventory that could be loaned to the hospital.
  4. Home ventilator patients. Many of these patients will have a back-up ventilator on hand in case of malfunction of their primary ventilator.
  5. Gas-powered ventilators. These are often stored in regional disaster caches. The are not a great substitute for a regular ventilator but may be better than rationing ventilators in times of extreme demand.

Alternates to Critical Care Physicians

In some countries, intensive care units are staffed by anesthesiologists but in the United States, ICUs are primarily staffed by critical care physicians. If COVID-19 results in a doubling or tripling of ICU beds, then there will need to be other physicians who can step in. Some of the possibilities include hospitalists, anesthesiologists, emergency medicine physicians, and sleep medicine specialists. Often, it is not necessarily the specific specialty of the physician but instead how old they are. Most internal medicine, surgery, and anesthesiologists do several months of residency training in intensive care units and so those physicians recently out of residency may be more able to stand in for critical care physicians.

The COVID-19 pandemic is not going to last forever but the next 2 months will bring challenges to our nation’s hospitals and particularly our intensive care units. By preparing now and establishing various metrics that would trigger use of these alternate resources, we will be able to match our communities’ COVID-19 needs to the critical care resources of our hospitals.

March 29, 2020

Intensive Care Unit

The Management Of Respiratory Failure In COVID-19 Patients

Every hospital in the United States is bracing for a potential deluge of patents with COVID-19 infection and many of these patients will require admission to our country’s intensive care units. There are not enough critical care physicians to manage all of these patients so it may be necessary for doctors and nurses who do not normally manage critically ill patients to step in. Although we hope that the seemingly draconian measures our countries leaders are taking will “flatten the curve” of the prevalence of COVID-19 in the United States and minimize the demand on our hospitals, it remains possible that the critical care crisis that has occurred in Northern Italy will happen here.

The Ohio State University Medical Center is taking a multi-faceted approach to the COVID-19 outbreak and one of the tasks that I was assigned was to create a webcast that could be used by physicians around the world who need to know how to manage COVID-19 patients in the intensive care unit. Rather than repeat everything from that webcast in this post, I’m giving you a link to the 1-hour webcast by my colleague, Dr. Rachel Quaney and myself. I’m hoping that this presentation will give physicians, nurses, and respiratory therapists the tools that they will need to improve the survival of these patients who are in our ICUs. Click this link to access the webcast.

March 20, 2020

Intensive Care Unit

Is Your Hospital Doing Too Many (Or Too Few) Tracheostomies?

In my last post, I ranked the highest paying DRGs from the Medicare Inpatient Charge Dataset that lists the average charges and average Medicare payments to hospitals for different diagnoses. One of the possible conclusions from analyzing this dataset is that we may be inadvertently incentivizing doing more tracheostomies.

When a patient is admitted to the hospital with respiratory failure requiring endotracheal intubation and is placed on a mechanical ventilator, we try to get that patient off of the ventilator as quickly as possible. If the patient cannot be extubated within a few days, there are 3 options for managing that patient:

  1. Leave the patient intubated with an endotracheal tube and on a mechanical ventilator for as long as it takes for them to get better and get off of the ventilator. The advantage of this approach is that it avoids doing a surgical procedure on the patient and once that patient is off of the ventilator, they can generally be discharged home or to a nursing home for recovery. The disadvantage is that sometimes it can take many days or weeks to “wean” the patient off of the ventilator and so that patient’s hospital length of stay can be quite long, resulting in higher hospital expenses, without higher Medicare or insurance payments to the hospital. In other words, this approach can result in the hospital losing money on that patient.
  2. Have a palliative care discussion with the patient or family and discontinue life support with an expectation of death. The advantage of this approach is that it provides realistic goals of care discussion with the family so that they can make informed decisions about the patients end-of-life care. Also, this can result in a shorter hospital length of stay, thus on the surface reducing the hospital’s expenses for that patient and improving the hospital’s financial margin. The disadvantage is that the patient dies.
  3. Place a surgical tracheostomy and then transfer that patient to another level of care facility for ventilator weaning, most commonly, a long-term acute care hospital. The advantage of this approach is that the patient’s hospital length of stay is shorter, thus reducing hospital expenses and improving the hospital’s financial margin. It can also be easier and more comfortable for the patient to wean from the ventilator when they have a tracheostomy. The disadvantage is that sometimes these patients never get better and it can give the patient or their family false hope of ever getting off of the ventilator – instead of prolonging the patient’s life, it can sometimes just prolong their death.

These advantages and disadvantages are what critical care physicians and palliative medicine physicians discuss with patients and their families every day in the intensive care unit. But there is another implication of the tracheostomy that no one ever talks about: the benefit of doing a tracheostomy to the hospital’s financial margin.

In the Medicare Inpatient Charge Dataset, the average medicare payment for any given DRG (diagnosis-related group) is listed by individual hospital and as an aggregate average for all hospitals in the United States. Let’s look at the 2 common diagnoses that result in a patient being admitted to the intensive care unit, intubated, and on a mechanical ventilator: sepsis and respiratory failure and then let’s look at the financial effect of whether or not those patients get a tracheostomy.

If a patient is admitted to the ICU with respiratory failure requiring mechanical ventilation but they are not septic, they will likely get assigned DRG 189: “pulmonary edema & respiratory failure” and the average hospital will get paid $7,799 from Medicare. However, if that same patient remains on mechanical ventilation for at least 4 days and gets a tracheostomy, then the hospital can assign DRG 004 and gets paid $71,098 from Medicare – nearly $62,000 more!

If a patient is admitted to the average hospital ICU with sepsis and is only on mechanical ventilation for less than 4 days, the hospital will use DRG 872 and get paid $6,392 from Medicare. If that same patient additionally has major complications or comorbidities (which is by far the more common situation), then the hospital can use DRG 871 and gets paid $11,632 from Medicare. If the patient with sepsis remains on the ventilator for more than 4 days, then the hospital can use DRG 870 and gets paid $40,174 from Medicare – nearly $29,000 more! However, if that same patient who is on the ventilator for more than 4 days gets a tracheostomy, then the hospital can bill DRG 004 and get paid $71,098 from Medicare – nearly $31,000 more!

So, what are the financial implications of all of this? The cynic in me can identify a few:

  1. When a patient looks like they are ready to wean from the ventilator after 3 days, if you can leave them on the ventilator 1 extra day, the hospital gets paid a lot more.
  2. When a patient is on a ventilator for at least 4 days, the hospital gets paid a lot more if that patient gets a tracheostomy.
  3. When a patient is on a ventilator and palliative care discussions result in that patient having life support terminally withdrawn, the hospital will save some money by reducing the ICU length of stay. However, the hospital loses even more money by not performing a tracheostomy, thus missing out on the opportunity to bill DRG 004.
  4. When a patient is admitted with respiratory failure, the hospital is better off financially if that patient gets a tracheostomy after 4 days on a ventilator and then gets discharged to a long-term acute care hospital (as opposed to giving that patient a week or 10 days to see if they can wean from the ventilator before doing a tracheostomy.

The hospital’s greatest financial margin occurs when all patients with respiratory failure get a tracheostomy after 4 days on the ventilator and then get discharged to a long-term acute care hospital on the 5th day.

The frequency of doing tracheostomies for patients with respiratory failure may be a marker of ICU quality of care – a lower frequency indicating that the hospital is more appropriately using palliative care resources and is successfully weaning patients from mechanical ventilation before needing a tracheostomy. However, a higher frequency of tracheostomy can be a marker of greater ICU profitability.

The Scottish economist and philosopher, Adam Smith, said about capitalism that the invisible hand of free market economies will drive business decision making. I cannot help but wonder what Adam Smith would say about how the invisible hand of healthcare economics drives tracheostomy decision making.

June 6, 2019

Inpatient Practice Intensive Care Unit

Should We Stop Using Intravenous Saline?

Saline has been the go-to intravenous solution for decades. Every year in the United States, more than 200 million liters of saline are given to patients. Two studies presented at this week’s American College of Chest Physicians meeting indicate that we may have it all wrong and that we should NOT be using saline for most patients.

Saline is an isotonic crystalloid solution meaning that it has the same osmotic pressure as blood. For years, we thought that isotonicity was all that was important and that the specific electrolyte constituents did not really matter. Now, it looks like it does matter. There are 3 commonly used isotonic IV crystalloid solutions: saline, lactated Ringer’s, and Plasmalyte. They have significantly different compositions as can be seen in this table. Of particular note, the concentration of chloride in saline is about 50% higher than the concentration of chloride in blood. This has raised questions about whether this chloride can be harmful by creating a hyperchloremic metabolic acidosis or by other adverse effects of excessive chloride.

In the SMART study, 15,802 patients admitted to the ICU were randomized to receive either saline or a balanced IV solution as their maintenance and resuscitation solution. The balanced solution was either Ringer’s or Plasmalyte, at the clinicians preference (Ringer’s was used 90% of the time and Plasmalyte was used 10% of the time). The results showed that patients receiving saline had a 15.4% incidence of a composite outcome of death or adverse renal events compared to 14.3% in patients receiving a balanced solution. This translates to a 1.1% increase in the composite score of death, need for dialysis, or persistent renal dysfunction. Patients who were septic had the greatest adverse outcome difference with saline compared to a balanced solution.

In the SALT-ED study, 13,347 patients admitted to a non-ICU nursing unit were randomized to receive either saline or a balanced IV solution. The main outcome was the “MAKE30” which was a composite score of hospital-free days and adverse kidney events. Once again, the patients receiving saline did worse with a 5.6% MAKE30 versus 4.7% for the patients receiving a balanced IV solution. The overall hospital length of stay was the same. Patients receiving saline had a significantly higher blood chloride level and lower blood bicarbonate level during their hospitalization.

These are pretty compelling studies and they build on other recent studies that have indicated that patients receiving saline have a worse outcome than those receiving balanced crystalloid solutions. But what about colloid solutions? One of the most common colloid solutions in use is hetastarch, but in a trial comparing hetastarch to crystalloid solutions in resuscitation of patients with sepsis, hetastarch also was associated with an increase in renal disease and an increase in death. A second study of hetastarch compared to crystalloid in 7,000 patients in an intensive care unit with multiple diseases (not just sepsis) also showed an increase in adverse renal events with hetastarch compared to crystalloid.

At this time, we still don’t know what the ideal intravenous fluid is for resuscitation and fluid maintenance. For example, there are no head-to-head comparison studies of lactated Ringer’s solution to Plasmalyte. Furthermore, there are any number of other crystalloid solutions that could be created using biologic electrolytes that have not yet been used in medicine and it is likely that one of these could be superior to any of our existing crystalloid solutions.

The recent Baxter saline bag shortage gives us an opportunity to begin to move away from saline to balanced crystalloid solutions. But the use of saline is so ingrained in medicine that change will not come easily or quickly. However, it is now time for us as hospital leaders to promote the use of lactated Ringer’s and Plasmalyte instead of saline.

November 2, 2017


Intensive Care Unit

The Cocaine Mule

This morning, I was contacted by the emergency department. We had a patient who came in with severe agitated delirium to the point that no combination of sedative and anti-psychotic medications could control him and keep him from harming himself or ER staff members. Ultimately, he had to be intubated and placed on continuous IV sedatives. It reminded me of a similar case from a couple of years ago.

A young man was brought into the ER by police after being found sitting on a park bench, yelling and shouting at no one in particular. He had no identification on him and we had no idea who he was or where he lived. In the ER, he was tachycardic, hypertensive, and had agitated delirium to the point that he, also, had to be intubated and admitted to the ICU. When he got up to the ICU, his tox screen came back positive for cocaine. So, we decided to leave him intubated, sedated, and paralyzed overnight until the cocaine wore off and then let him wake up and extubate him.

The next day, we tapered his sedatives off and almost immediately, he became tachycardic and hypertensive and he began to flail around in his ICU bed. So, we re-started his sedatives for another 24 hours. The next day, the same thing happened: shortly after stopping his IV sedatives, his heart rate and blood pressure shot up. Once again, we restarted his IV sedation and left him on the mechanical ventilator.

Later that afternoon, when the nurses were giving him a bath, one of the nurses noted a piece of rubber protruding from his anus. They called us over and we pulled out a broken condom that was partially full of white powder. We got an abdominal x-ray and found that his entire rectum was full of balloons. With the help of some laxatives, a couple of enemas, and with oversight by our hospital security staff, we were able to get him to expel all of the balloons full of cocaine.

It turned out that the broken cocaine condom had been continuously overdosing him with cocaine so that he never got a chance for the cocaine to wear off. He was a mule. He packed his rectum with balloons of cocaine to transport across the country. Drug mules will swallow balloons or tied-off condoms full of cocaine or heroin in order to smuggle them into the country. The best way to identify them is by x-ray or CT scan – cocaine is about the same density as stool whereas heroin is closer to the density of air. In order to get the balloons out, it is best to start with a standard laxative; oil-based stool lubricants can break down some of the rubber/plastic balloons risking balloon rupture in the colon and endoscopic methods or enemas can risk breaking the bags open. In this patient’s case, we used laxatives to get risk of the remaining bags and enemas to wash out any remaining cocaine from the broken bag.

For the patient this morning, his tox screen was negative for cocaine and was just positive for opioids, fentanyl, and cannabinoids which has become a fairly routine finding in patients presenting to the ER with delirium, coma, or cardiorespiratory arrest.

October 19, 2017