ICU Management & Practice, Volume 20 - Issue 2, 2020


COVID-19 poses several challenges and has made it difficult to measure ICU performance. Sticking to evidence-based interventions can go a long way in improving outcomes and resource utilisation.

How it Started

Taking care of a COVID-19 patient comes with an initial surprise: there is no standard mindset to guide the physician’s work before entering a COVID-19 patient’s room. After some initial hesitation, most physicians have dubious feelings. It is a mix of both relief (“it is just another sick patient, I have treated people before”) and fear (“what should I do next?”).

There were few reports early on in the pandemic. Scattered information from “happy hypoxaemic” patients, coupled with suggestions that early intubation was of key importance, the fear of aerosols, the multiple suggestions that anticoagulation should be done due to an abnormally high number of thrombotic events and even other proposed pharmacological treatments (from hydroxychloroquine to tocilizumab, passing by corticosteroids and other drugs) appeared in discussions, forums and some case reports. They were sometimes demanded by patients and families and even politically propelled. This all contributed to shifting the discussion from keeping standards of care to a search of finding a single treatment that would halt the disease.

As soon as COVID-19 blew up in Brazil in May, I was reallocated to what we call a “routine intensivist” (an intensivist that goes almost every working day to the ICU) to a dedicated 10-bed COVID-19 ICU in my former teaching hospital (Hospital das Clínicas from the University of São Paulo). This large teaching hospital, built more than 50 years ago, became a dedicated COVID-19 hospital. Many ICU beds were created, peaking close to 300. I was privileged to work under few resource constraints, but eventual drug shortages and irregular availability of some specific supports (such as continuous renal replacement therapy [RRT] and high-flow nasal catheters [HFNC]) were expected. Concerns were beyond patient care, including staff protection and wellbeing, and resource management.

How is my Unit Doing?

A constant concern of the intensivist is whether they are doing the right thing for their patients. In a scenario where ICU admissions are many and mortality is high, the impeding sense of despair is inevitable. Under normal circumstances, measuring and optimising ICU care is a long-term process, with trends in standardised mortality ratio and standardised resource use being tracked and adjustments being made (Salluh and Soares 2014; Rothen 2017; Zampieri 2020). This is impossible in a pandemic, let alone a pandemic of an unknown disease in a new ICU that has just opened, with a multidisciplinary team that never worked together before.

Traditional markers of performance seemed misleading. Crude mortality is not a useful parameter, specially if you are receiving patients late or only the sickest ones (which are eventually granted an ICU bed). Waiting for standardised mortality ratio and standardised resource use to become available was not feasible, because these measures take time and because they rely on well calibrated illness severity scores (which were unavailable; it was not to be expected that traditional illness severity scores, such as SAPS 3, would perform well for a single disease) (Rewa 2018). Need for readmission is also problematic, because pressure for beds was constant and because you are discharging patients to a ward staff that is as inexperienced with COVID-19 as you. Length of mechanical ventilation is also cumbersome, because few data were available. Although this data should all be measured, they should not (and were not) interpreted as direct performance measurements in this situation.

Rational Implementation

The key to delivering best care was, once again, to develop and implement a routine that guaranteed a minimum evidence-based acceptable care. This is the key point: “minimum.” It became clear that while we had few to no evidence on how to manage the COVID-19 disease, we had a good deal of evidence on how to manage patients with multiple organ failure, including respiratory failure. That became our focus in the new unit. Simply do what we know as supported by evidence and rest assured that, on average, we would be doing the best for most patients. We adopted the “zentensivist” approach with open arms (Siuba 2020). What a graceful surprise we had when we realised that we could do a lot by focusing in doing well what we knew and by being more patient and less reactive. Summarising previous evidence as you prepare for an unknown disease can be relieving for the staff and allows one to drift attention away for elusive therapies or magic bullets. A brief summary of what was known before the pandemic, caveats on applying evidence on COVID-19 considering resource optimisation and the proposed solution are shown in Table 1.

Improving patient safety also required us to establish a “tracking” algorithm for unexpected events. Tracing back the roots of a problem following a structured approach (like a fish bone diagram) can be helpful to find opportunities to perform simple interventions and, also equally important, to understand that not all adverse events can be completely solved solely by changing ICU behaviour. You have to accept that you cannot change everything and that some structural problems will prevail despite the best efforts. One example is shown in Figure 1, based on a real case experienced during the pandemic. Barotrauma may occur after intubation and it is tempting to attribute it to operator inexperience, improper airway management, poor sedation/blockade, excessive bag-mask ventilation, etc. However, the root cause of the problem may be much more complicated and include a complex chain of events that are not limited to the ICU itself. Tracing these problems can not only help improve immediate aspects of care but also provide feedback to the institution and decrease staff burden or feeling of guilt.

Finally, we also anticipated that some outstanding success cases could be problematic. This seldom discussed aspect of medical care is entangled with the hot hand fallacy. Some interventions performed as an exception attempt to improve a very sick patient's condition would eventually be seen by staff as effective if the patient survived. This could result in widespread intervention use in other scenarios with unknown clinical benefit (generalisation bias). Situations like this would be (and were) inevitable. When “off the grid” interventions seemed successful, one should temper their hardihood and calmly explain to staff that this was an exception, not a rule, and avoid creating new directions of protocol based on previous successful cases. Debriefing successful cases should be done with the same scrutiny as debriefing unsuccessful ones.

Final Remarks

The goal of any ICU remains the same, whether during a pandemic or not: produce survivors with the best quality of life possible at the fastest speed achievable by applying the most current evidence while using the smallest amount of resources. Sticking to evidence-based practices and reducing a hyper-reactive state allows one to remain focused on what is known and beneficial and may be able to improve both outcomes and resource utilisation. 

Key Points

  • COVID-19 has resulted in significant challenges for doctors in intensive care.
  • The search for a cure has shifted the discussion from maintaining standards of care to a search of finding a single treatment.
  • Under normal circumstances, measuring and optimising ICU care is a long-term process but this becomes impossible during a pandemic.
  • While we had few to no evidence on how to manage the COVID-19 disease, we had a good deal of evidence on how to manage patients with multiple organ failure, including respiratory failure.
  • Improving patient safety also required us to establish a “tracking” algorithm for unexpected events.


Cardoso LT, Grion CM, Matsuo T et al. (2011) Impact of delayed admission to intensive care units on mortality of critically ill patients: a cohort study. Crit Care, 15(1):R28. doi:10.1186/cc9975

Iwashyna TJ, Boehman A, Capelcelatro J et al. (2020) Variation in Aerosol Production Across Oxygen Delivery Devices in Spontaneously Breathing Human Subjects.

Krag M, Marker S, Perner A et al. (2018) Pantoprazole in Patients at Risk for Gastrointestinal Bleeding in the ICU. N Engl J Med, 379(23):2199-2208. doi:10.1056/NEJMoa1714919

RECOVERY Collaborative Group, Horby P, Lim WS et al. (2020) Dexamethasone in Hospitalized Patients with Covid-19 - Preliminary Report [published online ahead of print, 2020 Jul 17]. N Engl J Med. doi:10.1056/NEJMoa2021436

Rewa OG, Stelfox HT, Ingolfsson A et al. (2018) Indicators of intensive care unit capacity strain: a systematic review. Crit Care, 22(1):86. doi:10.1186/s13054-018-1975-3

Roca O, Caralt B, Messika J et al. (2019) An Index Combining Respiratory Rate and Oxygenation to Predict Outcome of Nasal High-Flow Therapy. Am J Respir Crit Care Med. 199(11):1368-1376. doi:10.1164/rccm.201803-0589OC

Rothen HU, Stricker K, Einfalt J et al. (2007) Variability in outcome and resource use in intensive care units. Intensive Care Med, 33(8):1329-1336. doi:10.1007/s00134-007-0690-3

Salluh JI, Soares M. (2014) ICU severity of illness scores: APACHE, SAPS and MPM. Curr Opin Crit Care, 20(5):557-565. doi:10.1097/MCC.0000000000000135

Siuba MT, Dugar S, Shekar K (2020) Treading Lightly in a Pandemic: #Zentensivist Reflections on COVID-19 [published online ahead of print, 2020 Apr 21]. Chest. 2020;S0012-3692(20)30745-5. doi:10.1016/j.chest.2020.04.013

Zampieri FG, Soares M, Salluh JIF (2020) How to evaluate intensive care unit performance during the COVID-19 pandemic. Avaliação do desempenho de unidades de terapia intensiva durante a pandemia da COVID-19. Rev Bras Ter Intensiva, 32(2):203-206. doi:10.5935/0103-507x.20200040