ICU re-organisation to face the first COVID-19 epidemic wave in a tertiary hospital.

Because of the emergence of a virus outbreak in Mulhouse, a city located 100 km south of Strasbourg, Alsace was one of the first regions in France affected by SARS-CoV-2 and patients with a new coronavirus disease (COVID-19). From the 24th of February 24 to the 15th of July, 2020, when the intensive care units (ICUs) of Strasbourg University Hospital faced the COVID-19 epidemic, we counted 1487 deaths in our region, i.e., an associated mortality rate of 788 deaths per 1,000, 000 inhabitants.

A total of 492 COVID-19 patients were hospitalised in the ICU during these 5 months. At the peak of the outbreak on the 2nd of April, there were 182 COVID patients in the ICUs, whereas our 2,160 bed-hospital usually has only 97 ICU beds (Fig. 1 ). The ICUs employ 292 nurses, 112 care assistants and 35 permanent doctors. To cope with this unprecedented epidemic wave, we drastically reorganised our ICUs.

Fig. 1

The black dashed curve represents our hospital’s ICU bed capacity (97 beds on the 24th of February, 207 beds on the 13th of April).

The blue curve represents the number of ICU patients (maximum on the 1st of April with 189 patients).

The red curve represents the number of ICU COVID-19 patients (maximum 181 patients on the 2nd of April).

The difference between the blue and the red curves (light blue area) represents the number of ICU non-COVID-19 patients.

At the start of the outbreak on the 24th of February, 97 ICU beds were functional and there were 91 non-COVID-19 patients and one COVID-19 patient.

At the peak of the epidemic on the 2nd of April, 198 ICU beds were functional and there were 3 non-COVID-19 patients and 181 COVID-19 patients.

The complex reorganisation process can be summarised in four main actions.

Firstly, non-emergency activities were restricted and then stopped, which made possible to reduce the use of ICU beds for non-urgent care and, above all, reposition medical and nursing staff from the operating theatres to the ICUs. In fact, 7 ICU beds that were closed due to a lack of nursing staff were quickly re-opened.

Secondly, we created 108 new ICU beds by equipping continuous care units with ventilators.

Thirdly, we evacuated 59 patients to other ICUs, in France or other countries.

Fourthly, all ICUs were reorganised, in particular by using the human resources (doctors, nurses) made available by stopping non-urgent care and by massive internal and external reinforcements (178 nurses, 77 care assistants and 48 doctors). This reorganisation limited exposure to the virus and reduced the burden on the ICUs.

Specific teams were trained to 1) condition patients in the operating theatre before their transfer to the ICU (orotracheal intubation, placement of vascular catheters), 2) place patients in a prone position and 3) provide daily information to patients' families.

Together, these measures helped us keep ahead of the epidemic. This lead was extremely short during the last week of March. At that time, we only had 2 or 3 unoccupied resuscitation beds in the whole facility. This short advance allowed us to avoid the dramatic situation of not being able to provide critical care because of a lack of space in our ICUs. The decrease in non-COVID-19 patients admitted to the ICU also contributed to maintaining a lead over the epidemic. From the 24th of March to the 15th of April, fewer than 20 ICU beds were occupied by non-COVID-19 patients, a decrease of almost 80% from the usual occupancy rate. Part of this decrease directly reflects the decline in non-emergency activities, particularly surgical activities that usually require ICU hospitalisation (e.g. cardiac surgery). In addition, it is possible that some frail patients who could have benefited from intensive care may not have sought emergency care during this period for fear of being hospitalised in a "COVID-19 hospital". The human cost of these non-hospitalisations has not yet been assessed.

We observed a 20% mortality rate for all ICU COVID-19 patients. This number should be interpreted with caution as 1) we managed the most serious patients and, in particular, those on ECMO (39 patients) in our hospital and 2) patients evacuated to other ICUs were less severely ill (mortality rate 14%). Nevertheless, this number is similar to that published in Italy where there was a 26% mortality rate (although 54% of the patients were still in the ICU at the time of publication of this article) [1], in the USA where the ICU mortality rate was 35.4% [2] and in the UK where mortality varied from 29.3% to 41.4%, depending on the use of dexamethasone [3].

The reorganisation of our hospital has enabled us to stay one step ahead of the epidemic. Nevertheless, this reorganisation has had other consequences, the significance of which has not yet been fully appreciated. Patients' families had only very restricted access to their loved ones and, in particular, we were unable to accompany the families of deceased patients as we usually do. This example shows that the response to the epidemic led to working conditions that were sometimes degraded compared to the high-quality standards with which we usually expect to work.

Conflicts of interest

The authors have no conflict of interest to disclose related to this topic.