BackgroundThe optimal isolation time of COVID-19 patients in intensive care unit (ICU) is debated. We investigated the impact of two different COVID-19 patient isolation time strategies on healthcare workers (HCW) contamination, intensity of nursing care and potential associated adverse events.MethodsWe prospectively included all consecutive COVID-19 patients and HCW in our ICU in the first two pandemic waves (March to May 2020 and August to November 2020). Specific isolation measures for COVID-19 patients were released after two negative RT-PCR assays in the first wave and 14 days after the onset of symptoms in the second wave. Contamination of HCW was assessed at the end of each pandemic wave by combining both a RT-PCR assay and a serological test.ResultsOverall, 117 COVID-19 patients and 73 HCW were included. Despite an earlier release from isolation after ICU admission in the second than in the first wave [6 (4–8) vs. 15 (11–19) days, p < 0.01], the proportion of HCW with a positive serological test (16 vs. 17%, p = 0.94) or with a positive RT-PCR assay (3 vs. 5%, p = 0.58) was not different between the two waves. Although a lower nurse-to-bed ratio, the intensity of nursing care was higher in the second than in the first wave. A longer isolation time was associated with accidental extubation (OR = 1.18, 95%CI:1.07–1.35, p = 0.005) but neither with ventilator-associated pneumonia nor with dysglycemia.ConclusionA shorter isolation time of COVID-19 patients in ICU was not associated with higher HCW contamination, while a longer isolation time seemed to be associated with higher accidental extubation.