Invasive Mechanical Ventilation and Nutrition Intake in Association with Serious Outcomes among COVID-19 ICU Patients

Background: Worse clinical outcomes have been reported among critically ill COVID-19 patients who are admitted to an inpatient intensive care unit (ICU), especially those who require invasive mechanical ventilation (IMV). Patients on IMV do not receive any food or medication by mouth, during which their energy and protein needs are dependent on enteral nutrition (EN) and propofol lipid emulsion delivery (and parenteral nutrition if needed). Underfeeding throughout the ICU stay has been reported, while overfeeding due to propofol delivery is also a concern, which were both reported to be harmful to critically ill patients. However, there are limited data on the nutrition intake of COVID-19 patients during IMV and the association with in-hospital mortality. Objectives: In this study, we examined whether IMV is associated with higher mortality and longer ICU length of stay in critically ill ICU patients admitted with COVID-19. We calculated the energy and protein intake from EN and the lipid emulsion for propofol and also evaluated the associations with in-hospital mortality.

Methods: We conducted a hospital-based retrospective study, using chart review of electronic medical records among COVID-19 patients admitted to an ICU at Harborview Medical Center (HMC) in Seattle, Washington, between March 5, 2020, through October 31, 2020. Multivariable logistic regression models were used to assess the associations between IMV and in-hospital mortality. Among those who survived to hospital discharge, multivariable linear regression was used to estimate the associations between IMV and ICU length of stay. Among those who received IMV during their hospital stay, we used multivariable logistic regression to estimate the associations between the nutrition intake during IMV and in-hospital mortality.

Results: Ninety-five patients were included in the chart review. The mean age was 60.9 [standard deviation (SD): 16.1] years and 78.9% were men. About half (50.5%) received IMV during hospitalization; 33.7% died prior to hospital discharge. After adjustment for demographics, the odds of in-hospital mortality was 3.28 (95% CI: 1.12, 9.59) times higher among patients who received IMV than those who did not (p = 0.03). The association was slightly attenuated after further adjustment for comorbidities [adjusted odds ratio (aOR) = 3.01 (95% CI: 0.97, 9.39), p = 0.06], prior medication use [aOR = 3.04 (95% CI: 0.96, 9.58), p = 0.06], and Sequential Organ Failure Assessment (SOFA) score [aOR = 2.92 (95% CI: 0.84, 10.15), p = 0.09]. Among the sixty-three patients who survived to hospital discharge, we found that patients receiving IMV were in the ICU 19.4 days longer (95% CI: 11.6, 27.3), than those who did not receive IMV after adjustment for demographics, comorbidities, prior medication use, and SOFA score (p < 0.001). Among the forty-three patients who received IMV and EN support during IMV, the in-hospital mortality was 44.2%. The mean total energy intake during IMV from EN and propofol administration was 1,890 (SD: 416.2) kcal/d. On average, this met 93.4 (SD: 16.9) percent of the energy intake goal set by the HMC clinical nutritionists. The mean energy intake per ideal body weight (IBW) was 26.9 (SD: 5.2) kcal/kg/d and 20.5 (SD: 11.8) percent of total energy, on average, was from propofol administration. The mean overall protein intake was 82.9 (SD: 21.1) g/d meeting, on average, 66.9 (SD: 16.7) percent of the protein intake goal. The mean protein intake per IBW was 1.2 (SD: 0.3) g/kg/d. We found no evidence that the amount of energy intake per IBW was associated with in-hospital mortality. However, for each 0.1 g/kg/d increase of protein, the odds of in-hospital mortality was 46 (95% CI: 13, 64) percent lower among the patients with higher protein intake after adjustment for demographics, comorbidities, SOFA score, and receipt of extracorporeal membrane oxygenation (ECMO) (p = 0.01). The association remained significant after further adjustment for total energy intake per IBW [aOR = 0.47 (95% CI: 0.23, 0.98), p = 0.04].

Conclusion: Our study found that receiving IMV was associated with higher in-hospital mortality after adjustment for demographics among COVID-19 patients admitted to an ICU. IMV was an independent predictor of longer ICU length of stay. Among patients who received IMV and EN during hospitalization, deficits were larger for protein intake than for energy during IMV. Higher protein intake during IMV was a protective factor for in-hospital mortality. More clinical attention should be placed on nutrition delivery during IMV, especially protein intake. Studies with a larger sample size, more rigorous study designs, using more accurately measured daily nutritional needs, and addressing the intra- and inter-individual nutrition variability are needed to further elucidate the impact of nutrition intake during IMV on clinical outcomes of COVID-19 patients.

Keywords: COVID-19, Critically ill patient, Invasive mechanical ventilation, Enteral nutrition, Propofol