Study design
This is a multicenter, non-interventional, prospective study in critically ill patients aged > 18-years with COVID19 (diagnosed according to reverse transcriptase-polymerase chain reaction (RT-PCR) obtained from nasopharyngeal or throat swabs), who were admitted to the ICU at two tertiary hospitals in Saudi Arabia from March 01, 2020, until December 31, 2020. We aimed to enroll as many patients as possible, with no predefined sample size. Patients were excluded if they did not receive corticosteroid therapy during the ICU period or labeled as "Do-Not-Resuscitate" code status within 24 hours of ICU admission.
Participants
Eligible patients were classified into two groups based on the timing of dexamethasone initiation during ICU stay (early vs. late initiation). The definition of early versus late corticosteroids initiation is debatable in critically ill patients in general and, more specifically, in COVID19 patients. We decided to choose three days cutoff margin for early vs. late initiation definition based on clinical judgment. It is well known that COVID19 related lung injury and its associated hyperinflammatory and overreacting immune response occur early in ARDS presentation. In our study, early initiation was defined as systemic dexamethasone initiation before ICU admission (e.g., wards) and continuation during ICU stay or newly initiated within three days of ICU admission, whereas late initiation was defined as dexamethasone after three ICU admission days. All included patients have received a dose of 6 mg IV once daily of dexamethasone or its equivalent based on the national protocol of the ministry of health (MOH) in Saudi Arabia 14. Patients were followed during ICU stay. The study was approved by King Abdullah International Medical Research Center in July 2020 (Ref.# RC20/430/R).
Setting
This study was conducted in two tertiary governmental hospitals; King Abdulaziz Medical City, Riyadh, and King Abdulaziz University Hospital, Jeddah. The primary site for this multicenter study was King Abdulaziz Medical City (Riyadh).
Data collection
We collected demographic data (See additional file 1), comorbidities, vital signs and laboratory tests, Acute Physiology and Chronic Health Evaluation II (APACHE II), Sequential Organ Failure Assessment (SOFA) and Nutrition Risk in Critically ill (NUTRIC) scores, Glasgow Coma Score (GCS), acute kidney injury, fluid balance, the needs for mechanical ventilation (MV) and MV parameters (e.g., PaO2/FiO2 ratio, FiO2 requirement) within 24 hours of ICU admission. Also, renal profile, liver function tests (LFTs), coagulation profile (i.e., INR, aPTT, fibrinogen), and inflammatory markers (CRP, procalcitonin) within 24 hours of ICU admission were collected. Tocilizumab use was recorded for the eligible patients. All patients were followed until they were discharged from the hospital or died during the in-hospital stay, whichever occurred first.
Outcomes
The primary endpoint was to evaluate the timing of dexamethasone initiation and its association with 30-day mortality in critically ill patients with COVID 19. The secondary endpoints were to assess the in-hospital mortality, hospital LOS, ICU LOS, MV duration, and, ICU-related complication (s) during ICU stay (i.e., acute kidney injury (AKI), liver injury, respiratory failure requires MV, thrombosis/infarction).
Definition (s)
- Acute kidney injury (AKI) was defined using Acute Kidney Injury Network (AKIN) definition15.
- Thrombosis/infarction was defined using the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD10-CM) code (i.e., myocardial infarction (MI), ischemic stroke, pulmonary embolism, deep vein thrombosis) during ICU stay16.
- Respiratory failure was defined as either hypoxemic respiratory failure (PaO2 < 60 mm Hg with a normal or low arterial carbon dioxide tension (PaCO2) or hypercapnic respiratory failure (PaCO2> 50 mm Hg) that requires mechanical ventilation.
- Acute liver injury was defined as alanine aminotransferase (ALT) exceeding three times the upper limit of normal or double in patients with elevated baseline ALT during the hospital stay.
Data management and Statistical analysis
We presented categorical variables as number (percentage), numerical variables (continuous variables) as mean and standard deviation (SD), or median and lower quartile (Q1) and upper quartile (Q3), as appropriate. The normality assumptions were assessed for all numerical variables using a statistical test (i.e., Shapiro–Wilk test) and graphical representation (i.e., histograms and Q-Q plots). We assessed model fit using the Hosmer-Lemeshow goodness-of-fit test. No imputation was made for missing data as the cohort of patients in our study was not derived from random selection.
We compared categorical variables using the Chi-square or Fisher exact test. We compared the normally distributed continuous variables using student t-test and other non-normally distributed continuous variables with the Mann-Whitney U test. Baseline characteristics, baseline severity, and outcome variables were compared between the two groups. Multivariate logistic and generalized linear regression were used to find out the relationship between the timing of initiation with different outcomes considered in this study after adjusting for patient’s age, gender, and mechanical ventilation needs within 24 hours of ICU admission 4. The odds ratios (OR) and estimates with the 95% confidence intervals (CI) were reported for the associations. We considered a P value of < 0.05 statistically significant, and we used SAS version 9.4 for all statistical analyses.
Propensity score matching Procedure (Proc PS match) (SAS, Cary, NC) with a 1: 4 ratio were used to match patients in the late group to early Dexamethasone group. A greedy nearest neighbor matching method was used in which one late use of dexamethasone (control) is matched with four patients in the early use of dexamethasone (treated) group; this eventually produces the smallest within-pair difference among all available pairs with treated patients. These patients are matched only if the difference in the logits of the propensity scores for pairs of patients from the two groups is less than or equal to 0.5 times the pooled estimate of the standard deviation.