Secondary outcomes
Regarding secondary outcomes, ASTRALI group showed significantly higher median of P/F ratios only at T48 (p= 0.003), T72 (p< 0.0001) and T96 (p< 0.0001). ASTRALI group showed significantly lower median of VP days (1 ± 4) than control group (4 ± 5) (p= 0.013). Both groups showed comparable medians of ICU stay (p= 0.649) and MV days (p= 0.611). These findings are supported by the knowledge of that ascorbic acid is a cofactor in the hydroxylase enzymes which are involved in the synthesis of important catecholamines (such as norepinephrine and vasopressin) and downregulation of hypoxia inducible factor 1α. (20) The small sample size used in the current study may affect the significance of outcomes such as MV days, because strong evidence that high dose IV ascorbic acid shortens MV days in critically ill patients was identified in a meta-regression analysis (2020). (45)
Zhang et al. (2021) showed sustained significant improvement of P/F ratio of ascorbic acid group (228.5 ± 72.6) than control (150.7 ± 75.3) (p= 0.01). (35) Saeidreza et al. (2021) studied the effect of IV ascorbic acid (1.5 gm/ 6hrs for 5 days) in COVID-19 patients. Results showed improved peripheral capillary oxygen saturations in ascorbic acid arm over controls. (46)
In Wang et al. (2019) meta-analysis, results showed that ascorbic acid was associated with a decreased MV days, but whether this reflected improved oxygenation was uncertain. (47) Another meta-analysis in 2015 showed significant reduction of VP requirements in critically ill patients. (48) In Zabet et al. (2016) study, IV 25 mg/kg every 6 hrs ascorbic acid in septic shock patients showed lower MV days (standardized mean difference, SMD = -0.73 [95% CI: -1.5-0.04]) and reduced vasopressor requirements than control group (49.64 ± 25.67 Vs. 71.57 ± 1.60 h, p = 0.007). (49)
In contrast, Matthew Li et al. (2021) did not to find any improvement in VP use or ICU days in ascorbic acid group (1.5 gm/6 hrs) over controls with COVID-19. (50) In CITRIS-ALI trial (2019), ascorbic acid group showed similar oxygenation index and VP use at T96 and T168 of follow up. (43) Also, Flower et al. (2014) showed nearly similar MV days (SMD=0.15 [95% CI: -0.70–1.000]) and VP requirements (SMD= -0.49 [95% CI: -1.35–0.37]) (51)
In terms of mortality, after Kaplan-Meier analysis, survival curves were compared. ASTRALI group showed significantly higher 7-days survival than control group (p= 0.007). Then, Cox regression was used as multiple comparisons (HR= 0.305 [95% CI: 0.120–0.775],p  = 0.008). ASTRALI group was associated with 75.7% reduction of 7-days mortality (OR=0.243 [95% CI: 0.082 – 0.721], p= 0.011). This reduction in mortality was partially explained by IL-10 levels after treatment as it was associated with 6.1% reduction (OR=0.939 [95% CI: 0.894 – 0.987],p= 0.013). Receiving ascorbic acid “ASTRALI group” was a fair predictor for 7-days survival (AUROC=0.668, [95% CI: 0.538–0.797],p = 0.019). Also, post treatment IL-10 at a cut-off value of equal or more than 19.55 pg/mL was a good predictor for 7-days survival (AUROC=0.708, [95% CI: 0.584–0.831], p = 0.004).
This finding reflects that ascorbic acid may be associated with better 7-days survival via enhancing IL-10 levels in such patients Although our findings showed short term mortality benefit, results failed to find this mortality benefit after 28-days follow-up. This may be due to the small sample size and long-term ICU complications evolved due to heterogeneity in comorbidities between groups such as diabetes. The prevalence of diabetes was nearly doubled in ASTRALI group than controls (45% Vs. 20%) (p= 0.031).
CITRIS-ALI trial (2019) showed that 28-days mortality was lower in the ascorbic acid group (29.8%) than placebo (46.3%) (p= 0.03). Although this study failed to show different primary outcomes, the Kaplan-Meier survival curves for the two groups were significantly different (p= 0.01) in favor of ascorbic acid group than control. (43) Wang et al. Meta-analysis (2019) showed that IV 3-10 gm/day ascorbic acid for critically ill patients was associated with reduced mortality (OR=0.25 [95% CI: 0.14–0.46)]). (47) In Zabet et al. (2016) study, IV ascorbic acid showed significantly reduced 28-days mortality than control group (14.28% Vs. 64.28%, p  = 0.009) (OR=0.09 [95% CI: 0.01 – 0.59]). (49)
In contrast, Zhang et al. pilot trial (2021) showed no difference in 28-days mortality between ascorbic acid group (26 [95% CI: 9.0-28.0]) and control group (22 [95%CI: 8.50-28.0]) (p=  0.57). (35) But this trial showed imbalanced sex distribution across groups. Control group was supported with significantly positive fluid balance than ascorbic acid group on day 3 (463 Vs. -240, p= 0.02). Also, treatment with high dose ascorbic acid was started 10 days after symptoms.
Also, Matthew Li et al. (2021) showed increased in-hospital mortality in ascorbic acid group (88% Vs. 79%, p= 0.049) than control group. (50) But this study had several limitations starting from very small size (n=8 for ascorbic acid group and n=24 for control), retrospective design, and bias liable design. Very late ascorbic acid treatment was initiated, and heterogeneity of baseline characteristics was extended even after propensity score matching to form a control group.
Ferrón-Celma et al. (2009), studied the effect of IV 450 mg/d ascorbic acid in septic Patients after abdominal surgery. No difference was detected in 7-days mortality (OR=2.25 [95% CI: 0.38 – 13.47]). (52) In addition to the small dose of ascorbic acid used, this study had very small sample size (n=10) and blinding of outcomes assessment was not clear.