Vitamin D Supplementation in Critically Ill Patients: A Meta-Analysis of Randomized Controlled Trials

Randomized controlled trials (RCTs) have reported conflicting outcomes with the use of vitamin D in critically ill patients. With reporting of newer RCTs, we conducted this updated meta-analysis. Electronic databases were searched for RCTs comparing vitamin D with placebo in critically ill patients admitted to the intensive care unit (ICU). A random-effects meta-analysis was used to calculate the risk ratio (RR) and standardized mean difference (SMD) with a 95% confidence interval (CI). Eleven RCTs with a total of 2,187 patients (vitamin D: n = 1,120; placebo: n = 1,067) were included. Vitamin D when compared to placebo was associated with the decreased duration of mechanical ventilation (SMD = -0.50; 95% CI = [-0.97, -0.03]; p = 0.04) and ICU stay (SMD = -0.60; 95% CI = [-1.03, -0.16]; p = 0.007) without any difference in the mortality (RR = 0.85; 95% CI = [0.68, 1.04]; p = 0.12) and length of hospital stay (SMD = -0.21; 95% CI = (-0.51, 0.09); p = 0.18]. Subgroup analysis showed that parenteral vitamin D may reduce the risk of mortality (RR = 0.54; 95% CI = [0.35, 0.83], p = 0.005). Vitamin D supplementation in critically ill patients decreases the duration of mechanical ventilation and ICU stay. Further studies should identify specific groups of patients who will derive the most benefit from vitamin D supplementation.


Introduction And Background
Vitamin D deficiency is commonly noted in critically ill patients and has been shown to be associated with worse clinical outcomes. Studies have shown that vitamin D deficiency is associated with increased risk for respiratory infections, increased length of stay, and mortality [1][2][3]. Apart from its role in bone and calcium metabolism, vitamin D works as an immuno-modulator, decreases inflammatory cytokines, and may have a role in lung protection [4,5]. Vitamin D supplementation has been shown to be beneficial in many noncritical care settings like colon cancer and inflammatory bowel disease [6][7][8].
Randomized controlled trials (RCTs) assessing the role of vitamin D in critically ill patients had shown conflicting results [9][10][11][12]. Some trials suggested possible benefits in mortality, while others suggested decreased intensive care unit (ICU) stay. Current guidelines do not suggest routine measurement of vitamin D levels or its supplementation in critically ill patients. Previous meta-analyses have also shown variable results regarding the role of vitamin D supplementation; some of the major limitations of those metaanalyses were smaller study sample sizes and a limited number of included studies [13][14][15]. One of the recently published and the largest RCT till date, the VIOLET trial, showed a numerical increase in the mortality in patients receiving vitamin D supplementation compared to placebo [16]. With the reporting of newer RCTs and concern for worse outcomes with vitamin D supplementation, we performed this updated meta-analysis to evaluate the role of vitamin D in critically ill patients.
Two reviewers (SS and SS) independently screened the study reports for eligibility, assessed the risk of bias, and collected data from each eligible study. Any differences between the two reviewers were resolved with consensus after discussion with the third reviewer (AR). From the eligible RCTs, data on study characteristics like study design, year of publication, inclusion and exclusion criteria, sample size, follow-up period, baseline patient characteristics, treatment data, and clinical outcomes at the longest available follow-up were obtained. Subgroups were made based on the route of vitamin D administration: enteral or parenteral route. The outcomes of interest were mortality, duration of mechanical ventilation, ICU stay, and hospital stay.

Study Analysis
This meta-analysis was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [17]. We used Cochrane Review Manager, version 5.4 (Cochrane, London, United Kingdom), for study analysis [18]. Mean and standard deviations were extrapolated from median and range using the statistical method outlined by Hozo et al. [19]. For dichotomous clinical data, pooled risk ratio (RR) and 95% confidence interval (CI) were calculated using the random-effects models with the Mantel-Haenszel method. For continuous variables, we computed standardized mean difference (SMD) with 95% CI using the inverse variance method. A p-value of 0.05 or less was considered statistically significant. Study heterogeneity was assessed by calculating I-squared statistic; heterogeneity was considered significant in the case of I 2 > 50%. Sensitivity analysis was performed by excluding each trial from the final analysis. Forest plots were generated to demonstrate the relative effect size of vitamin D supplementation versus placebo for individual clinical endpoints.

Results
The initial search yielded 1,267 studies out of which 11 RCTs were identified [9][10][11][12]16,[20][21][22][23][24][25]. Figure 1 shows the PRISMA flow diagram for search strategy. Data for one of the trials was accessed using an abstract [22]. A total of 2,187 patients with 1,120 in the vitamin D arm and 1,067 in the placebo arm were included in the final analysis. The mean duration of follow-up was two months for all the trials except one which was until hospital/ICU stay [23]. The mean age of the study population was 58 years, and 61% of the study population were men. Vitamin D formulation was enteral (oral or via feeding tube) in six studies; parenteral (intravenous or intramuscular) in four studies, and both in one study. Baseline vitamin D levels were assessed in six studies. Among the 11 studies, six studies included patients admitted with any medical cause requiring ICU level of care, three studies included sepsis/septic shock patients, one study with only neurocritical care patients, and one study included patients exclusively with ventilator-associated pneumonia.  Eleven studies reported mortality with a total of 534 deaths (vitamin D group: n = 260 and placebo group: n = 274) out of the 2,187 patients. Compared to placebo, vitamin D supplementation was associated with a decrease in mortality, but the difference was statistically not      Table 2 shows sensitivity analysis by excluding each trial from the final analysis. Exclusion of the VIOLET trial shows a significant reduction in the risk of mortality in the vitamin D group compared to placebo.      Source: References [9][10][11][12]16,[20][21][22][23][24][25].

Discussion
In this present meta-analysis of 2,187 patients from 11 RCTs, we evaluated the role of vitamin D supplementation in critically ill patients. The main findings are that vitamin D supplementation in critically ill patients was associated with reduced duration of mechanical ventilation and ICU stay. There was no significant difference noted in mortality and length of hospital stay. The parenteral route of vitamin D administration was associated with a reduction in the risk of mortality, duration of mechanical ventilation, and ICU stays as noted in the sub-group analysis of limited patients.
Increased incidence of vitamin D deficiency has been shown to be associated with critically ill patients with some studies reporting a prevalence close to 80% (1-3, 26-28). Vitamin D plays an important role in inflammatory pathways [4]. It inhibits various inflammatory cytokines such as IL-1 α, IL-1β, tumor necrosis factor (TNF)-α and affects T-lymphocyte differentiation by inhibiting IL-12 release from dendritic cells [29].
Decreased vitamin D levels are associated with a lower level of cathelicidin peptide that acts against infectious agents and has been associated with worse outcomes [30,31]. Several retrospective and prospective studies have reported that reduced levels of vitamin D were associated with increased mortality among critically ill patients [32][33][34][35]. Critically ill patients with sepsis, acute kidney injury, and other medical conditions have worse outcomes when they have concomitant vitamin D deficiency [32,36,37]. Increased rate of infection, ICU length of stay, duration of mechanical ventilation, and hospital stay along with increased health-care costs have been associated with vitamin D deficiency in critically ill patients [38,39].
So far, RCTs have been inconclusive regarding vitamin D supplementation in critically ill patients. Earlier trials had several limitations including variability among the trials in terms of dosages of vitamin D, route of administration, and the limited number of patients included. The VITdAL-ICU (vitamin D deficiency in critically ill patients) trial was one of the first large-scale trials, which looked at two different vitamin D deficiency groups (25(OH)D <20 ng/ml and <12 ng/ml). They found no overall difference in the six-month mortality but did find statistically significant mortality reduction in the subgroup analysis of the severely deficient vitamin D group (<12 ng/ml) [9]. However, the VIOLET trial failed to show any mortality benefits [19]. Important differences between the VITdAL-ICU and the VIOLET trial include the inclusion of patients early in the course of critical illness, using a vitamin D cutoff of <20 ng/ml and not providing additional vitamin D supplementation in the VIOLET trial. Our largest meta-analysis failed to show any mortality benefits, but unlike the VIOLET trial, we found a non-significant trend toward decreased mortality in the vitamin D group. Interestingly, sensitivity analysis after excluding the VIOLET trial shows a reduction in mortality with vitamin D supplementation. In our subgroup analysis, parenteral vitamin D supplementation showed significant mortality benefits; however, this subgroup involves less than 15% of the total study population, and none of the major trials (VITdAL-ICU or VIOLET) were included.
Our meta-analysis shows vitamin D supplementation reduces the duration of mechanical ventilation and ICU stay in critically ill patients, though the results are associated with significant heterogeneity. These results are important as both the long duration of mechanical ventilation and ICU stay are associated with increased long-term morbidity and mortality [40,41]. Additional sensitivity and subgroup analyses show a reduction in mechanical ventilation duration mainly in the parenteral subgroup, while the reduction in length of ICU stay is noted in both routes of vitamin D supplementation. The VIOLET trial was not included in these analyses as the length of ICU stay was not reported in the trial and only a small portion of patients in the trial had post-randomization mechanical ventilation.
When compared to previous meta-analyses [13][14][15], this updated analysis has some major differences. First, our study has a much larger population; second, many newer and more diverse clinical trials are included in this analysis, and finally, our results demonstrate benefits in terms of the duration of mechanical ventilation and ICU stay. Currently, two large-scale RCTs are ongoing, which will further improve the evidence regarding the role of vitamin D in critically ill patients [42,43].

Conclusions
Vitamin D supplementation in critically ill patients decreases the duration of mechanical ventilation and the length of ICU stay. Vitamin D may reduce mortality in specific groups of critically ill patients. Our study has some important limitations. Despite including 11 trials, we still have a small population of patients in this study. There are many disparities among the trials in terms of dose and route of vitamin D supplementation, checking vitamin D levels in some trials, outcomes reported, and duration of follow-up. We included a very diverse population of critically ill patients like patients with neuro-critical illness, sepsis, ventilatorassociated pneumonia, and trauma patients. All these differences led to significant heterogeneity among trials. Nevertheless, the results of this study are important and are hypothesis-generating regarding the duration of mechanical ventilation and ICU stay and the role of parenteral vitamin D supplementation. Further large-scale RCTs are needed to identify specific groups of critically ill patients who will most likely benefit from vitamin D supplementation.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.