Abstract

Background. Although vaccine rollout for COVID-19 has been effective in some countries, there is still an urgent need to reduce disease transmission and severity. We recently carried out a meta-analysis and found that pre- and in-hospital use of statins may improve COVID-19 mortality outcomes. Here, we provide an updated meta-analysis in an attempt to validate these results and increase the statistical power of these potentially important findings. Methods. The meta-analysis investigated the effect of observational and randomized clinical studies on intensive care unit (ICU) admission, tracheal intubation, and death outcomes in COVID-19 cases involving statin treatment, by searching the scientific literature up to April 23, 2021. Statistical analysis and random effect modeling were performed to assess the combined effects of the updated and previous findings on the outcome measures. Findings. The updated literature search led to the identification of 23 additional studies on statin use in COVID-19 patients. Analysis of the combined studies (; 3,238,508 subjects) showed no significant effect of statin treatment on ICU admission and all-cause mortality but a significant reduction in tracheal intubation (, 95% CI: 0.54-0.99, , studies). The further analysis showed that death outcomes were significantly reduced in the patients who received statins during hospitalization (, 95% CI: 0.50-0.58, , studies), with no such effect of statin therapy before hospital admission (, 95% -1.37, , studies). Conclusion. Taken together, this updated meta-analysis extends and confirms the findings of our previous study, suggesting that in-hospital statin use leads to significant reduction of all-cause mortality in COVID-19 cases. Considering these results, statin therapy during hospitalization, while indicated, should be recommended.

1. Introduction

As of May 1, 2021, 152,038,419 people have been infected by the SARS-CoV-2 virus, the cause of Coronavirus Disease 2019 (COVID-19) [1, 2]. This translates to nearly 2% of the world population and accounts for a doubling in the number of cases over the last 6 months [3]. The number of people who have died in association with a COVID-19 diagnosis has now reached 3,194,337, which translates to a death rate that has held steady over the last 6 months at 2.1% of the cases. However, since December 2020, we have seen the rollout and administration of multiple vaccines against COVID-19 disease, due to an unprecedented and coordinated effort across the world. Although some countries with advanced vaccination programs have seen a reduction in COVID-19 case numbers, there is still an urgency to control disease spread and reduce its severity worldwide.

While waiting for increased vaccinations across the globe, one way of achieving this is through repurposing existing therapeutics. We recently carried out a meta-analysis, which identified significant reductions in intensive care unit (ICU) admission and death outcomes in COVID-19 patients taking statins [4]. Most importantly, this analysis also found that mortality was reduced most profoundly in those patients who were administered statins in-hospital (by 60%), compared to those who were already taking statins prior to hospital admission (by 23%). If confirmed, this would represent an important step forward in the treatment of COVID-19 disease severity. However, this latter finding was accounted for by only three studies with significant heterogeneity between them [4]. In addition, a recent meta-analysis by Hariyanto and Kurniawan [5] indicated that statin use has nothing to do with the composite adverse outcomes of COVID-19, including the risk of mortality. However, the study showed that despite the presence of COVID-19 infection, patients with dyslipidemia should continue to take statins as this is beneficial for cardiovascular outcomes.

Here, we provide an updated meta-analysis to further compare statin use on ICU admission, tracheal intubation, and death outcomes in COVID-19 patients. It was of particular interest to compare in-hospital vs. prehospital statin treatment on these outcomes.

2. Methods

2.1. Search Strategy

This meta-analysis was performed according to PRISMA guidelines. The searches were reconducted using Web of Science, PubMed, Scopus, and ProQuest databases for targeted articles up to April 23, 2021 (previous searches had been performed up to November 2, 2020). The population, intervention, comparison, and outcome (PICO) criteria were, respectively, patients infected with qPCR-confirmed SARS-CoV-2, statin therapy, SARS-CoV-2 patients who were not treated with statins, and intensive care unit (ICU) admission, tracheal intubation, and mortality.

The main aim was to further elucidate if statin therapy is associated with the improvement of outcomes in COVID-19 patients. The keywords were chosen as described previously to account for the various names of SARS-CoV-2 and statins [4]. For comprehensive screening of target articles, we first carried out searches without consideration of specific outcomes. Next, we identified three outcomes (ICU admission, tracheal intubation, and mortality) that could be used in a well-powered meta-analysis.

2.2. Eligibility Criteria

The inclusion criteria were (1) observational studies and randomized clinical trials testing the effect of statins on COVID-19 and (2) studies including ICU admission, tracheal intubation, and mortality outcomes. Articles were excluded if they were (1) clinical case reports, literature reviews, and preclinical investigations and (2) studies which did not incorporate statin nonusers as controls.

2.3. Quality Assessment

Assessment of study quality was performed separately by two authors (FHB and AVA), applying the Newcastle-Ottawa Scale (NOS) for cohort studies, and disagreements were resolved as above. The assessment categories were (1) selection of study groups, (2) comparability of groups, and (3) ascertainment of either the exposure (case-control studies) or outcome (cohort studies) of interest. These were rated from 0 to 3 stars as an indication of quality. This translated to a total of 0 to 9 stars per article.

2.4. Statistical Analysis

The analyses were conducted as described previously [4]. Briefly, data extraction for the main outcomes was performed, and random effect meta-analysis was conducted, by applying the restricted maximum likelihood method [6], to account for unknown, unregistered, or unpublished studies. Heterogeneity between studies was determined using the Cochran test, tau-squared (), -squared (), and -squared () statistics. Significant results and values higher than 75% were considered heterogeneous while represented perfect homogeneity [7]. Publication biases were displayed using funnel plots, and regression-based Egger’s [8] and nonparametric rank correlation-based Begg’s [9] tests were applied as a measure of small-study effects. A nonparametric “trim and fill” method was used to account for publication bias, and modified effect sizes were estimated. Common effect sizes were displayed using an odds ratio (OR) with 95% confidence interval (CI) for the outcomes, and forest plots were used to illustrate the significance of the results. Subgroup analyses were performed for those studies reporting in- or prehospital use of statins.

3. Results

3.1. Literature Search

Supplementary Figure 1 shows the flowchart of the study selection process. A total of 1,234 records were initially searched from PubMed (), Scopus (), Web of Science (), and ProQuest (), and 8 studies were identified through other sources. The full list of records was reviewed with 144 duplicate studies omitted from the study, leaving 1,090 records. Following this, articles were screened by titles and abstracts, and the full texts of the remaining 323 studies were evaluated for eligibility. This left 71 studies for the final stringent screen. Finally, 47 studies were included, which met the eligibility criteria. Odds ratios (ORs) were extracted to evaluate the effect of statin use in patients with COVID-19 on ICU admission (), tracheal intubation (), and death (). The general characteristics of included studies are given in Table 1. In addition, quality assessment of studies was done by the Newcastle-Ottawa scale (Supplementary Table 1).

3.2. ICU Admission

As shown in Figure 1(a), the risk of ICU admission between statin and nonstatin users in patients with COVID-19 was not significant. The OR from 17 studies was 0.99 (95% CI: 0.77-1.27, ) with significant heterogeneity between studies , ). Assessment for bias by Egger’s () and Begg’s () tests did not find significant small-study effects, and visual analysis of the funnel plot showed some publication bias effects (Figure 1(b)).

3.3. Tracheal Intubation

As shown in Figure 2(a), the risk of tracheal intubation between statin and nonstatin users in patients with COVID-19 was significantly different. The risk of tracheal intubation in patients with COVID-19 who used statins was significantly reduced by 27% compared with those who did not take statins. The OR from 10 studies was 0.73 (95% CI: 0.54-0.99, ), with significant heterogeneity between studies ). Small-study effects were not significant as shown by Egger’s () and Begg’s () tests, and the funnel plot suggested no publication bias (Figure 2(b)). Thus, the results were not extended to account for publication bias.

3.4. Death

As shown in Figure 3(a), the risk of mortality between statin and nonstatin users in patients with COVID-19 was not significant. The OR from the 41 studies which determined the effect of statins on mortality was 0.96 (95% CI: 0.77-1.18, ), with significant heterogeneity between studies , . Assessment for bias by Egger’s () and Begg’s () tests showed no significant small-study effects, and visual inspection of the funnel plot suggested no publication bias (Figure 3(b)). When the analysis was restricted to studies in populations with cardiovascular disease () and diabetes (), total death was found to be reduced in the former ( (95% CI: 0.45-0.85, )) but not the latter ( (95% CI: 0.46-2.41, )).

The risk of mortality in patients with COVID-19 who used statins before hospital admission was not significantly different from those who did not take statins (, 95% -1.37, , 29 studies) but with significant heterogeneity between studies , ) (Figure 4(a)). Analysis using Egger’s () and Begg’s () tests found no significant small-study effects, and the funnel plot showed no publication bias (Figure 4(b)). In the subgroup of studies conducted in populations with cardiovascular disease (; , 95% -1.02, ) or diabetes (; , 95% -3.44, ), there was no significant effect of prehospital statin use on mortality.

We also analyzed mortality risk in COVID-19 patients who received statins only after hospital admission. This allowed analysis of a new total of 7 studies which found a significant reduction in mortality compared with those who did not take statins (, 95% -0.58, ), with no significant heterogeneity between studies (, ) (Figure 5(a)). Egger’s () and Begg’s () testing showed no significant small-study effects, and the funnel plot suggested no publication bias (Figure 5(b)).

4. Discussion

Our updated meta-analysis found no significant reductions in ICU admission and mortality outcomes in COVID-19 patients who used statins, compared to those who were not on these drugs. Interestingly, a significant reduction of all-cause mortality with statins was observed in patients with cardiovascular disease; however, due to the limited number of studies included, this still needs to be confirmed. The subgroup analysis also showed that administration of statins during hospitalization was associated with a significant 46% reduction in mortality, in line with the findings of our previous study [10]. Conversely, we found that use of statins prior to admission had no significant effect on the mortality outcomes. What is additionally important, statin therapy also reduced tracheal intubation by 27%.

One possibility for these differences in mortality outcomes could be associated with the type of statin used across different studies. As the characteristic of the included studies were varied, this gives rise to bias which makes it difficult to draw firm conclusions. Expectedly, differential physiochemical characteristics of statins can affect the potency of their well-known pleiotropic actions [1118]. For example, one study found that treatment with simvastatin or atorvastatin led to a reduction in mortality of COVID-19 patients, compared to cases given pravastatin or rosuvastatin [19]. In addition, the CORONADO study showed that treatment with statins was associated with increased mortality in COVID-19 patients with preexisting diabetes [20], although another study found that statin use reduced mortality in a similar patient group [21]. Again, this might have been due to the use of different statins as information regarding the statin type was not listed in the CORONADO study. Another possibility for the lack of effect of prehospital use of statins on mortality outcomes in COVID-19 patients could be due to the preexistence of diseases such as obesity, hypertension, cardiovascular disorders, and metabolic diseases, which are significant risk factors for severe outcomes [2224]. This could be explained by the possibility that any potential benefit of statins could be nullified by the presence of comorbidities. Finally, the observed benefit in terms of reducing the incidence of tracheal intubation deserves further investigation. This benefit might imply that statin therapy is particularly beneficial in reducing the serious complications of COVID-19 like intubation which is closely related to death. This notion is in line with the observed mortality benefit in patients receiving statins during hospitalization.

The currently updated meta-analysis had several limitations. First and foremost, only associations are given since it was not possible to investigate a cause-and-effect relationship involving statin use. Secondly, we do not have data from the included studies on the preparations of statins that were used in COVID-19 patients, which is a reason we cannot make any conclusions whether there are differences in the outcomes between hydrophilic and lipophilic ones. Thirdly, potential effects of preexisting or postdiagnosis development of comorbidities such as acute respiratory distress, coagulation disorders, or insulin resistance cannot be excluded. Fourthly, the findings were not adjusted for other medication use, which may also have affected outcomes. Finally, although the number of studies that we identified which investigated in-hospital use of statins was more than doubled in this updated meta-analysis [10, 21, 2529], this was still likely to have been statistically underpowered.

In conclusion, this updated meta-analysis further supports our previous finding that administration of statins during hospitalization is associated with reduced mortality of patients diagnosed with COVID-19 disease. Thus, further clinical studies are warranted to determine the timing of statin administration, recommended preparations, and doses, as well as potential effects of preexisting medical conditions and prescribed medications on clinical outcomes in COVID-19 patients. Most importantly, such studies will provide critical insights and outline strategic measures and patient-specific treatment approaches to successfully control the current devastating COVID-19 outbreak. It is hoped that such studies will help to pave the way for better preparedness in the likely event of future pandemics. However, more randomized clinical trial studies are needed to confirm these results.

Data Availability

No original (raw) data was produced for this systematic review.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Supplementary Materials

Supplementary Table 1: quality assessment of studies by Newcastle-Ottawa scale (NOS). Supplementary Figure 1: flow chart of the study selection process. (Supplementary Materials)