DISCUSSION
First of all, we would like to evaluate whether our analysis was consistent and reliable. In this point, lab parameters that are known as “mortality/severity indicators” such as WBC, PLT, CRP, PCT, SEDIM, D-dimer, LDH, Fibrinogen, Troponin, ALT, AST, and Ferritin were compared between the patient and control group (12) (13). The comparison analysis revealed no significant difference regarding neither of these parameters and we consider that this result is a reliable indicator for the comparability of DEX between these groups. Besides, since no significant difference was found with respect to these lab parameters between the groups, our result revealing no significant difference regarding mortality would also be expected and is compatible with other analyses. As for the effects of applied treatments upon mortality rates, most of the patients received standard therapies that were recommended by the guidelines of the Ministry of Health in our country at that time. The recommended therapies for the patients with severe COVID-19 were favipiravir, steroids (dexamethasone ≥ 8mg of methylprednisolone ≥ 40 mg), and LMWH at the time at which the study group was treated in the ICUs. As you can see from Table 3 more than 90% of the patients received these standard therapies (percentage of the patients receiving dexamethasone + pulse steroid was between 85-90) revealing that there was no significant difference regarding the applied treatments between the groups. There was also no significant difference in terms of median ages, gender distribution, place of treatment (all patients were treated in the ICUs), and intubation rates between the groups. Therefore, without the effect of these parameters and applied treatments except DEX, the comparison of DEX treatment and its effects on the outcomes in our study group is reliable.
Nowadays, it is very well know that advanced age, male gender, and comorbidities such as DM, HT, and CAD are risk factors for advanced disease in COVID-19 (14). Concurrently, in our study group which consists of patients with severe COVID-19, median age was 69.5. While the median age was higher in the control group (70) than in the patient group (67), this difference did not reach a significant level (p=0.094, p>0.05). Male predominance was seen in both groups as expected. In a prospective observational study in which 63 of 89 patients with severe COVID-19 were treated with DEX, HFNC and long periods of PP, mean age was 67 ± 12 and 74% of the treatment group consisted of male patients (15). These results are very similar to our study. We would also like to point out that PP up to 16 hours a day was also applied to all patients in our study who were able to tolerate as a part of standart therapy.
When we made a literature search regarding to the articles studying theurapeutic value of DEX in patients with COVID-19, we could not find any randomised controlled trials conducted with large patient populations. Except one study which we mentioned above (15), a few case reports, case series, and letters to editor constitute all studies about this issue (15) (16) (17) (18) (19) (20) (21). Among these, the study conducted by Taboada et al. was the most appropriate candidate to compare our results with. In that study, clinical characteristics of 63 patients treated with long PP, DEX, and HFNC, 20 of whom with failed therapy and 43 of whom with successful therapy were compared. 20.6% of the patients either died or remained in the ICU while 30.2% of the patients required endotracheal intubation. In a case series in which 11 patients were treated with helmet CPAP in the prone position with DEX infusion, 2 of these patients died and 3 of them required invasive ventilatory support (18% and 27% respectively)(17). However, in our study, the intubation rate was 72.4% and the mortality rate was 74.6% for all patients. This huge difference might have been caused by a few factors. Firstly, COVID-19 daily cases were very high in our country at that time at which our study was conducted (22). Therefore, even some patients with SpO2<90% and/or respiratory rate (RR)>20/min had to be treated in the ward since the ICUs were full of patients with severe COVID-19. This situation might have caused a delay in the transfer of the patients from emergency rooms or wards to the ICUs. And this delayed start of ventilatory support and DEX infusion might have caused these high mortality and intubation rates in our study group. However, in the beforementioned studies, Paternoster et al. and Taboada et al. reported that they dedicated special beds called high dependency units (HDU) for the patients requiring helmet CPAP and as soon as the SpO2 of the patients dropped to below 90% or PaO2/FiO2 dropped to below 200 mmHg they transferred them to ICUs. This rapid transfer process and easy access to intensive care resources might have been the reason for their lower mortality and intubation rates. Secondly, the known side effects of DEX infusion are mostly cardiac such as bradycardia and hypotension (23). Although we could not reach the data regarding the discontinuation rate of DEX during the treatment from the medical files of the patients, as ICU doctors we had to decide discontinuation of DEX infusion on several occasions especially due to bradycardia (HR<50 beat/min). If the intolerance rate would have been higher in our study group, this could have caused shorter durations of DEX infusion and NIMV therapies and thus, higher mortality and intubation rates.
None of the studies mentioned above compared the parameters of the patients who received DEX therapy versus patients who received standard therapy in patients with COVID-19. From this point of view, our study is unique and can contribute to the literature. Although mortality and intubation rates were lower numerically in our study group, the difference was not statistically significant. Although there was no significant difference regarding the duration of intubation between the groups, the time between ICU admission and intubation was found to be longer in the patient group (Table 1). This might be the most important result of our analyses in this study since treatments to avoid intubation or prolong the time before intubation are of vital importance and potentially life-saving due to high mortality rates reported for intubated patients with severe COVID-19 (12) (13). With this result, we think that DEX proved its therapeutic value even if not as a mortality and intubation rate reducer, but as a time prolonger before intubation in our study. The median time between ICU admission and intubation was found as 3 days in a similar study in which all 63 patients received long PP periods and HFNC treatment (15). Maybe this time prolongation effect before intubation would rebound on future trials with large sample sizes as decreased intubation and mortality rates.
Comparison of lab parameters revealed no significant difference between the groups which was considered as a healthy indicator for comparability of DEX between the groups as we mentioned before. In the study conducted by Taboada et al., lab parameters such as lymphocyte count, LDH, D-dimer, CRP, PCT, and Ferritin were compared between the group with response to DEX, HFNC, and PP therapy and the group with no response to therapy. Neither of these comparisons revealed a significant difference between the groups (15).
Our control group who were treated without DEX contained significantly more patients with DM, and CHF than the patient group (Table 2). In the similar study, no significant difference was found regarding co-existing conditions between the groups (15). HT-hyperlipidemia (15) and HT-COPD (17) were the two most common comorbidities in similar studies whereas HT and DM were the two most common co-existing condition in our study group. Predominance of the patients with DM and CHF in our control group might be caused by their fulminant course since these two co-existing conditions are known as risk factors for higher mortality (24) (25). Our hypothesis is that their more rapid progression to endotracheal intubation and death as soon as they were transferred to ICUs might prevent them from receiving DEX treatment. Congruently, significantly less requirement of the patients under regular insulin therapy for DEX infusion may also be explained by this hypothesis. Our second hypothesis for the patients under insulin treatment is that they might be more familiar to ICU setting and thus less agitated/anxious since these patients might be frequently admitted to ICUs because of diabetic ketoacidosis. In the similar study, no significant difference was found between successful and failed therapy groups regarding home treatments, however, they included only ACEIs, anticoagulants, corticosteroids, and statins in the analysis (15).
Among all the treatments applied to the patients with severe COVID-19, only IVIG and CPAP were significantly associated with DEX infusion (Table 3). 4 of 5 patients (80%) were treated with DEX infusion. Although IVIG group has a small number of patients the percentage is high. When we searched for the possible etiology of this situation we found out that headache, flushing, malaise, and pyretic reactions are common side effects of this therapy (26) (27). This side effects may easily cause delirium, anxiety, and agitation when occured. Our second hypothesis for these patients is that since IVIG is a very expensive treatment option, clinicians might have applied this valuable option to younger patients with less risk factors. Therefore these patients could have remained in NIMV support for longer and received DEX infusion. Since one of the most important indications of DEX infusion is maintaining cooperation with NIMV techniques, it was not surprising for us to see more patients receiving CPAP in the patient group. However, what is surprising was that not seeing more patients receiving HFNC in the patient group. Because the standard treatment algorithm for patients with severe respiratory failure was PP + HFNC with or without sedatives, if not responsive PP + CPAP, if not responsive endotracheal intubation respectively. Possible reasons for this condition might include the lack of equipment, presentation of the patient to the ICUs at the end stages of ARDS and thus directly applying CPAP or intubation by skipping the first step (HFNC), and the choice of clinicians as a result of their experience (found ineffective?). In a similar study in which all the patients received HFNC support, there was not any significant difference regarding hospital medical treatments between the successful and failed therapy groups (15). What should not be overlooked at this point is that patients receiving NIMV support constituted only 26 of the 45 patients treated with DEX which means 19 of the patients received DEX with the indication of either PP without NIMV support and/or delirium. This number was more than what we expected before the analysis.
Generally, the studies we mentioned above report that DEX can be a beneficial option in patients with severe COVID-19 especially to increase the patients’ tolerance to NIMV support and long periods of PP. However, one letter to the editor drew attention to DEX-associated hyperpyrexia as a side effect and dosage of infusion as a concern for the withdrawal phenomenon (20). The author recommends using DEX with caution because of this potential side effect and narrow therapeutic index. In our study, no apparent beneficial effect of DEX has been shown on mortality and intubation rates, however, there are limitations of our study. Performing in a single center, the missing data regarding withdrawal rates and causes of DEX infusion, being retrospective, and having a small sample size were the limitations of our study.