An elevated WBC count may be the result of various inflammatory responses, including both bacterial and viral infections (11). In the context of COVID-19, although the association between a high WBC count upon admission and the increased severity and mortality has been widely reported (12-14), only a limited number of articles provide information on the progression of WBC count during the stay in the ICU (15-22).

Even in early reports, the WBC count was shown to remain at high levels, peaking on day 10 of illness (15) or on day 7 following admission (16), while decreasing at the time of discharge only in less severe cases (17). In an early cohort study from Wuhan, China (18), the WBC count was shown to be significantly higher in non-survivors compared to survivors for 19 days following disease onset. On the other hand, in a study including 548 patients with COVID-19 from a national cohort in China, an increased neutrophil count exhibited an upward trend in non-survivors, whereas the neutrophil count was stable or exhibited a downward trend in survivors (19). In a large European-based cohort providing the almost real-time assessment of 639 patients critically ill due to COVID-19, the WBC count increased over time, peaking between days 2 and 3 following admission to the ICU (20). More specifically, in non-survivors, the WBC count was persistently higher during the first 7 days of ICU stay (20). Furthermore, in the multicenter study by Zanella et al (21), including 1,260 patients critically ill with COVID-19, the WBC count upon admission to the ICU was higher in non-survivors, whereas the neutrophil count was not. However, the trend in the WBC count during the stay in the ICU did not differ significantly between the survivors and non-survivors (21). By contrast, in a study from New York, USA identifying clinical markers that demonstrated a temporal progression associated with mortality, the WBC count was found to gradually increase prior to mortality (22).

Lymphocyte count. Absolute lymphopenia has been recognized as a common feature, particularly in patients with severe COVID-19, associated with an increased risk of developing adverse outcomes (5,21). Dynamic changes in the lymphocyte count over time have been reported in a limited number of studies, as indicated below.

In the large multicenter study by Xie et al (23), persistent lymphopenia was observed in both survivors and non-survivors during the first 14 days in the ICU, whereas in another study (24), time-dependent analysis revealed that the lymphocyte counts differed significantly between survivors and non-survivors, although no significant difference was observed over time. Persistent lymphopenia was also reported by Zhou et al (25) during the first 3 weeks following the onset of critical illness, and this was more profound in non-survivors. By the day of discharge, lymphocytes had returned to normal levels only in the less severe cases. Additional data have also revealed marked lymphopenia over time, which is more severe in non-survivors (17,18,26,27), suggesting that this variable may be used as a laboratory marker to distinguish patients with COVID-19 who are at a high risk of mortality at any time point during their clinical course. In another study by Chen et al (19), when addressing the dynamic changes in different biomarkers, the survivors were found to exhibit an increasing trend for lymphocytes (1.2-fold in midterm, and 1.4-fold at the end of hospitalization); in non-survivors, the lymphocytes remained at low levels without a noteworthy increase (19). Finally, in the large study by Zanella et al (21), daily values of lymphocytes were associated with survival; however, no significant difference was found in the progression of the lymphocyte count as a prognostic marker for outcomes.

NLR. The NLR, calculated by dividing the absolute neutrophil count by the absolute lymphocyte count, is considered to be linked to innate immunity, and serves as a useful biomarker for the evaluation of systemic inflammatory responses in patients with sepsis (28). An increased baseline NLR has been well-described in severe COVID-19 (18,29-33) and has also been found to be associated with poor outcomes (34,35). The evolution of NLR over the clinical course of severe COVID-19 has been analyzed in the studies discussed below.

In an early retrospective study from China (36), an increasing trend in the NLR in non-survivors, significantly differed compared with that in survivors. In addition, the peak value during hospitalization was higher in non-survivors, indicating a role of the NLR time course in the risk of mortality. Similarly, Zanella et al (21) reported an increased NLR in non-survivors at the time of admission to the ICU, as well as during the stay in the ICU. Both daily values and the slope of the ratio over time were associated with survival. In accordance, in the study by Wendel-Garcia et al (20), the NLR was persistently increased and was significantly higher in non-survivors in the ICU compared to survivors during the first 7 days in the ICU. A persistently high NLR was also observed in a small study, including 24 patients with COVID-19 in the ICU, peaking on day 10 of illness (15).

By using the maximum value of NLR during the first 3 days after being diagnosed with severe COVID-19, Ma et al (37) reported the ability of this biomarker to discriminate patients with moderate-severe acute respiratory distress syndrome eligible for veno-venous extracorporeal membrane oxygenation. Finally, in the study by Chen et al (19), NLR along with leukocytes, and other laboratory variables were maintained at significantly lower levels or exhibited a slight downward trend in survivors. By contrast, the NLR exhibited an upward trend or maintained higher levels in non-survivors.

Platelet count. Platelets are key regulators of thrombosis, inflammation and immunity and as such, they contribute to the pathophysiology of COVID-19 and the development of COVID-19-associated complications (38,39). Moreover, an abnormal platelet count, particularly thrombocytopenia, is relatively common in patients in the ICU (40). Temporal changes in the platelet count over the course of severe COVID-19 have been reported in a limited number of studies.

A previous study demonstrated that the trajectory of the platelet count during 28 days of hospitalization for patients with critical illness due to COVID-19 was persistently lower in critically ill patients and in non-survivors (39). Consistently, other studies have demonstrated that the platelet count is significantly decreased in non-survivors and the temporal changes in the levels of this marker differ markedly between survivors and non-survivors throughout the clinical course (26,41).

In a study on the coagulation profiles of critically ill patients due to COVID-19 stratified by a sequential organ failure assessment (SOFA) score >10 or ≤10(42), the platelet count increased from baseline to day 14 in both groups; a lower count was observed over time in patients with a SOFA score >10 compared to those with a score ≤10(42). The platelet count was increased in all patients, with ICU survivors presenting consistently higher counts during the first 7 days in the large cohort study by Wendel-Garcia et al (20). In another study, the median values of platelets were maintained within the normal range in survivors, significantly higher than those in non-survivors among 195 critically ill patients with COVID-19 over the course of days 1 to 28(25). However, the platelet count was not independently associated with mortality in the multivariable analysis. Similarly, in another study (18) the survivors exhibited a significantly higher level of platelets upon admission and an increasing trend during hospitalization, whereas non-survivors had a lower level of platelets upon admission, which further decreased afterwards.

D-dimer levels. D-dimer, a fibrin degradation product, represents a non-specific acute phase reactant, the levels of which may be elevated in acute inflammatory illnesses, as well as in venous thromboembolism. In the setting of COVID-19, elevated values of D-dimer are common, particularly in severe cases. However, the clinical relevance of elevated D-dimer levels is multifaceted. Increased D-dimer values may reflect disease activity and cannot be solely attributed to venous thromboembolic complications (43). Various studies have confirmed elevated baseline D-dimer levels as a predictor for both mortality and complications. However, the dynamic trend of D-dimer levels in patients with COVID-19 has been analyzed in a limited number of studies, as described below.

In a previous study, the D-dimer level, although not different upon admission, exhibited an increase over time in non-survivors in an early study from China (18), as well as in another study from Italy (44). Similarly, in another study significant dynamic changes in D-dimer levels over the first 14 days in the ICU were observed in non-survivors (23). In the large study by Wendel-Garcia et al (20), the D-dimer levels remained elevated throughout the first 7 days in the ICU in patients with unfavorable outcomes. In another small study (45) evaluating the coagulation function in 40 patients with COVID-19, the D-dimer values on days 5 and 10 were persistently high, although lower than those on the day of admission. Dynamic changes in the D-dimer levels in 577 patients admitted to the ICU due to COVID-19 were also observed in another study; these levels were higher and a higher increase rate was observed in non-survivors compared with survivors, indicating the utility of dynamic changes of D-dimer levels (46). In addition, in a study aiming to predict the development of venous thromboembolism in patients critically ill with COVID-19, the median D-dimer levels significantly increased on days 4 and 8 post-ICU admission in the patients who developed thromboembolism compared to those who did not (47).

A previous study demonstrated that D-dimer levels, measured every other day after admission, were persistently higher in mechanically ventilated patients with COVID-19 compared to those under non-invasive ventilation (16). Finally, another study also demonstrated the dynamic profile of coagulation parameters, tracked from days 1 to 14 after admission at 3-day intervals, in 183 patients with COVID-19, and revealed significantly higher D-dimer levels in non-survivors compared to survivors upon admission, as well as in late hospitalization (48). However, in another study (17), although higher D-dimer values upon ICU admission were observed in non-survivors compared to survivors, the change from admission to the 7th day did not differ between them. In accordance, in the large study by Zanella et al (21) evaluating the daily values and trends over time of relevant laboratory parameters in 1,260 patients in the ICU with COVID-19, the D-dimer values upon admission to the ICU were higher in non-survivors than in survivors; these values then decreased in both groups, but were not associated with survival in joint modeling. Similarly, elsewhere, although it was the single largest identifier of COVID-19 status, the D-dimer trajectory was not associated with outcomes (49).

Fibrinogen. Fibrinogen is a key coagulation factor in the common pathway of the coagulation cascade. It is an acute phase protein with a high molecular weight and is a known key regulator of inflammation (50). In the context of COVID-19, the plasma levels of fibrinogen are frequently increased and are associated with excessive inflammation, disease severity and ICU admission. To date, the temporal trends of fibrinogen levels over the course of the disease have only rarely been described in patients with COVID-19 in the ICU.

In the early study by Corrêa et al (42), reporting laboratory tests at baseline and on days 1, 3, 7 and 14 in patients stratified by a SOFA score >10 or ≤10, the highest fibrinogen values were observed on day 1 in the group with a SOFA score ≤10 and on day 3 in the group with a SOFA >10. On day 14, a more pronounced decrease in plasma fibrinogen levels was observed in patients with a SOFA score ≤10(42). In another study, the fibrinogen levels, although persistently high, did not differ between survivors and non-survivors both upon admission and on day 7(17). Consistently, individual trajectories for fibrinogen values did not exhibit any association with mortality in a study evaluating coagulation and endothelial function biomarkers in 14 patients with COVID-19(49). Finally, it was also previously demonstrated that the fibrinogen plasma levels were persistently high over a follow-up period of 20 days after ICU admission in patients with COVID-19 either with or without the diagnosis of venous thromboembolism (47).

LDH is a cytoplasmic enzyme which is crucial for cellular energy production. LDH plasma concentrations increase upon cell injury as a result of inflammation or hypoxia. In the context of COVID-19, elevated levels of LDH are considered to reflect lung and more widespread tissue damage. High levels of LDH upon admission have been used as a discriminatory laboratory marker for severity (5). The time course of LDH levels has been reported in a limited number of studies, as described below.

In an early report from Japan (15), LDH levels were elevated in the initial phase of illness and subsequently decreased in patients admitted to the ICU due to COVID-19. In the study by Chen et al (22), assessing the time course of various variables as a function of days to outcome, LDH levels exhibited a sharp change on the day of or a day prior to death, relatively to the values of survivors at the same time points. Persistently high LDH levels or increasing trends in LDH in non-survivors were reported in the large study by Wendel-Garcia et al (20), as well as in the study by Ouyang et al (26). Similarly, in another study by Xie et al (23), the LDH levels significantly decreased in survivors, but remained higher in non-survivors. Notably, in that study, the LDH levels, over time, exhibited significant differences between survivors and non-survivors (23). In addition, elsewhere, the time-dependent analysis of LDH levels reveled significant differences between survivors and non-survivors, as well as across time (24).

CRP. CRP, an acute phase serum protein, serves as a marker of the degree of inflammation. CRP has been identified as a key biomarker whose levels increase significantly in patients with severe COVID-19 and determines the progression of the disease (51-53). In a limited number of studies, the trend of CRP during the disease course has been reported, as described below.

In a small study from the first COVID-19 pandemic wave, CRP levels were elevated in the initial phase of illness and peaked on day 10 of illness (15). Consistently, longitudinal changes in CRP levels from the onset of symptoms exhibited similar trends in cases of either ICU admission or mortality (54).

In a European cohort of patients critically ill with COVID-19(20), CRP levels peaked on day 3 of stay in the ICU. In addition, ICU non-survivors were characterized by a significant increase in CRP levels, whereas the initial CRP levels did not differ between survivors and non-survivors. Similarly, in a multicenter study from North Italy (21) CRP levels at ICU admission were equally elevated in survivors and non-survivors, although continuously decreased over time in survivors. Τhe trends in CRP levels over the entire ICU stay were predictive of mortality (21). In another study, CRP levels decreased from admission to the 7th day of hospitalization in survivors, whereas an increase was observed in non-survivors (20). Moreover, in another study, the time dependent analysis of CRP levels did not reveal significant differences between survivors and non-survivors, although significant differences were observed over time (55); however, in another previous study, CRP values in non-survivors gradually increased with differences early or prior to mortality, relative to those of survivors (22). Finally, in the multicenter study by Wendel-Garcia et al (56), which provided the dynamics of the disease characteristics of patients critically ill with COVID-19 over the course of the pandemic, CRP levels during the first 5 days of ICU stay progressively displayed a more prominent decreasing trend over the duration of the pandemic. As regards the differences between survivors and non-survivors over time, CRP dynamics over the first days after ICU admission exhibited a more pronounced decline in non-survivors after the first pandemic wave, as compared to survivors (56).

It should be noted that dexamethasone and tocilizumab, both recommended for the treatment of severe COVID-19 since August, 2020(57) and January, 2021(58), respectively, are known to suppress the pro-inflammatory response, including CRP levels. Therefore, the more pronounced decrease in CRP levels during the stay in the ICU may reflect the systematic initiation of corticosteroids and/or tocilizumab. Zacharias et al (59) recently evaluated the effects of dexamethasone and tocilizumab on the trajectory of CRP levels among patients with critically ill COVID-19. Sequential CRP data were available in 174 patients receiving dexamethasone and in 40 patients receiving tocilizumab. Among the patients who received dexamethasone, CRP levels were significantly higher among the non-survivors. A significant reduction in CRP levels was observed in the first 3 days of treatment among the survivors and non-survivors, whereas over the subsequent week, the CRP levels increased among non-survivors, but not in survivors (59).

Serum albumin. Hypoalbuminemia upon hospital admission, possibly in the context of changes in vascular permeability leading to a greater capillary leakage of albumin (60), among other factors, such as the underlying nutritional status, nitrogen balance, or renal replacement therapy (61,62), has been associated with poor outcomes in patients COVID-19 (63-65). Data on changes in albumin levels over the course of severe COVID-19 are limited, as discussed below.

Albumin kinetics in patients critically ill with COVID-19 compared with patients critically ill with sepsis or other causes have recently been studied (66). Albumin levels were found to decrease rapidly following admission in patients with COVID-19 regardless of outcome, whereas the recovery in albumin levels predicted clinical improvement in this cohort. Notably, the decline, nadir and recovery of albumin levels in patients with COVID-19 were more pronounced compared with those patients with illnesses of non-COVID-19 etiology (66). In another study (26), in non-survivors, the serum albumin values were below the normal reference range and were significantly lower in non-survivors than in survivors, suggesting that a decline in liver synthesis function may be a key factor for COVID-19-associated mortality. However, elsewhere (20), albumin levels were found to decrease over time in all patients over the first 7 days.