The main goal of this study was to evaluate the prevalence of, the risk factors for, and the prognosis of DVT in patients with ARDS and to determine whether predictors of DVT differ between those with direct and indirect ARDS. We eventually enrolled 225 patients with ARDS in this study, 111 of whom had direct ARDS and 114, indirect ARDS. The prevalence of DVT on ultrasound scans in the overall group of patients with ARDS was as high as 40.0%, followed by an undifferentiated prevalence between the cohorts with direct and indirect ARDS (39.6% vs 40.4%; P = 0.913). Advanced age, serum creatinine level, and IMV were independently associated with DVT in the overall ARDS group as well as in the direct ARDS cohort. In the indirect ARDS cohort, however, increased DVT was only associated with advanced age and IMV. Patients with DVT had more adverse outcomes than those without
DVT, not only in the overall ARDS cohort but also in the direct and indirect ARDS groups. To our knowledge, this research is the first systematic description of DVT in patients with ARDS and of distinct associations among clinical characteristics and DVT in patients with direct and indirect ARDS.
In 2002, Greets et al. reported that the rates of objectively confirmed DVT in 4 prospective studies ranged from 13–31% [26]. In recent years, some research showed that, despite the use of guideline-recommended thromboprophylaxis, the incidence of DVT is still as high as 14–37.2% in critically ill patients [2, 3]. Zhang et al. pointed out that the cumulative incidence of VTE at 7, 14, 21, and 28 days was 4.45%, 7.14%, 7.53%, and 9.55%, respectively, in patients admitted to ICUs in China, even though the patients received guideline-recommended thromboprophylaxis [4]. Several factors probably account for the notably higher prevalence of DVT in our patients. First, most of the previously mentioned studies focused on critically ill patients who were in the ICU for other reasons. ARDS is a more serious type of critical illness that shows an overwhelming systemic inflammatory process accompanied by alveolar epithelial and vascular endothelial injury and an abnormal blood coagulation mechanism associated with significant death and may have a higher risk of DVT. Multiple studies have also suggested that the incidence of DVT in patients with ARDS from coronavirus disease 2019 or influenza A (H1N1) was as high as 42.2–85.4% [27–30]. These conditions indicate that direct ARDS itself may be a risk factor for DVT. As our results show, both direct and indirect ARDS had extremely high incidence rates of DVT. Second, some researchers defined VTE as a pulmonary embolism, proximal DVT, and/or symptomatic distal DVT, thereby excluding asymptomatic isolated distal DVT, which could probably be identified only by screening ultrasound [3]. Furthermore, the heterogeneous patient populations, such as those with different primary conditions, different numbers of days in the hospital, and different preventive measures, may represent a variety incidence.
Advanced age is a well-recognized risk factor for DVT in hospitalized patients, especially in critically ill patients, which has been included in a variety of thrombosis prevention scoring systems [22, 23]. As expected, in this study, the independent association of increased DVT with advanced age was found in both the direct and indirect ARDS cohorts. Interestingly, however, the contribution of advanced age to DVT differed in the different ARDS cohorts. The prevalence of DVT increased more significantly with advancing age in patients with direct ARDS than in those with indirect ARDS. The reason for this phenomenon may be partly, as previous studies have shown [14, 31] that, in our study, the patients in the indirect ARDS group also displayed more severe disease (higher APACHE II scores and higher SOFA scores) than those in the direct ARDS group, so the effect of advanced age on the overall condition of indirect ARDS was relatively small.
We found an independent association between serum creatinine levels and DVT in our patients. To our knowledge, however, this study is the first to assess differences in DVT related to renal function in ARDS by direct or indirect etiology. We found that the independent association between the serum creatinine level and the incidence of DVT in ARDS is modified by the underlying ARDS risk factors, with the protective effect on DVT of higher levels of serum creatinine being limited to patients with direct ARDS. However, we did not find a correlation between serum creatinine level and DVT in patients with indirect ARDS, which may be due to the more serious renal impairment and coagulation dysfunction in indirect ARDS, thus weakening the correlation between these two factors. Renal function was associated with dysregulation in coagulation in proportion to the severity of the renal impairment [32]. Some studies have demonstrated that chronic kidney disease and acute kidney injury (AKI) are independent risk factors for VTE [33, 34]. Al-Dorzi et al. pointed out that, for critically ill patients, neither AKI nor end-stage renal disease was an independent risk factor for VTE [35]. McMahon et al. found that AKI increases the risk for hospitalization-related VTE in a large, heterogeneous population that includes medical and surgical patients. However, this relationship was not seen in patients with traumatic injuries [36]. Some studies have shown that LWMH may have different levels of bioaccumulation in the case of renal insufficiency [37, 38]. The study by Cook et al. indicated that the incidence of DVT for patients with renal insufficiency in ICU who received dalteparin 5,000 IU once daily was 5.1% [39], which was far lower than that in the overall population of critically ill patients who received preventive treatment recommended by the guidelines [2–4]. So we speculat that the same dose of LWMH may play a stronger role in the prevention of DVT in the case of renal insufficiency. Unfortunately, due to the retrospective nature of the study, the decrease of LWMH metabolism in patients with AKI and higher level of serum creatinine was based on the conjecture of clinical data analysis, and we did not detect the activity of anti-factor Xa.
ARDS is a clinical syndrome with high mortality manifested by severe acute hypoxemia, which usually requires MV, especially IMV [8]. With IMV, sedation and immobilization are often performed simultaneously, which would aggravate blood stasis and increase the risk of DVT. Some studies have shown that IMV is a high-risk factor for DVT [28, 40]. Knudso et al. pointed out that IMV administered for more than 3 days is an independent risk factor for VTE [41]. As the duration of IMV increased, the risk of DVT increased [3]. Our research showed that both IMV and sedation were risk factors for DVT. Because all sedated patients in our study were treated with IMV, we only included IMV in the multivariate regression analysis. The results showed that IMV was an independent risk factor for DVT in both direct and indirect ARDS cohorts. However, in our study, compared with patients in the non-DVT group, the duration of IMV in the DVT group did not increase significantly, possibly because our small number of cases resulted in no statistically significant difference.
In direct and indirect ARDS cohorts, neither the APACHE II score nor the SOFA score was associated with the occurrence of DVT, presumably because the serum creatinine level, which was negatively correlated with the occurrence of DVT, was included in these two scoring systems [24, 25], thus weakening the correlation between severity scores and DVT.
Differences in predictors of DVT between direct and indirect ARDS partly support the growing body of literature suggesting that there are subphenotypes of ARDS that affect clinical outcomes [12–14, 16]. Our results suggest that subgroup analyses of ARDS are probably beneficial for stratifying and predicting the risk of DVT. We used age, IMV, and serum creatinine levels to predict DVT in the overall and the direct and indirect ARDS cohorts, respectively, and found that, in ARDS, the combined application of these indicators was not inferior to the current commonly used thrombus prediction scores, such as the Padua prediction score [23] and the Caprini score [22], for screening for DVT. Especially for direct ARDS, the combination of age, IMV, and serum creatinine level yielded a sensitivity of 81.8% and a specificity of 69.6% for scanning for DVT. A possible reason is that the Padua prediction score and the Caprini score apply to the general medical and surgical patients in the hospital. As a serious clinical pathophysiological syndrome with an overwhelming inflammatory response and coagulation abnormalities, ARDS has unique clinical characteristics and serious complications. The predictive value of the commonly used thrombus prediction method may be limited to screening for DVT in a patient with a critical illness such as ARDS.
Similar to the results of some previous studies [29, 42, 43], our results showed that DVT was associated with adverse outcomes in all the ARDS cohorts. Although there was no significant difference between length of stay in the hospital and length of stay in the ICU, Kaplan-Meier curves showed that the 28-day survival rate of patients with DVT was significantly lower than that of patients without DVT in all the ARDS cohorts.
Our study has some limitations. First, our sample size was small, which may underestimate the influence on DVT of factors such as obesity, being bedridden, and the insertion of a central venous catheter. Second, some patients had ultrasound scans only in the early stage of ARDS and did not have continuous dynamic monitoring, which may cause the incidence of DVT to be underestimated. Third, due to the critical condition of patients with ARDS, CTPA examinations were restricted. We performed CTPA examinations on only 3 patients with a high suspicion of PE and then confirmed the diagnosis of PE, which significantly underestimated the incidence of PE. Finally, this study is a retrospective study. We hope to conduct a prospective investigation to further clarify the incidence of DVT in patients with different subtypes of ARDS, to determine the corresponding risk factors, and to explore optimized individualized preventive measures in the case of ARDS to reduce DVT-related adverse prognoses.