Peripherally Inserted Central Venous Catheter in Upper Extremities Could be a Risk for Deep Vein Thrombosis in Lower Extremities and D-Dimer in Neurology Department

Background: With the in-depth study of Peripherally Inserted Central venous Catheter (PICC) related venous thrombosis, it is found that the incidence of lower extremity deep venous thrombosis (LEDVT) in patients with PICC in upper extremities is higher than that in patients without PICC . However, there is no explanation for this clinical phenomenon that PICC related venous thrombosis seems to have exceeded the range of PICC travel . The purpose of this study is to elucidate this association between PICC in upper extremities and LEDVT by observing the changes of D-dimer . Methods: This was a retrospective cohort study of adults in Neurology department who underwent Color Doppler ultrasound and D-dimer test between 1 April 2017 to 1 April 2020. We analyzed the related factors of LEDVT and the change of D-dimer value, compared the changes of D-dimer before and after PICC insertion, and evaluated the predictive value of D-dimer in patients with and without PICC. Results: (cid:0) It was found that the presence of PICC increased the risk of lower extremity venous thrombosis by 7 times (OR = 7.048 [95% CI: 4.486-11.074]; (cid:0) It was found that the presence of PICC promoted the increase of D-dimer value (OR = 5.133 [95% CI: 3.072-8.575]). (cid:0) For patients without LEDVT, the level of D-dimer in patients with PICC was higher than that in patients without PICC (P < 0.05). (cid:0) The level of D-dimer after PICC insertion was signicantly higher than that before PICC insertion(P < 0.05). (cid:0) In patients with PICC, the AUC value of D-dimer in the diagnosis of LEDVT was 0.657 (95% CI: 0.549-0.765), and the negative predictive value was 82.35%. Whitney rank sum test, according to normality of their distribution. Risk factors for lower extremity deep vein thrombosis(LEDVT) and D-dimer group were studied by performing univariate and multivariate logistic. A Student’s t-test for association between D-dimer value in patients with PICC and those without PICC. A Student’s t-test for the changes of D-dimer value before and after PICC insertion. Risk assessments are presented as odds ratios (ORs) with 95% condence intervals (CIs). Receiver operating characteristic (ROC) analysis was used to plot the sensitivity and specicity of the with PICC related / non- PICC related D-dimer level respect to LEDVT. PICC related and non- PICC related D-dimer level as an independent biomarker for LEDVT was evaluated according to its area under the curve (AUC). The Youden’s index were determined to identify the optimal cut-off point and its sensitivity, specicity, negative predictive value, positive predictive value. All statistical analyses were conducted with SPSS (Version 18; SPSS, Central South University,Hunan, China), A P-value of < 0.05 (two-tailed) was considered to be signicant.


Introduction
Peripherally Inserted Central venous Catheter (PICC) is a catheter with tip located in superior vena cava or inferior vena cava through superior limb basilic vein, median cubital vein, cephalic vein, brachial vein, external jugular vein (newborns PICC insertion can also through great saphenous vein of lower limb, temporal vein of head, retroauricular vein, etc.). As a new infusion tool, PICC can reduce the pain of repeated puncture and reduce the incidence of drug extravasation, which is welcomed by clinical medical staff, patients and family members. However, PICC insertion and displacement will damage the vascular intima [2], The presence of PICC occupies part of the inner diameter of the vein [3,4], coupled with the patients' own diseases and medication induced hypercoagulable state [5], just meet the Virchow's triad [6]. The incidence of PICC related upper extremity deep venous thrombosis(UEDVT) was 51.4% − 71.9% [7][8][9]. Among them,patients who appear peripheral symptoms like limb swelling, local tenderness, skin temperature rise, skin color cyanosis, limb dysesthesia and dyskinesia or shoulder discomfort were 2.0% − 27.9% [10][11][12].
Previous studies have suggested that PICC related venous thrombosis is mainly mural thrombus, which is mainly limited to the venous route where the catheter is located [13][14][15]. With the deepening of the study, some data show that the incidence of lower extremity deep venous thrombosis (LEDVT) in patients with PICC is higher than that of patients without PICC, and PICC related venous thrombosis seems to have exceeded the range of the venous route where the catheter is located [2] [6,[16][17][18]. How to explain this clinical phenomenon? The purpose of this study is to elucidate this association by observing the changes of D-dimer. D-dimer is a speci c degradation product of cross-linked brin. The increase of D-dimer re ects the enhancement of coagulation and brinolysis system, and can be used as a sensitive index of hypercoagulability [14,19] The normal value of D-dimer can be used to assist in the exclusion of venous thrombosis [20] [21] [22] [23] [24]. Previous studies have shown that D-dimer can predict catheter-related venous thrombosis, and when venous thrombosis occurs, the D-dimer value can also increase [19,25]. It is not clear whether D-dimer is the medium or by-product of catheter-related venous thrombosis events [26]. Based on the above understanding, we conducted a cohort study to to elucidate the association between PICC in upper extremities and LEDVT by observing the changes of D-dimer in neurology department.

Patients And Methods
Patients and study design

Statistical analysis
This was a case cohort study, descriptive statistics were reported as means and standard deviations for continuous variables and frequencies for categorical variables. Proportions of events were compared by use of the Pearson χ2 test. Continuous variables were compared by use of Student's t-test or the Mann-Whitney rank sum test, according to normality of their distribution. Risk factors for lower extremity deep vein thrombosis(LEDVT) and D-dimer group were studied by performing univariate and multivariate logistic. A Student's t-test for association between D-dimer value in patients with PICC and those without PICC. A Student's t-test for the changes of D-dimer value before and after PICC insertion. Risk assessments are presented as odds ratios (ORs) with 95% con dence intervals (CIs). Receiver operating characteristic (ROC) analysis was used to plot the sensitivity and speci city of the with PICC related / non-PICC related D-dimer level respect to LEDVT. PICC related and non-PICC related D-dimer level as an independent biomarker for LEDVT was evaluated according to its area under the curve (AUC). The Youden's index were determined to identify the optimal cut-off point and its sensitivity, speci city, negative predictive value, positive predictive value. All statistical analyses were conducted with SPSS (Version 18; SPSS, Central South University,Hunan, China), A P-value of < 0.05 (two-tailed) was considered to be signi cant. In this study, 1360 patients with D-dimer value recorded within 7 days before the diagnosis of ultrasound, 597 patients (43.90%) had D-dimer > 0.5 mg/L.This includes 101 patients with PICC,80 patients(79.21%) had D-dimer > 0.5 mg/L. Univariate logistic regression analysis showed that age,male, malignant tumor, PICC, recent surgery and diabetes were associated with LEDVT and the presence of PICC promoted the increase of D-dimer value by 5 times(OR = 5.467 [95%CI: 3.338-8.956],P = 0.000);After adjusting gender, age and other factors, multivariate logistic regression analysis found that age,PICC, recent surgery and renal dysfunction were associated with the LEDVT and the presence of PICC promoted the increase of Ddimer value by 5 times (OR = 5.133 [95% CI: 3.072-8.575], P = 0.000). (Table 2) Table 2 Logistic regression analysis of related factors of D-dimer 3.The relationship between D-dimer value in patients with PICC and those without PICC For patients were diagnosed LEDVT, there was no signi cant difference in D-dimer between patients with PICC and patients without PICC; for patients were diagnosed not LEDVT, the level of D-dimer in patients with PICC were higher than that patients without PICC, the difference was statistically signi cant (P < 0.05). (Table 3)

4.Comparison of D-dimer before and after PICC insertion
The level of D-dimer after PICC insertion was signi cantly higher than that before PICC insertion (P < 0.05) (

Discussion
Foreign body in vascular system is the most important independent risk factor of upper extremities deep vein thrombosis (UEDVT) [27][28][29][30]. CVC increased the risk of UEDVT by 7 times (odd ratio [or] 7.3, 95% con dence interval [5.79, 9.21]; P < 0.0001) [29]. PICC is placed in a much smaller vein than in CVC, and the risk of DVT in PICC is 2.5 times higher than that in CVC [31], of which 33-60% were asymptomatic [32]. It is easy to understand that PICC is the most important risk factor for UEDVT [33,34], but the presence of PICC in this study increased the risk of LEDVT by 7 times (or = 7.048 [95% CI: 0.05] .It has been speculated that endothelial injury, vascular reactivity and coagulation promotion may lead to PICC related deep venous thrombosis, which extends beyond the vascular bed of PICC itself, but lacks systemic coagulation indicators to con rm this association.
D-dimer is a brin degradation product, which can be determined in the blood after the blood clot is degraded by the brinolysis process [35]. The formation and decomposition of blood clot is a dynamic process, and D-dimer can be elevated in various situations such as trauma and in ammation. Therefore, this study included the D-dimer values of these patients 7 days before the diagnosis of LEDVT after PICC insertion. It was found that the increase of D-dimer value was closely related to PICC insertion.In order to further understand the relationship between PICC and D-dimer value and LEDVT, we compared the Ddimer values of patients with PICC and patients without PICC under thrombus or non thrombus state, and found that the D-dimer value of patients with PICC in non thrombus group was higher than that in patients without PICC, and the difference was statistically signi cant. Considering that there are many and complex factors affecting the formation of LEDVT, it is di cult to judge whether the D-dimer values of patients with PICC is higher than that patients without PICC in the non thrombus group, because the PICC insertion population is more inclined to hypercoagulable patients or the in uence of PICC itself.
Therefore, in this study, D-dimer values within 7 days before and 7 days after PICC insertion were included in the analysis. It was found that the D-dimer value after PICC insertion was signi cantly higher than that before PICC insertion,which supported the view that PICC led to the increase of D-dimer.
Thrombosis is a natural process, which is activated by internal and external pathways. Both pathways contain a stimulating event, which starts a series of coagulation in the body and eventually forms brin rich thrombus. Endothelial injury is often an inciting event for our body's natural coalescence cascade to be activated. In order to maintain balance, the body starts appropriate brinolysis or blood clot formation destruction mechanism. DVT is formed when the mechanism that makes the body easy to coagulate is not hindered, or when the mechanism of decomposing blood clots is overloaded [36]. On the other hand, PICC occupies nearly half of the inner diameter of the vein, resulting in local blood ow slowing down, which can produce micro venous thrombosis, activate the coagulation system in the process of re ux, so as to produce a larger range of deep venous thrombosis [36]. We have reason to believe that the operation of PICC insertion itself or the existence of PICC leads to the general increase of D-dimer value.
Under the joint action of other factors, some patients eventually form deep venous thrombosis beyond the PICC vascular bed.
At the same time, we compared the predictive value of D-dimer in the diagnosis of LEDVT in patients with PICC and without PICC. The results showed that when the patients with PICC,the AUC value of D-dimer in the diagnosis of LEDVT was 0.657 (95% CI: 0.549-0.765), and the negative predictive value was 82.35% .when the patients without PICC, the AUC value of D-dimer in the diagnosis of LEDVT was 0.800 (95% CI: 0.769-0.830), and the negative predictive value was 93.25%. It is suggested that the D-dimer value of patients with PICC is not suitable to be used as a key index to exclude LEDVT.
Our study has important limitations. First of all, we did not conduct UEDVT screening, which may have missed a lot of asymptomatic UEDVT data. Therefore, we can not explain whether PICC as an important risk factor causes UEDVT rst and then the LEDVT, or that compared with the upper limb, the lower extremity venous valves are more and fragile, and the blood ow rate is slower, which only causes the LEDVT. Secondly, this study is limited to neurology patients, and there may be speci c bias of the disease itself. Finally, this is a retrospective study,the D-dimer values were collected within 7 days before and after PICC insertion,during this period, other events that could affect the increase of D-dimer were not excluded.
These limitations notwithstanding, our study has important strengths:1.The study shows that PICC in upper extremities could be a risk for LEDVT in Neurology department,suggesting that we should not only pay attention to the complications of PICC side limbs, but also pay attention to the compound risks after PICC insertion, especially the possibility of LEDVT. For patients with high risk of thrombosis, it is necessary for specialized nursing team to weigh the risks and bene ts of PICC, and to consider alternative vascular access schemes for high-risk patients; 2. For the rst time, the systemic coagulation index D-dimer was used to preliminarily explain the possible relationship between PICC in upper extremities and LEDVT. It can be considered try to use prophylactic anticoagulant regimen when patients at high risk of thrombosis need PICC insertion;3.It was found that the predictive value of D-dimer in patients with PICC is lower than that in patients without PICC,It is suggested D-dimer value is not suitable to rule out PICC associated LEDVT, so as to avoid missed diagnosis of PICC related LEDVT, which may cause adverse sequences or even life threading Area under ROC curve of D-dimer in the diagnosis of LEDVT in patients without PICC