Effect of Physical Intervention on Preventing Upper Limb Venous Thrombosis of Patients with PICC: A Systematic Review and Meta-Analysis of Randomized Trials

Objective: To evaluate the effect of physical methods on the prevention of venous thrombosis in patients with peripherally inserted central catheter (PICC). Methods: Randomized controlled trials meeting the inclusion and exclusion criteria were retrieved from the following databases: Cochrane li-brary, Pubmed, EMbase, Web of science, Ovid, CBM, CNKI, Wanfang, VIP, then Review Manage (RevMan) 5.3 software was used for data analysis. Results: A total of 24 RCTs including 3496 patients were analyzed in the study. The results of meta-analysis showed that various forms of upper limb movements could effectively decrease the incidence of venous thrombosis in the patients with PICC [RR = 0.23, 95% CI (0.16, 0.33), P < 0.01], upper limb movements could effectively improve the mean blood flow velocity of basilic vein [MD = 1.65, 95% CI (1.19, 2.11), P < 0.01]. Conclusion: Upper limb movements can effectively decrease the incidence of phlebothrombosis and improve the mean flow velocity of basilic vein in the patients undergoing PICC. Other forms of physical prevention methods need to be further studied due to insufficient sample size.


Introduction
Peripherally inserted central catheter (PICC) is a kind of central venous catheter punctured by peripheral veins, such as the basilic vein, cubital median vein, ce-ent modes and frequencies. Regular exercise: the nurse only carries on the health education of the physical exercise mode to the patient, but does not supervise the implementation. 5) Outcome indicators: the main outcome indicators in this study were the incidence of venous thrombosis and the mean velocity of blood flow. Venous thrombosis and venous velocity were determined by color Doppler examination. Normal vein ultrasound showed thin, smooth vein wall, no echo in the lumen, the probe pressure lumen can be compressed, the color blood flow filling good. In the case of venous thrombosis, the inner diameter of the lumen increases, the wall thickens, there are low or moderate echoes around the catheter, and color Doppler flow signals are filled with defective or no color flow signals at all. Venous velocity was compared with that of the patient at steady state.

Search Strategy
1) Searchable database: This study systematically searched relevant published original literatures in Chinese and English databases, including Pubmed, Web of sciene, EMbase, Ovid, Cochrane library, CBM, CNKI, Wanfang, VIP. The retrieval time is from their inception until August 2019. 2) Search terms: "arm exercise/make a fist/arm training/upper limb movement/limb function exercise", "PICC/peripherally inserted central catheter", "phlebothrombosis/venous thrombosis/thrombus of vein/venous embolism/deep vein thrombosis", "prevention and control".

Study Selection and Quality Evaluation
1) Study screening and data extraction: According to the consistent inclusion and exclusion criteria, the data was selected and extracted by two researchers independently, the two researchers have learned knowledge about evidence-based medicine, and received standardized training. Extract contents including basic information (author, year of publication, country, baseline situation, etc.), research design, sample size, interventions, outcome indicators, etc. Any disagreement was resolved by discussion with a third reviewer. 2) The two researchers independently evaluated the included studies according to Cochrane Assessment Manual 5.1.0 Assessment Criteria. The evaluation items included: random sequence generation, allocation hiding, blind method, data integrity of outcome indicators, selective reporting of findings and other sources of bias. A: fully meet the above criteria, the risk of a variety of bias is low; B: partially meet the above criteria, bias risk is moderate; C: completely dissatisfied with the above criteria, bias risk is higher. Both A and B studies were included in this study. rane collaboration network. Weighted mean difference (MD) was used for measurement data, such as the venous blood flow velocity. Relative Risk (RR) was used for enumeration data. 95% Confidence interval (CI) was calculated for all analyses. Firstly, Chi-square (X 2 ) was used to confirm whether there was heterogeneity among the studies. If there was no significant statistical heterogeneity (P > 0.1, I 2 < 50%), the fixed effect model was adopted. If there was heterogeneity (P < 0.1, I 2 ≥ 50%), the random effect model was used to combine with the effect size. Descriptive analysis was adopted if heterogeneity was too obvious and can't be judged.

Study Selection and Study Characteristics
A total of 579 trials were initially retrieved, 334 are Chinese and 245 are English. After screening, 24 literatures were finally included, and the specific steps of inclusion were shown in Figure 1. Three literatures [5] [6] [7] divided the intervention group into two groups (experimental group 1 and experimental group 2). The two groups had different exercise frequency, in order to make the data clearer and the consistency of intervention measures and outcome indicators, the two groups were divided into two RCTs. The basic features of the included articles are shown in Table 1.

Methodological Quality Evaluation of the Included Literature
The methodological quality of the included 24 trials was all of grade B. The baseline levels of the subjects were compared. Two of the studies [7] [8] made an intentionality analysis of the missing patients, and the specific evaluation indicators and results were shown in Table 2.         Figure 4. Descriptive analysis showed that all the 6 studies indicated that the velocity of axillary vein after exercise was faster than that of conventional exercise, and the difference was statistically significant (P < 0.05).

Effect of Bioelectrical Stimulation on Thrombosis 1 literature [23] evaluated the effect of electrical stimulation on vein of patients
with PICC catheterization. It was concluded that the difference of axillary vein blood flow velocity and the incidence of deep venous thrombosis between the experimental group and the control group was statistically significant.

Effect of Microwave on Vein Thrombosis in Patients with PICC Catheterization 2 literatures [24] [25] evaluated the effect of microwave on vein thrombosis in
patients with PICC catheterization. One literature [24] showed that the incidence of venous thrombosis in microwave group was lower than that in control group, and the difference was statistically significant (P < 0.05). One literature   [25] evaluated the effect of microwave on the diameter of basilic vein in patients with PICC catheterization. The result showed that the inner diameter of cardinal vein in microwave group was significantly dilated with higher success rate at a time than that of control group and the difference was statistically significant (P < 0.05). Due to the different outcome indexes of the two literatures, meta-integration could not be carried out, and only descriptive analysis was conducted.

Effect of Pressure on Venous Thrombosis in Patients with PICC
3 domestic studies [26] [27] [28] evaluated the effect of pressure on venous thrombosis in patients with PICC. Heterogeneity showed that fixed effect model should be selected to analyze (I 2 = 0%, P = 0.96). The results showed that the incidence of thrombosis in the pressure group was lower than that of the conventional exercise group, and the difference was statistically significant [RR = 0.22, 95% CI (0.09, 0.55), P < 0.01], as shown in Figure 5.

Sensitivity Analysis and Publication Bias
Funnel plot analysis was performed on 15 literatures [5] [9]- [22]. Through sensitivity analysis about the above combined results, the meta-analysis results had no change, suggested that the results of this study are stable. Scatter plots show a symmetrical distribution. The test results of 15 studies are symmetrical without publication bias, as shown in Figure 6.

Effect of Catheter Material on Venous Thrombosis in PICC Patients
In this paper, 14 [   20% (6/30) and 10.3% (7/68), which was the same as foreign brands (8% -50%) [32]. Meanwhile, Seckold, T. pointed out that there was no difference in the total incidence of complications of catheters in terms of materials, but the types of complications were different. The incidence of complications of PICC catheters of different brands is shown in Table 3.

Effect of Physical Exercise on the Incidence of Thrombosis in Patients with PICC
There are many ways to prevent thrombus in patients with PICC. The results of meta-analysis show that all kinds of upper limb movements can reduce the incidence of thrombus in patients with PICC and accelerate the blood flow speed of venous catheterization. Although there are different forms of upper limb movements, they can through the movement of upper limbs to accelerate the blood flow speed of the limb on the side of catheterization, promote the venous and lymphatic circulation, so as to reduce the incidence of thrombus. Some studies have confirmed that [33] [34], compliance of exercise is positively related to the incidence of thrombus in patients with PICC, patients with good compliance have better exercise effect than patients with regular exercise, and the incidence of thrombus is low. It may be due to the lack of supervision of the patients in the routine exercise group, few of them aware the importance of exercise, fear of the adverse consequences of the activity on the body, resulting in the failure of the exercise intensity and frequency to reach the standard, therefore, the blood flow of the limb on the catheter side is slow, which is easy to cause thrombosis. This study failed to confirm that bioelectric stimulation and microwave can reduce the incidence of venous thrombosis in patients with PICC, which may be related to the fact that this study is less involved in related studies, the implementation process of bioelectricity is relatively cumbersome, and the cost is high, the number of patients received in clinical is less or the intervention time is short, etc., and it is expected that large sample and high-quality randomized controlled trials to be further explored.
The relationship between pressure and venous thrombosis is uncertain. Domestic studies [26] [27] [28] have confirmed that intermittent pressure can reduce the incidence of venous thrombosis. This study found that the upper limb movements can effectively improve the average blood flow velocity of the vein and slow down the formation of thrombus. Therefore, effective education should be provided to patients with PICC, reasonable limb movements and physical therapy should be guided to reduce their fear of exercises and physical pain, and improve patients' compliance and initiative of exercise. However, the incidence of PICC catheterization in foreign countries is low, only one study included, so the sample size is small, and in the course of the study, the treatment group was forced to suspend the experiment due to the high incidence of thrombus, and the integrity of the outcome data was affected. Therefore, more high-quality RCTs and multilingual literatures are needed to evaluate the effect of intermittent pressure on thrombosis.
the comparing between domestic and foreign brands, to promote the development of intravenous therapy specialty in China.

Limitations of This Study
Most RCTs included are Chinese, only one [8] is English, and this study only retrieves the published Chinese and English trials, some trials may not be included. There are great differences in the sample and methodology of the included studies, and the intervention schemes are not consistent, which may have a certain impact on the results of the study. Some studies did not illustrate the generation of random sequence and blinding of allocation, which may also affect the results of Meta analysis. In addition, the quality of the studies included in this study is low, there is a certain risk of bias, and the effect of exercise needs to be further studied and practiced. It is suggested that future research should standardize the interventions. For example, the exercise of mode, method, frequency, intensity, measurement time and standard of outcome indicators, etc., should be refined and specific economic conditions and willingness of the patients should to be taken into consideration, so as to improve the consistency of the prevention. In the future, we can expand the sample size and carry out high-quality randomized controlled trials to validate the effect of standardized physical prevention on venous thrombosis of patients with PICC, and provide more comprehensive theoretical support for clinical nursing staff.