Establishment and validation of a predictive model for peripherally inserted central catheter-related thrombosis in patients with liver cancer

Abstract

BACKGROUND

Peripherally inserted central catheters (PICCs) are commonly used in hospitalized patients with liver cancer for the administration of chemotherapy, nutrition, and other medications. However, PICC-related thrombosis is a serious complication that can lead to morbidity and mortality in this patient population. Several risk factors have been identified for the development of PICC-related thrombosis, including cancer type, stage, comorbidities, and catheter characteristics. Understanding these risk factors and developing a predictive model can help healthcare providers identify high-risk patients and implement preventive measures to reduce the incidence of thrombosis.

AIM

To analyze the influencing factors of PICC-related thrombosis in hospitalized patients with liver cancer, construct a predictive model, and validate it.

METHODS

Clinical data of hospitalized patients with liver cancer admitted from January 2020 to December 2023 were collected. Thirty-five cases of PICC-related thrombosis in hospitalized patients with liver cancer were collected, and 220 patients who underwent PICC placement during the same period but did not develop PICC-related thrombosis were randomly selected as controls. A total of 255 samples were collected and used as the training set, and 77 cases were collected as the validation set in a 7:3 ratio. General patient information, case data, catheterization data, coagulation indicators, and Autar Thrombosis Risk Assessment Scale scores were analyzed. Univariate and multivariate unconditional logistic regression analyses were performed on relevant factors, and the value of combined indicators in predicting PICC-related thrombosis in hospitalized patients with liver cancer was evaluated using receiver operating characteristic (ROC) curve analysis.

RESULTS

Univariate analysis showed statistically significant differences (P < 0.05) in age, sex, Karnofsky performance status score (KPS), bedridden time, activities of daily living impairment, parenteral nutrition, catheter duration, distant metastasis, and bone marrow suppression between the thrombosis group and the non-thrombosis group. Other aspects had no statistically significant differences (P > 0.05). Multivariate regression analysis showed that age ≥ 60 years, KPS score ≤ 50 points, parenteral nutrition, stage III to IV, distant metastasis, bone marrow suppression, and activities of daily living impairment were independent risk factors for PICC-related thrombosis in hospitalized patients with liver cancer (P < 0.05). Catheter duration of 1-6 months and catheter duration > 6 months were protective factors for PICC-related thrombosis (P < 0.05). The predictive model for PICC-related thrombosis was obtained as follows: P predictive probability = [exp (Logit P)]/[1 + exp (Logit P)], where Logit P = age × 1.907 + KPS score × 2.045 + parenteral nutrition × 9.467 + catheter duration × 0.506 + tumor-node-metastasis (TNM) staging × 2.844 + distant metastasis × 2.065 + bone marrow suppression × 2.082 + activities of daily living impairment × 13.926. ROC curve analysis showed an area under the curve (AUC) of 0.827 (95%CI: 0.724-0.929, P < 0.001), with a corresponding optimal cut-off value of 0.612, sensitivity of 0.755, and specificity of 0.857. Calibration curve analysis showed good consistency between the predicted occurrence of PICC-related thrombosis and actual occurrence (P > 0.05). ROC analysis showed AUCs of 0.888 and 0.729 for the training and validation sets, respectively.

CONCLUSION

Age, KPS score, parenteral nutrition, TNM staging, distant metastasis, bone marrow suppression, and activities of daily living impairment are independent risk factors for PICC-related thrombosis in hospitalized patients with liver cancer, while catheter duration is a protective factor for the disease. The predictive model has an AUC of 0.827, indicating high predictive accuracy and clinical value.

Key Words: Liver cancer; Peripherally inserted central catheters; Thrombosis; Model; Verify

Core Tip: In this study, we comprehensively analyzed the influencing factors of peripherally inserted central catheters-related thrombosis in hospitalized patients with liver cancer. We constructed a predictive model to accurately predict the risk of thrombosis and validated its performance. Age, Karnofsky performance score, parenteral nutrition, tumor-node-metastasis staging, distant metastasis, bone marrow suppression, and activities of daily living impairment were identified as independent risk factors, while catheter duration was a protective factor. The predictive model demonstrated high accuracy, with an area under the curve of 0.827, indicating its potential clinical value in enhancing the quality of care for peripherally inserted central catheters patients with liver cancer.

INTRODUCTION

According to global cancer statistics[1], as of 2020, approximately 906000 new cases of liver cancer were reported worldwide, resulting in 830000 deaths. Characterized by its propensity for metastasis and high invasiveness, liver cancer poses significant challenges in selecting treatment strategies[2]. Given the necessity for long-term treatment in liver cancer patients, venous catheterization often emerges as the preferred clinical approach to facilitate continuous drug infusion and chemotherapy. Peripherally inserted central catheter (PICC) is commonly employed for venous catheterization and involves the insertion of a catheter into the superior vena cava through a small incision in the elbow vein[3]. Catheter-related thrombosis, wherein thrombi form in the vicinity of the catheter, can lead to severe complications, such as pulmonary embolism and superior vena cava obstruction[4]. Previous studies indicated[5] that the incidence of PICC-related thrombosis ranges from 0.47% to 50.00%, with the majority being asymptomatic covert types, while only 2% to 26% of patients exhibit evident clinical symptoms. Consequently, accurately assessing the risk of PICC-related thrombosis in hospitalized patients with liver cancer holds particular significance.

Although some studies have delved into PICC-related thrombosis in liver cancer patients, an effective predictive model for the occurrence of this condition is currently lacking. Hence, this study endeavors to analyze the influencing factors of PICC-related thrombosis in hospitalized patients with liver cancer and to formulate a precise and dependable predictive model to assist clinicians in proactively identifying high-risk patients and implementing effective preventive measures in clinical practice.

MATERIALS AND METHODS

General information

We conducted a retrospective analysis of clinical data of liver cancer patients who received treatment at Chengdu Shangjinnanfu Hospital, West China Hospital of Sichuan University from January 2020 to December 2023.

The inclusion criteria encompassed patients hospitalized with liver cancer confirmed by pathology or imaging[2]; aged 18 years and above; undergoing PICC insertion for chemotherapy during hospitalization; and capable of furnishing complete clinical data. Exclusion criteria included pregnant and lactating women; patients with hematological diseases (such as hemophilia, thrombocytopenic purpura) or undergoing anticoagulant therapy; patients with a history of thrombosis; patients with an expected hospital stay less than 3 days; and patients unable to provide comprehensive clinical data or declining participation in the study. This study was approved by the Ethics Committee of West China Hospital of Sichuan University (No. 2023-1712), and the Ethics Committee agreed to waive informed consent.

Statistical analysis requirements stipulate that the sample size for variables in the risk factor study should be at least 5 to 10 times the number of variables. Approximately 22 predictive variables were identified based on literature review on factors influencing controlled randomized trial. These variables included age, sex, education level, hypertension, smoking, Karnofsky performance status (KPS) score, bedridden time, body mass index (BMI), impairment in activities of daily living, parenteral nutrition, vascular placement, arm, lumen, thrombosis history, PICC placement history, D-dimer, catheter duration, catheter-related complications, catheter diameter, tumor-node-metastasis (TNM) staging, distant metastasis, and bone marrow suppression. The sample size (n) ranged from approximately 110 to 220 cases. Thirty-five cases of PICC-related thrombosis in hospitalized patients with liver cancer were amassed, and 220 patients who underwent PICC placement during the same period but did not develop PICC-related thrombosis were randomly selected as controls. A total of 255 samples were amassed, with 255 cases utilized as the training set and 77 cases collected anew as the validation set in a 7:3 ratio.

Methods

Study tools: General patient information included sex, age, education level. PICC catheter information included placement history, placement arm, placement vein diameter, catheter duration, other related complications of PICC catheter. Patient disease-related data included number of comorbidities (e.g., hypertension), thrombosis history, smoking history, BMI, daily bedridden time (within the past week), hematological examination indicators (D-dimer). The Autar Thrombosis Risk Assessment Scale included seven modules such as age, physique, activity level. Each module was assigned a score of 0 to 7 based on specific risk factors. The cumulative score was used to assess the risk of thrombosis: Low (7–10 points), moderate (11–14 points), or high (≥ 15 points)[6]. PICC catheter-related thrombosis occurrence was evaluated by recording the B-ultrasound diagnosis results[7].

Data collection: Data were collected by the researchers themselves at Chengdu Shangjinnanfu Hospital, West China Hospital of Sichuan University PICC Outpatient Clinic. The process included filling out forms for general patient information, basic information of PICC placement, reviewing patient disease-related data, and using the Autar thrombosis assessment scale to assess thrombosis risk. Ultrasound Doppler equipment was used to conduct vascular ultrasound examinations on patients. During the examination, patients were exposed from the waist up and laid flat on the examination bed. The limb on the side of the catheter placement was externally rotated by 90° to fully expose the vein of the limb. The thrombosis diagnostic criteria included the following[8]: Dilation of the vein lumen, visible thrombus echo within the vein lumen, inability of the vessel to deform or be compressed when pressure is applied with the probe, and blood flow signal filling defect or bypass or absence.

Statistical analysis

SPSS 27.0 software was used for analysis. In the statistical analyses, normality was tested for continuous variables using the Shapiro-Wilk test. Quantitative data were expressed as (mean ± SD) and bilateral independent sample t test analysis. Categorical data were expressed as [n (%)] and compared using χ² tests. The receiver operating characteristic (ROC) curve was used to calculate the optimal cut-off value, sensitivity, and specificity. An area under the curve (AUC) > 0.9 indicates high predictive value, an AUC of 0.7-0.9 indicates moderate predictive value, and an AUC of 0.5-0.7 indicates low predictive value. A significance level of P < 0.05 was considered statistically significant. Variables with P < 0.05 were included in the binary logistic regression model to determine the independent risk factors for PICC-related thrombosis in hospitalized patients with liver cancer.

RESULTS

Incidence of PICC-related thrombosis in hospitalized patients with liver cancer

Among the 255 cases, 35 cases had PICC catheter-related thrombosis, with an incidence rate of 13.73%. Twenty-seven cases (77.14%) were asymptomatic thrombosis, and 8 cases (22.86%) were symptomatic thrombosis. Among the eight patients with symptomatic thrombosis, specific symptoms included pain, swelling, and numbness in the fingers, with 2 cases of pain, 5 cases of swelling, and 1 case of finger numbness.

Comparison of clinical data between thrombosis and non-thrombosis groups in the training set

Univariate analysis showed statistically significant differences (P < 0.05) in age, sex, KPS score, bedridden time, activities of daily living impairment, parenteral nutrition, catheter duration, distant metastasis, and bone marrow suppression between the thrombosis group and the non-thrombosis group. Statistically significant differences were not observed in other aspects (P > 0.05; Table 1).

Table 1 Single factor analysis, n (%).

Variable	n	Thrombus (n = 35)	Non-thrombosis (n = 220)	Statistical value	P value
Age (years)				χ2 = 4.270	0.039
< 60	150	15 (42.86)	135 (61.36)
≥ 60	105	20 (57.14)	85 (38.64)
Sex				χ2 = 4.785	0.029
Male	159	16 (45.71)	143 (65.00)
Female	96	19 (54.29)	77 (35.00)
Educational level				χ2 = 3.062	0.080
Primary and below	104	19 (54.29)	85 (38.64)
Junior high school and above	151	16 (45.71)	135 (61.36)
Hypertension				χ2 = 0.632	0.426
No	154	19 (54.29)	135 (61.36)
Yes	101	16 (45.71)	85 (38.64)
Smoking				χ2 = 0.520	0.471
No	146	22 (62.86)	124 (56.36)
Yes	109	13 (37.14)	96 (43.64)
KPS score (points)				χ2 = 5.022	0.025
≤ 50	186	31 (88.57)	155 (70.45)
> 50	69	4 (11.43)	65 (29.55)
Time in bed (hour/day)				χ2 = 4.935	0.026
≤ 12	146	14 (40.00)	132 (60.00)
> 12	109	21 (60.00)	88 (40.00)
BMI (kg/m2)				χ2 = 0.352	0.553
< 30	99	12 (34.29)	87 (39.55)
≥ 30	156	23 (65.71)	133 (60.45)
Disability of self-care ability				χ2 = 10.737	0.001
No	220	24 (68.57)	196 (89.09)
Yes	35	11 (31.43)	24 (10.91)
Parenteral nutrition				χ2 = 6.068	0.014
No	245	31 (88.57)	214 (97.27)
Yes	10	4 (11.43)	6 (2.73)
Vascularization				χ2 = 6.149	0.105
Venae magnalis	173	20 (57.14)	153 (69.55)
Brachial vein	54	7 (20.00)	47 (21.36)
Cephalic vein	16	5 (14.29)	11 (5.00)
Median vein	12	3 (8.57)	9 (4.09)
Arm				χ2 = 0.832	0.362
Left arm	142	17 (48.57)	125 (56.82)
Right arm	113	18 (51.43)	95 (43.18)
Lumen				χ2 = 0.811	0.368
Single lumen	250	35 (100.00)	215 (97.73)
Bicavate	5	0	5 (2.27)
History of thrombosis				χ2 = 0.483	0.487
No	252	35 (100.00)	217 (98.64)
Yes	3	0	3 (1.36)
PICC catheterization history				χ2 = 3.029	0.082
No	246	32 (91.43)	214 (97.27)
Yes	9	3 (8.57)	6 (2.73)
D-D (mg/L)				χ2 = 0.206	0.650
< 30	111	14 (40.00)	97 (44.09)
≥ 30	144	21 (60.00)	123 (55.91)
Tube time (month)				χ2 = 6.582	0.037
< 1	147	14 (40.00)	133 (60.45)
1-6	96	20 (57.14)	76 (34.55)
> 6	12	1 (2.86)	11 (5.00)
Catheter-related complications				χ2 = 0.209	0.648
No	133	17 (48.57)	116 (52.73)
Yes	122	18 (51.43)	104 (47.27)
Catheter diameter				χ2 = 0.647	0.421
4F	251	35 (100.00)	216 (98.18)
5F	4	0	4 (1.82)
TNM
I-II	133	11 (31.43)	122 (55.45)	χ2 = 6.985	0.008
III-IV	122	24 (68.57)	98 (44.55)
Distant metastasis				χ2 = 10.587	0.001
No	194	19 (54.29)	175 (79.55)
Yes	61	16 (45.71)	45 (20.45)
Myelosuppression				χ2 = 10.610	0.001
No	210	22 (62.86)	188 (85.45)
Yes	45	13 (37.14)	32 (14.55)

KPS: Karnofsky performance status; PICC: Peripherally inserted central catheter; TNM: Tumor-node-metastasis; BMI: Body mass index.

Multivariate regression analysis

Table 2 shows specific variable assignment details. Multivariate regression analysis showed that age ≥ 60 years, KPS score ≤ 50 points, parenteral nutrition, stage III to IV, distant metastasis, bone marrow suppression, and activities of daily living impairment were independent risk factors for PICC-related thrombosis in hospitalized patients with liver cancer (P < 0.05). Catheter duration was a protective factor for PICC-related thrombosis in hospitalized patients with liver cancer (P < 0.05; Table 3).

Table 2 Variable assignment.

Variable	Content	Assign
Y	Whether blood clots	No thrombus = 0, thrombus = 1
X1	Age	< 60 = 0, ≥ 60 = 1
X2	Sex	Male = 0, female = 1
X3	KPS score	> 50 = 0, ≤ 50 = 1
X4	Time in bed	≤ 12 hours/day = 0, > 12 hours/day = 1
X5	Parenteral nutrition	No = 0, yes = 1
X6	Tube time	< 1 = 0, 1-6 = 1, > 6 = 2
X7	TNM staging	I-II = 0, III-IV = 1
X8	Distant metastasis	No = 0, yes = 1
X9	Myelosuppression	No = 0, yes = 1
X10	Disability of self-care ability	No = 0, yes = 1

KPS: Karnofsky performance status; TNM: Body mass index.

Table 3 Multi-factor regression analysis.

Factor	B	SE	Wald χ2	P value	OR	95%CI
Constant	11.176	1.675	44.321	< 0.001	0.000	-
Age ≥ 60 years	0.644	0.308	4.342	0.036	1.907	1.038-3.491
KPS score ≤ 50 points	2.261	2.263	4.338	0.038	2.045	1.276-3.899
Parenteral nutrition	2.249	0.791	8.092	0.005	9.467	2.011-43.596
Duration of the tube is from 1 to 6 months	-0.681	0.321	4.556	0.032	0.506	0.271-0.955
Tube time > 6 months	-1.682	0.778	4.681	0.030	0.185	0.042-0.856
Phase III to IV	2.025	2.021	4.349	0.032	2.844	1.865-3.621
Distant metastasis	0.726	0.308	5.556	0.018	2.065	1.131-3.778
Myelosuppression	0.735	0.318	5.298	0.022	2.082	1.114-3.892
Disability of self-care ability	2.633	1.086	5.899	0.016	13.926	1.661-116.254

KPS: Karnofsky performance status; OR: Odds ratio.

Establishment and validation of the predictive model

Significant data from the baseline were included in the binary logistic regression analysis to obtain predictive model for PICC-related thrombosis in hospitalized patients with liver cancer: P predictive probability = [exp (Logit P)]/[1 + exp (Logit P)], where Logit P = age × 1.907 + KPS score × 2.045 + parenteral nutrition × 9.467 + catheter duration × 0.506 + TNM staging × 2.844 + distant metastasis × 2.065 + bone marrow suppression × 2.082 + activities of daily living impairment × 13.926. The relationship between the P predictive probability obtained from patients and thrombosis was used to construct the ROC curve, with an AUC of 0.827 (95%CI: 0.724–0.929, P < 0.001), corresponding to the optimal cut-off value of 0.612, sensitivity of 0.755, and specificity of 0.857 (Figure 1A). Calibration curve analysis showed good consistency between the predicted occurrence of PICC-related thrombosis and actual occurrence in the validation set (P > 0.05; Figure 1B). ROC analysis results showed AUCs of 0.888 and 0.729 for the training and validation sets, respectively (Table 4, Figure 2).

Figure 1 Receiver operating characteristic curve and decision curve. A: Receiver operating characteristic curve of peripherally inserted central catheter-associated thrombus predicted by combined indicators in inpatients with liver cancer; B: Decision curve of the prediction model.

Figure 2 Receiver operating characteristic curve and calibration curves for training set and verification set. A: Receiver operating characteristic (ROC) curve for training set; B: ROC for verification set; C: Calibration curves for training set; D: Calibration curves for and verification sets. AUC: Area under the curve.

Table 4 Value of prediction model for peripherally inserted central catheter-associated thrombosis in inpatients with hepatocellular carcinoma.

Index	AUC	Sensitivity (%)	Specificity (%)	95%CI	P value
Training set	0.888	90.76	87.77	0.851-0.965	< 0.001
Validation set	0.729	87.89	85.45	0.822-0.942	< 0.001

AUC: Area under the curve.

DISCUSSION

Among 255 cases, 35 had PICC catheter-related thrombosis, with an incidence rate of 13.73%. These findings are consistent with previous relevant reports[9]. This study found that symptomatic thrombosis in patients mainly presented as pain, swelling, and numbness in the fingers, with 2 cases of pain, 5 cases of swelling, and 1 case of finger numbness. In clinical practice, patient complaints should be highly regarded, especially for those who only present with symptoms of venous inflammation. Local treatment alone should not be applied, and the existence of thrombotic risk should not be overlooked. Instead, the cause should be promptly investigated, and a diagnosis should be made based on relevant examinations to ensure the safe use of catheters.

This study found that age ≥ 60 years is an independent risk factor for PICC-related thrombosis in hospitalized patients with liver cancer (P < 0.05). As age increases, endothelial cells release more procoagulant factors, so blood is more prone to coagulation; a hypercoagulable state is one of the risk factors for thrombosis formation. Shi et al[10] also showed that age > 60 years is a major risk factor for thrombosis formation in Chinese PICC chemotherapy patients. Therefore, the occurrence of thrombosis should be given attention in elderly patients after PICC catheter insertion. Patients should be taught to identify abnormal conditions in the limb with the catheter, seek medical attention promptly if abnormalities occur, be encouraged to drink water to reduce blood viscosity, and perform appropriate active and passive limb exercises daily to promote blood circulation and reduce the risk of thrombosis.

The KPS score is an important indicator used to assess the effectiveness of tumor treatment and the functional status of patients[11]. A higher score indicates better physical condition and functional ability. This study found that a KPS score ≤ 50 points is an independent risk factor for PICC-related thrombosis in hospitalized patients with liver cancer (P < 0.05). Patients with a KPS score below 50 have poorer physical condition and are more prone to muscle atrophy, metabolic abnormalities, and blood circulation disorders, leading to slower blood flow, endothelial cell damage, and increased platelet activity, making thrombosis more likely to occur[12]. Moreover, patients with low KPS scores may require prolonged use of PICC, thereby increasing the risk of thrombosis. Therefore, effective preventive and treatment measures should be taken to reduce the incidence of thrombosis in patients with KPS score ≤ 50 points.

This study found that parenteral nutrition is an independent risk factor for PICC-related thrombosis in hospitalized patients with liver cancer (P < 0.05). Parenteral nutrition is often used for tumor patients who cannot eat or have insufficient food intake and for those who experience adverse reactions such as vomiting, diarrhea, leading to malnutrition. These patients often have hypovolemia, and the infusion of high-concentration nutrients will further increase blood viscosity[13]. Increased blood viscosity slows blood flow and increases the risk of thrombosis formation. Therefore, close monitoring of patients' blood viscosity and thrombosis risk is required during parenteral nutrition therapy for tumor patients. Corresponding measures, such as appropriately diluting parenteral nutrition solution and implementing thrombosis prevention measures (such as anticoagulation therapy), should be implemented to reduce the risk of thrombosis-related events[14].

Stage III-IV liver cancer represents the advanced stage of liver cancer development and is associated with severe conditions and poor physical status; patients in this stage are susceptible to various factors, including the risk of thrombosis formation[15]. This study found that stage III-IV liver cancer is an independent risk factor for PICC-related thrombosis in hospitalized patients with liver cancer. Patients with stage III-IV liver cancer often have impaired liver function, poor blood circulation, abnormal platelet and coagulation factor levels, and liver cancer itself, which may affect the function of endothelial cells and increase the risk of thrombosis formation[16]. Additionally, patients with stage III-IV liver cancer typically require longer periods of PICC insertion, prolonging endothelial cell damage and thrombosis formation, which is also a significant contributing factor to thrombosis occurrence[17]. Therefore, enhanced thrombosis risk assessment and prevention measures, including anticoagulant therapy and regular PICC catheter replacement, are necessary for patients with stage III-IV liver cancer undergoing PICC treatment. Moreover, active disease control and improvement of patient physical status during liver cancer treatment are essential to reduce the risk of PICC-related thrombosis.

Distant metastasis is an independent risk factor for PICC-related thrombosis in hospitalized patients with liver cancer (P < 0.05). Patients with distant metastasis in liver cancer often have severe conditions, suboptimal health status, reduced immune function, and increased platelet counts, all of which are risk factors for thrombosis formation and increase the risk of thrombosis occurrence[18]. During PICC insertion, patients with distant metastasis in liver cancer often require longer catheterization procedures, leading to significant vascular damage. Additionally, the formation of biofilms on the inner wall of the PICC catheter and excessive coagulation and aggregation of blood in the catheter can increase the risk of thrombosis formation[19].

Bone marrow suppression is an independent risk factor for PICC-related thrombosis in hospitalized patients with liver cancer (P < 0.05). This condition refers to a decrease or inhibition of hematopoietic stem cell function in the bone marrow, leading to dysfunctions in the blood and immune systems[20]. In patients with liver cancer, bone marrow suppression may be caused by factors such as liver function failure and tumor infiltration into the bone marrow[21]. Bone marrow suppression leads to a decrease in platelet count and impaired coagulation function in patients with liver cancer, thereby increasing the risk of thrombosis formation[22].

Impaired activities of daily living are independent risk factors (P < 0.05). In patients with liver cancer, impaired activities of daily living may result from disease progression, liver function failure, adverse reactions to treatment, and other factors. Impaired activities of daily living may lead to prolonged bed rest and reduced mobility, thereby increasing the risk of blood stasis and thrombosis formation[23]. The characteristics of the PICC catheter itself may also increase the risk of thrombosis in patients with impaired activities of daily living. For example, the use of PICC catheters may limit the patient’s range of motion, slow blood circulation, and further increase the risk of thrombosis formation. Therefore, when treating hospitalized patients with liver cancer who have impaired activities of daily living with PICC catheters, necessary rehabilitation care and support should be provided as much as possible to promote patient mobility and reduce the risk of blood stasis and thrombosis formation.

The duration of catheterization is a protective factor for PICC-related thrombosis in hospitalized patients with liver cancer (P < 0.05). The occurrence of PICC-related thrombosis is mostly concentrated in the early stages of catheterization. After catheterization, patients undergo chemotherapy, during which a large number of tumor cells die, releasing a large amount of active substances, leading to the hypercoagulable state of the blood. Additionally, catheterization causes varying degrees of endothelial damage, which promotes platelet aggregation. Moreover, patients in the early stages of catheterization may reduce the activity of the cathetered limb due to lack of knowledge and fear of adverse effects on the catheter, thereby increasing the risk of thrombosis occurrence[24,25]. Therefore, after catheterization, patient education should be strengthened, appropriate levels and frequencies of activity should be advised to alleviate psychological pressure and reduce risk factors for thrombosis formation.

Based on the relationship between the predicted probability (P) obtained from patients and thrombosis, an ROC curve was constructed, with an AUC of 0.827 (95%CI: 0.724–0.929, P < 0.001), corresponding to an optimal cut-off value of 0.612, sensitivity of 0.755, and specificity of 0.857. Hence, the predictive model has high accuracy and reliability in identifying PICC-related thrombosis in hospitalized patients with liver cancer.

CONCLUSION

In summary, age, KPS score, parenteral nutrition, TNM stage, distant metastasis, bone marrow suppression, and impaired activities of daily living are independent risk factors for PICC-related thrombosis in hospitalized patients with liver cancer, while the duration of catheterization is a protective factor. Moreover, the predictive model has an AUC of 0.827, indicating high predictive accuracy and clinical value. Most thromboses in patients are asymptomatic, so patients should undergo regular vascular ultrasound examinations during catheterization, especially for patients with high-risk factors. Patient complaints should be given attention, and patient education should be strengthened. However, this study was only a cross-sectional survey; hence, further large-scale prospective studies are needed to clarify risk factors for prevention and early identification of PICC-related thrombosis.