Safety and risk factors of PICC insertion in hematologic malignancy patients with thrombocytopenia: a retrospective study

Abstract

Hematology patients often have severe thrombocytopenia and coagulation dysfunction. Despite the risk of bleeding, PICC insertion is still needed to reduce chemotherapy-related adverse reactions. Currently, there are no definitive indicators for predicting whether PICC placement increases bleeding risk in these patients. Additionally, large-scale studies on the safety of PICC insertion in hematology–oncology patients with platelet counts ≤ 50 × 10⁹/L and/or coagulation abnormalities are lacking. To evaluate the safety of PICC insertion in hematological oncology patients with a platelet concentration ≤ 50 × 10⁹/L and/or coagulation dysfunction through a large-sample retrospective study and to identify risk factors for bleeding after PICC placement. We retrospectively analyzed 1511 hematological patients with a platelet count ≤ 50 × 10⁹/L who underwent PICC insertion from February 2017 to January 2020, focusing on risk factors for bleeding within 24 h post-insertion. Multivariate analysis identified independent risk factors for bleeding within 24 h after PICC insertion as PT ≥ 13 s, platelet count ≤ 25 × 10⁹/L, and ADL score ≤ 95 points. Based on these three risk factors, patients were divided into low-risk (0–1 risk factors, 1% bleeding rate) and high-risk (≥ 2 risk factors, 7% bleeding rate) groups. When 0–1 risk factors are present, PICC insertion is safe even for patients with low platelet counts. If ≥ 2 risk factors exist, the decision to insert a PICC should consider the patient’s clinical symptoms and needs. Blood product intervention does not affect the bleeding rate.

Introduction

PICCs are utilized in a broad spectrum of applications, extending to clinical fields beyond the intensive care unit. The application scope of PICC has expanded to clinical fields beyond intensive care. It is simple to operate and has low risks [1–4]. Thrombocytopenia is common in critically ill patients, either already present at admission or acquired or worsened during the ICU stay. Currently, up to 67% of all platelets are used for managing patients with hematologic malignancies, while the remainder are used in cardiac surgery (7–10%) and intensive care (5–9%)^{5, 6}. One characteristic of PICC is its low risk of procedure-related trauma.

Patients with minor bleeding can achieve local hemostasis easily [7, 8]. Many studies have assessed the bleeding risk of central venous access device insertion in thrombocytopenic patients. In a prospective study of 105 patients, Ray and Shenoy found that despite an average platelet increase of only 11.5 × 10⁹/L after transfusion in patients with thrombocytopenia (50 × 10⁹/L), there were no significant bleeding complications requiring intervention [9]. The Society of Interventional Radiology (SIR) Practice Standards Committee recently released consensus guidelines stating that PICC placement is a low-bleeding-risk procedure that is easy to monitor and control [10]. Transfusion for any platelet count below 50 × 10⁹/L carries known risks and costs and imposes extra resources and blood safety demands on hospitals and treatment teams. Few studies have specifically evaluated the safety of PICC insertion in severely thrombocytopenic patients. In the study by Barrera et al., patients with an initial platelet count of less than 20 × 10⁹/L required platelet transfusion [11]. In 93% of these patients, the median number of platelet transfusions was 6 units. In a study by Potet et al., among cancer patients with severe thrombocytopenia, especially those with a platelet count less than 20 × 10⁹/L, the incidence of adverse events after PICC placement was low and limited to minor adverse.

hemorrhagic events [12]. Strahilevitz et al. reported on 40 patients with thrombocytopenia (platelet counts less than 10 × 10⁹/L) who underwent PICC insertion, finding that complications remained very low despite thrombocytopenia [13]. In 2019, the PICC team at the University of Maryland Medical Center (UMMC) lowered the prophylactic platelet count threshold for PICC insertion from 50,000/µL to 10,000/µL, which significantly reduced prophylactic platelet transfusions without a notable increase in PICC-insertion-related bleeding rates [14]. However, the above studies had small sample sizes and limited clinical guidance significance. Currently, large-scale studies systematically evaluating the safety of PICC insertion in hematology–oncology patients with platelet counts ≤ 50 × 10⁹/L and/or coagulation abnormalities are lacking, especially regarding key issues such as the incidence of adverse hemorrhagic events and transfusion requirements. This study aims to provide an evidence-based foundation for clinical practice through a retrospective analysis with a larger sample size.

Methods

Patients

We retrospectively collected data from 1511 patients with hematological malignancies and platelet counts < 50 × 10⁹/L who underwent PICC insertion using the ultrasound-guided modified Seldinger technique at the Department of Hematology, Peking University People’s Hospital, between February 2017 and January 2020. All patients provided informed consent before beginning treatment. This study protocol was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Peking University People’s Hospital.

PICC insertion

A standardized procedure combining ultrasound guidance, the modified Seldinger technique (MST) [15] and ZIM [16], along with a stratified hemostasis strategy, ensured safe PICC insertion for patients with hematological malignancies and thrombocytopenia (platelet count < 50 × 10⁹/L), with a significantly lower complication rate than traditional methods. Of the 1511 PICC insertions, 68 were silicone (4.5%) and 1443 were polyurethane (95.5%). The procedure was carried out by vein therapy specialists from the Chinese/Beijing Nursing Association [17]. Using ultrasound-guided MST, a 100% success rate was achieved (first-attempt puncture, single-attempt catheter placement). Among the 1,511 patients in this study, 1,011 underwent first-time PICC insertion, while 500 were post-transplant patients requiring re-catheterization, specifically intended for chemotherapy, and also for purposes such as fluid replacement therapy and poor vascular access. The catheter tip was positioned at the caval–atrial junction (CAJ) [18], as confirmed by X-ray [19]. Key data recorded included the puncture vein, date, catheter scale, tip position, and bilateral upper-arm circumference (compared with pre-insertion). Basic assessments included age, condition (hematological malignancy type and stage), allergy history, and vein therapy plans. Vein assessments evaluated skin integrity at the puncture site, vein elasticity, and diameter (with the basilic vein prioritized). Special assessments considered previous cannulation history, vein thrombosis history, coagulation function (INR, APTT), and platelet count (with a focus on bleeding risk for platelet counts < 50 × 10⁹/L). Under treatment-need satisfaction, the thinnest lumen catheter (e.g., 4Fr) was chosen to reduce thrombosis risk. The pre-puncture length was measured from the pre-puncture point (2 fingerbreadths below the elbow crease) along the vein to the right sternoclavicular joint and then down to the 3rd intercostal space. Using the ZIM-area puncture method, the area from the elbow to 21 cm under the axilla was divided into 3 segments, with the middle segment (green zone) selected as the optimal puncture point to lower phlebitis risk. Using a 21 G needle in the ZIM green zone, the success of vein puncture was confirmed via visualization technology. After the needle entered the blood vessel, the angle was reduced, the guidewire was inserted, and the expander and cannula were reassembled. Through the blunt scaling of the expander along with the cannula, they were slowly pushed into the blood vessel while advancing over the guidewire. After the expander was removed, continuous pressure was applied to the proximal end of the punctured vein (three-finger method), and the PICC was inserted into the cannula [20]. The gold standard for catheter-tip localization was post-insertion chest X-ray in the anteroposterior and lateral positions, confirming the catheter tip at the CAJ during maximal inspiration and breath-holding. An auxiliary technique, intracardiac electrogram (ECG), was used for real-time localization (optional) [21]. As malignant tumors can cause thrombocytopenia and coagulation dysfunction due to long-term chemotherapy, anticoagulation, and the disease itself, bleeding is a common issue in these patients [22]. The “sandwich” compression method involves covering the puncture site with alginate dressing, folding three layers of sterile gauze and fixing with transparent dressing, followed by two more layers of gauze and transparent dressing, and then applying moderate pressure with an elastic bandage (without hindering return flow). In our center, standard care after PICC insertion includes firm, manual compression of the puncture site for a full 20 min in every patient; no shorter or longer intervals were used. For platelet counts < 20 × 10⁹/L, heat application was delayed for 24 h to avoid early-stage bleeding. Heat application was started 2 h post-insertion above the puncture site (30 min/session, 4 sessions/day×3 days). Wet hot compresses after catheterization prevented phlebitis. Grip-strength exercises (30 repetitions/set, 4 sets/day) were performed to promote venous return. Quantitative grip strength exercises were used to reduce the incidence of PICC-related thrombosis and infection [23].

Data collection and analysis

Patient data collection included general information (sex, age, diagnosis, ADL at admission and discharge, ECOG score, BMI, height, weight, routine blood test results, coagulation tests), catheterization details (catheter type, puncture method and site, arm circumference, vein details (name, diameter, depth, blood flow), catheter length and external length, chest X-ray positioning, electrocardiogram localization, puncture attempts, first-time catheterization, and guidewire insertion attempts), thrombosis-related data (anticoagulant use, thrombosis history before and after catheterization, thrombosis occurrence time, site, and management), blood-transfusion details (platelet, plasma, prothrombin complex, and fibrinogen doses within 24 h before and after catheterization), and puncture-site bleeding (type, severity, exudation).

“Minor bleeding” was defined as petechiae, purpura < 1 inch, and exudation < 30 min. Purpura > 1 inch with exudation > 30 min was also categorized as minor bleeding. “Clinically significant bleeding” referred to limb-threatening deep-tissue bleeding, severe hemodynamic instability (systolic or diastolic blood pressure drop > 30% unresponsive to blood products), fatal bleeding, or limb swelling. Bleeding during the procedure, as described in the PICC insertion instructions, was classified as “immediate”. Bleeding within 24 h after PICC insertion was considered “within 24 hours post-procedure” ¹⁴.

Statistical analysis

Descriptive statistics were used to summarize covariates. Categorical covariates are reported as percentages and counts. Continuous variables are reported as medians and ranges. Pearson’s chi-square test was used to analyze categorical covariates. Student’s t (normal distribution) or Mann‒Whitney U (non-normal distribution) tests were used to compare continuous covariates between groups. Logistic regression models were applied for univariate and multivariate analyses to identify covariates associated with bleeding within 24 h after catheterization. The variance inflation factor (VIF) was estimated to check for multicollinearity among the covariates included in the Cox model. Covariates with P < 0.1 in the univariable analyses were included in the multivariable analyses. A 2-sided p < 0.05 was considered significant. SPSS 22.0 (SPSS, Chicago, IL) and GraphPad Prism 8 (GraphPad Software Inc., La Jolla, CA) were used for analysis and graphing.

Results

Patients

A retrospective analysis was conducted on data from 1511 patients with hematological malignancies and platelet counts ≤ 50 × 10⁹/L who underwent PICC insertion at Peking University People’s Hospital between February 2017 and January 2020. The cohort included 901 males and 610 females, with a mean age of 41.3 ± 17.69 years. Disease distribution was as follows: acute myeloid leukemia (AML) in 909 (60%), acute lymphoblastic leukemia (ALL) in 308 (20%), lymphoma in 65 (4%), multiple myeloma in 30 (2%), myelodysplastic syndrome in 89 (6%), aplastic anemia in 80 (5%), and other types in 30 (2%) (Table 1). A platelet count ≤ 25 × 10⁹/L was present in 52% (786) of the patients, and a platelet count between 25 × 10⁹/L and 50 × 10⁹/L was present in 48% (725) of the patients.

Table 1.

Patient characteristics

Factors	No. (%)
Gender, n (%):
Male	901 (60%)
Female	610 (40%)
Haematological disease
Acute myeloid leukaemia	909 (60%)
Acute lymphocytic leukaemia	308 (20%)
Lymphoma	65 (4%)
Multiple myeloma	30 (2%)
Myelodysplastic syndrome	89 (6%)
Aplastic anemia	80 (5%)
Other	30 (2%)
Total	1511

Identification of risk factors

Univariate analysis was performed on the clinical data of 1511 patients, including sex; age; pre- and post-admission ADL, BMI, puncture site (distance above the elbow in cm); arm circumference, pre-procedural WBC, Hb, PLT, PT, FIB, and APTT; and the use of anticoagulants, platelets, fibrinogen, prothrombin complex, and plasma (Table 2). We further compared 24-hour bleeding rates across subgroups: acute leukemia vs. non-acute leukemia (4% vs. 3%, P = 0.808) and first-time vs. non-first-time catheterization (3% vs. 4%, P = 0.521). Univariate analysis identified hypo-albuminaemia as the only factor significantly associated with bleeding within 24 h post-PICC (P = 0.001); multivariate analysis using logistic regression subsequently revealed three independent risk factors for bleeding within 24 h post-PICC insertion: PT ≥ 13 s, PLT ≤ 25 × 10⁹/L, and ADL ≤ 95 (Table 3).

Table 2.

Characteristics of patients with minor haemorrhage adverse events

Factors	Total N=1511	Bleeding group within 24 hours No. (%) N=55	Non-bleeding group within 24 hours No. (%) N=1456	P value
Gender, n (%):				0.537
Male	901 (60%)	35 (4%)	866 (96%)
Female	610 (40%)	20 (3%)	590 (97%)
Age, year median (range)	42 (2, 87)	39 (7-76)	42 (2, 87)	0.229
pre - admission ADL, median (range)	95 (10, 100)	90 (20, 100)	95 (10, 100)	0.001
BMI, median (range)	23 (10, 39)	22 (14, 29)	23 (10, 39)	0.118
Height, cm median (range)	168 (90, 193)	170 (132, 189)	168 (90, 193)	0.092
Weight, kg median (range)	64 (13, 124)	63 (34, 90)	64 (13, 124)	0.748
Puncture location (cm above the elbow), cm median (range)	9 (3, 24)	9 (6, 16)	9 (3, 24)	0.102
Arm circumference, cm median (range)	26 (12, 36)	25 (19, 30)	26 (12, 36)	0.167
WBC before PICC, × 109/L median (range)	3 (0, 395)	3 (0, 217)	3 (0, 395)	0.502
Hemoglobin before PICC, g/L median (range)	74 (35, 161)	65 (35, 189)	74 (36, 161)	0.001
Platelet before PICC, × 109/L median (range)	25 (1, 50)	14 (1, 50)	26 (1, 50)	＜0.0001
PT, s median (range)	13 (2, 39)	13 (11, 25)	13 (2, 39)	＜0.0001
APTT, s median (range)	30 (1, 68)	31 (22, 42)	30 (1, 68)	0.561
FIB, mg/dL median (range)	293 (11, 896)	281 (111, 705)	293 (11, 896)	0.528
D-D, ng/mL median (range)	314 (3, 252485)	508 (29, 38772)	311 (3, 252485)	0.617
Anticoagulant drugs before PICC, no (%)	22 (1%)	3 (5%)	19 (1%)	0.021
Thrombus events before PICC, no (%)	45 (3%)	2 (4%)	43 (3%)	0.771
Pre-PICC PLT infusion, no (%)	344 (23%)	20 (36%)	324 (22%)	0.016
Pre-PICC plasma infusion, no (%)	90 (6%)	7 (13%)	83 (6%)	0.036
Pre-PICC prothrombin complex concentrate infusion, no (%)	53 (4%)	5 (9%)	48 (3%)	0.029
Pre-PICC FIB infusion, no (%)	62 (4%)	3 (5%)	59 (4%)	0.659

Abbreviations: centimeters (cm), white blood cell (WBC), blasts in bone marrow (BM), Hemoglobin (Hb), Platelets (PLT), Prothrombin time (PT), International normalized ratio of prothrombin (INR), Fibrinogen (FIB), Activated partial thromboplastin time (APTT), D-Dimer (D-D)

Table 3.

Multivariate analysis

Outcome	Hazard ratio (95%Confidence interval)	P value
Bleeding group within 24 hours
pre - admission ADL ≤ 95	1.882 (1.065-3.327)	0.03
PLT ≤ 25×109/L	2.901 (1.558-5.399)	0.001
PT ≥ 13s	2.440 (1.384-4.304)	0.002

Abbreviations: Platelets (PLT), Prothrombin time (PT), International normalized ratio of prothrombin (INR), Fibrinogen (FIB), Activated partial thromboplastin time (APTT), D-Dimer (D-D)

Risk stratification

Based on the three risk factors, patients were stratified into low-risk (0–1 risk factors, 1% bleeding rate) and high-risk (≥ 2 risk factors, 7% bleeding rate) groups. The chi-square test revealed a significant difference between the two groups (P < 0.001).

Discussion

This study used a large retrospective sample to examine bleeding risk factors in hematological cancer patients with platelet counts ≤ 50 × 10⁹/L who underwent PICC insertion. Three risk factors for bleeding within 24 h post-insertion were identified: PT ≥ 13 s, platelet count ≤ 25 × 10⁹/L, and pre-admission ADL score ≤ 95. Risk stratification based on these factors revealed a 1% bleeding rate in low-risk patients (0–1 risk factors) and a 7% bleeding rate in high-risk patients (2–3 risk factors).

Our findings align with those of previous studies, which indicate that coagulopathy and thrombocytopenia are not absolute contraindications for PICC insertion. Instead [24], technical optimization (e.g., ultrasound guidance) and risk stratification are crucial. This study refined risk assessment by incorporating ADL and a multifactorial model that included pre-procedural blood count and coagulation test results, providing a basis for managing high-risk patients. Future research should verify the mechanistic link between ADLs and bleeding and evaluate the cost-effectiveness of stratification strategies for reducing blood product use.

This study revealed 3 risk factors for bleeding within 24 h after PICC insertion. Patients with 0–1 risk factors have a low bleeding rate, which is consistent with the literature. For example, Potet et al. (2013) and Amirahmadi et al. (2022) reported that even with platelet counts as low as 10,000/µL, PICC insertion caused only minimal oozing or small hematomas, with an incidence below 5%; thus, platelet transfusion is not routinely needed [14]. Strahilevitz et al. (2001) reported that in AML patients with platelet counts less than 50 × 10⁹/L, PICC-related complications were still controllable, supporting the safety of PICC insertion in low-risk patients [13]. This finding indicates that for low-risk patients, PICC insertion is safe and does not overcorrect coagulation parameters. In high-risk patients (≥ 2 risk factors), although the bleeding rate is higher, all events were minor, such as local oozing, which matches the “low-risk procedure” definition. Potet et al. (2013) reported that in patients with coagulopathy (INR ≥ 2 or on anti-platelet therapy), bleeding events were mainly minor oozing, with a 5.5% incidence and no major bleeding. Additionally, the literature stresses that ultrasound-guided techniques and standardized procedures can significantly reduce mechanical complications (e.g., arterial puncture), further supporting the safety of PICC insertion in high-risk patients. Therefore, as long as procedures are standardized, PICC insertion remains feasible for high-risk patients [12].

This study is in line with existing research, indicating that for low-risk patients with prolonged PT or thrombocytopenia, the bleeding risk after PICC insertion is low and coagulation correction is not needed. Using small-caliber catheters (4–5 Fr) with ultrasound guidance is key to reducing bleeding risk rather than just depending on blood transfusions. This finding highlights the importance of technical and procedural optimization for reducing bleeding risk and avoiding unnecessary blood product interventions. This study innovatively includes ADL in risk assessment. ADL-impaired patients have a higher local bleeding probability indirectly due to less activity, leading to slow venous blood flow or a higher risk of catheter displacement from improper care. Although the biological mechanism needs more research, this offers a novel risk assessment angle. Unlike previous studies that assess risks based on single indicators (e.g., platelet count < 50 × 10⁹/L), this study’s model with PT, platelet count, hemoglobin level, and ADL can better identify high-risk groups. For example, a patient with a platelet count of 30 × 10⁹/L and prolonged PT may have a higher bleeding risk than one with a platelet count of 10 × 10⁹/L alone. This approach overcomes single-indicator limitations and better reflects clinical complexity. This study also suggests differentiated bleeding management for high-risk patients. Despite the higher bleeding rate in these patients, all events were minor, such as local oozing. This matches the “low-risk procedure” definition. (2013) reported that even in patients with coagulopathy (INR ≥ 2 or on antiplatelet therapy), bleeding events were usually minor (5.5% incidence) without major bleeding observed. The literature highlights that ultrasound-guided techniques and standardized procedures can significantly reduce mechanical complications (e.g., arterial puncture), further supporting the safety of PICC insertion in high-risk patients. Therefore, as long as procedures are standardized, PICC insertion is feasible for high-risk patients. Furthermore, this study defines “high-risk.” Traditional “high-risk” often means a platelet count < 20 × 10⁹/L or INR > 2. However, our study defines it through multifactor analysis (e.g., thrombocytopenia + prolonged PT + abnormal ADL), which better reflects clinical reality. This stratification method improves the precision of bleeding risk assessment, offering a more scientific basis for clinical decisions.

This study aligns with existing research indicating that ultrasound-guided PICC insertion significantly reduces mechanical complications such as arterial puncture. In addition, standardized procedures, including careful puncture site selection, proper catheter fixation, and post-procedural care, are also crucial for minimizing bleeding events. Together with our findings, these observations underscore the importance of technical and procedural optimization in enhancing PICC-insertion safety. In clinical practice, this reduces patient transfusion dependence, decreases transfusion-related risks, and conserves medical resources. Multifactorial assessment can identify patients previously deemed at high risk for bleeding as potentially suitable candidates for PICC insertion. This not only broadens the applicability of the PICC but also offers patients safer and more effective treatment options.

Among the limitations, this was a retrospective study, and our research is prone to selection and information biases. Additionally, being conducted within a single center restricts the generalizability of the findings to other populations and health care settings. The studied patient cohort may also have unique characteristics that limit the representativeness of the results. Moreover, the study only assessed short-term outcomes, specifically bleeding events within 24 h post-PICC insertion, and did not include late-onset complications such as hematomas, infections, or thrombosis. Future studies should have longer follow-up periods to comprehensively evaluate the safety and efficacy of PICC insertion. Finally, the risk stratification model developed in this study requires external validation in different populations and health care settings, necessitating multicenter studies to confirm its reliability and applicability in diverse clinical contexts.

In summary, our study used multivariable stratification to identify PT, platelet count, and pre-admission ADL as independent risk factors for bleeding after PICC insertion. For high-risk patients, clinical decisions need to be made cautiously by considering the patients’ clinical manifestations, symptom severity, and treatment requirements. Moreover, our results indicate that there is no statistically significant link between bleeding risk and blood product intervention.

Author contributions

Author Contributions: All authors have read and agreed to the published version of the manuscript.Z.K.: designed the study, searched, collected, and analyzed the data, provided detailed data arrangement, and wrote the manuscript.H.H.: searched, collected, and analyzed the data, revised the manuscript.W.X.L: searched, collected, and analyzed the data.W.Y.:searched, collected, and analyzed the data.W.W.J: searched, collected.H.X.:searched, collected.L.J.:searched, collected. Y.X.: supervised the project and revised the manuscript. All authors had full access to all the data in the study, and the corresponding author had final responsibility for the decision to submit for publication. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.

Funding

This work was supported by the Peking University People’s Hospital Research and Development Funds (RDN2022-27).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Informed consent

Informed consent was obtained from all individual participants or their guardians included in the study.

Institutional review board statement

The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Institutional Review Board of Peking University People’s Hospital (approval number 2023PHD010-001) on 18 Sep 2023.

Competing interests

The authors declare no competing interests.