Ultrasonographic Thrombosis Rates Associated with Midline Catheters in Patients with Advanced Chronic Kidney Disease: A Prospective Cohort Study

Abstract

Background: Preserving arm veins is important for arteriovenous fistula (AVF) creation in patients with advanced chronic kidney disease (CKD), as mature AVF is the preferred hemodialysis access. Midline catheters introduced into AVF candidate veins may cause thrombosis, hindering AVF formation. This study aims to determine ultrasonographic rates of midline-associated upper extremity deep venous thrombosis (UE-DVT) or superficial venous thrombosis (SVT) in patients with advanced CKD. Methods: We conducted a prospective study involving subjects with advanced CKD, who had a point-of-care ultrasound-guided midline placed in an arm vein. Within 35 days of midline insertion, participants underwent routine bilateral UE venous duplex ultrasound. The primary outcome was a composite occurrence of UE-DVT/SVT ipsilateral to the midline. Comparative analyses were performed based on patient demographics and device-specific variables. Results: 49 subjects with advanced CKD received midlines. The median midline catheter dwell time was <6 days for 15/49 patients (30.6%). The primary outcome occurred in 15/49 patients (30.6%), mostly asymptomatic thrombosis. No significant associations were found between outcomes and patient or device characteristics. Conclusions: Our study identified frequent use of midlines with short dwell times in subjects with advanced CKD which calls into question proper device selection. In this cohort, midline-associated arm clots were frequent.

1. Introduction

According to the 2025 United States Renal Data System annual report, the number of individuals with end-stage kidney disease (ESKD) in the United States reached an all-time high of 831,192 by the end of 2023. The rate of prevalent ESRD was 2327 per million people. Of all reported ESKD cases, approximately 57.1% were on in-center hemodialysis (HD) and 1.7% were on home HD [1]. In HD management for ESKD patients, various dialysis access modalities exist including arteriovenous fistula (AVF), arteriovenous graft (AVG), or central venous catheter (CVC). When compared to other dialysis access types, a mature AVF exhibits superior outcomes including decreased morbidity and mortality, reduced need for interventions, and enhanced long-term patency [2,3,4].

The National Kidney Foundation (NKF) Clinical Practice Guidelines recommend protecting both central and peripheral arteries and veins from damage in patients undergoing dialysis or those with CKD anticipating future dialysis [2]. Safeguarding these vessels in individuals with ESKD and severe CKD is crucial to prevent damage that might hinder AVF development. Despite guidelines advocating for reduced CVC use, approximately 85% of ESKD patients in the U.S. start dialysis utilizing either a tunneled or non-tunneled catheter, with approximately 24.1% still reliant on catheter access after 12 months [1].

Peripherally inserted central catheters (PICCs) and midline catheters are commonly used for venous access in clinical settings; however, they can pose risks to vein integrity due to heightened risk of thrombosis and stenosis. These catheters can cause endothelial injury, venous flow alterations and foreign body-induced hypercoagulability [5,6,7,8,9,10,11]. Concerns are rising about the impact of midline catheters on venous outcomes in advanced CKD; however, there is limited literature elucidating their thrombotic effects, especially in this population.

In 2012, the American Society of Nephrology’s Choosing Wisely campaign advised vein preservation in CKD stages 3–5 by discouraging the use of PICCs, midlines and subclavian vein catheters. It also recommended early nephrology consultation to aid in vein preservation, thereby enhancing the likelihood of establishing a functional AVF [12]. Subclavian catheters have become a rare event in dialysis patients. However, PICCs and midlines are still commonly used in advanced CKD.

Midline catheters dwell in the deep veins of the upper extremity and are peripherally inserted, but they are shorter than PICCs. Midlines are less invasive since they do not enter the central veins. Thus, they do not pose the risk of causing central venous stenosis or central-line associated bloodstream infections [10,13]. Guidelines from the Healthcare Infection Control Practices Advisory Committee recommend using a midline or PICC instead of a short peripheral intravenous (IV) catheter when the duration of IV therapy will likely exceed six days [13].

A recent meta-analysis compared outcomes of midlines vs. PICCs and showed fewer catheter-related bloodstream infections with midlines compared with PICCs. In regard to risk of thrombosis, similar rates of DVTs were observed between groups but higher rates of superficial venous thrombosis (SVTs) were seen in patients using midlines [14]. If an SVT occurs in a cephalic or basilic vein, it could hinder usage of that vein for AVF creation. It remains unclear whether midlines are safer in patients with advanced CKD who may need an AVF for dialysis due to its potential thrombotic risk.

2. Materials and Methods

2.1. Study Design, Setting and Participants

We conducted a prospective cohort study including patients with either CKD stage 5 or ESKD, which we will refer to as patients with advanced CKD, admitted to the University of Miami Hospital from January 2022 to December 2023. All patients had a point-of-care ultrasound (POCUS)-guided midline catheter insertion. The procedure was performed by a specialized vascular access team composed of a group of nurses with comprehensive vascular access training and at least ≥2 years of experience managing vascular access. Real-time POCUS guidance was used to map the arm veins often using a transverse/short-axis view, then an introducer needle was guided using the Modified Seldinger Technique to thread a guidewire. Then, the midline catheter was inserted. The arm veins routinely assessed during the initial POCUS for midline insertion were the basilic, brachial, cephalic, and axillary veins. The basilic vein was usually considered the preferred first-line vein for midline catheter placement because it is generally larger, straighter and farther from arteries and nerves; it was considered ideal when the vein-to-catheter ratio was acceptable (≤45%). When the basilic vein was unsuitable, the brachial and cephalic veins were generally considered second and third choice, respectively. The tip of the midline catheter generally ended in the mid axillary vein.

Patients who were found to have thrombosis in the cephalic, basilic, brachial or axillary veins during the initial POCUS did not receive a midline catheter on the thrombosed arm as this was considered a contraindication. The upper extremity POCUS-guided midline catheter insertion was generally unilateral. Only if a thrombosed arm vein was found in an upper extremity, then a contralateral upper extremity POCUS was performed to determine the patient’s candidacy to receive a midline on that arm. The decision to place a midline catheter was at the discretion of the clinical provider in charge of the patient during the inpatient hospital stay. The specific clinical indication for midline placement was not recorded. All the enrolled patients were prospectively followed and received a routine bilateral upper extremity venous mapping Duplex ultrasound within 35 days post midline catheter insertion, irrespective of thrombosis symptoms. The arm veins routinely assessed during the US Duplex venous mapping were the ulnar, radial, basilic, cephalic, brachial, axillary, subclavian and internal jugular veins. While POCUS is not equivalent to a Duplex venous ultrasound, it substantially lowers the likelihood of a pre-existing venous thrombosis in the cannulated veins. Therefore, we believe that it is reasonable to describe these thrombotic events as incident and temporally associated with midline catheter placement, but not definitely causal given the study design. We decided to use bilateral upper extremity Duplex venous mapping as the follow up ultrasound modality since this would allow us to screen for incident UE-DVT or SVT after midline catheter insertion while simultaneously providing valuable information on the quality of the arm veins (e.g., luminal size, distensibility with manual occlusion of blood flow) which could be helpful to determine what is the most appropriate vein to be used for future AV fistula creation. The examiners were not blinded to the midline catheter insertion side. This study was conducted in adherence with the Declaration of Helsinki and approved by the Institutional Review Board at the University of Miami Human Subject Research Office (IRB #20221045, 8 November 2022). All participants provided written informed consent prior to enrollment.

2.2. Inclusion Criteria

We prospectively enrolled adults (≥18 years) admitted during the study period to our institution, diagnosed with ICD-10 diagnosis of either CKD stage 5 or ESKD, who had a POCUS-guided midline catheter insertion in an upper extremity vein in the hospital (not present on admission). All patients were able to provide written consent to participate in the study.

2.3. Exclusion Criteria

We excluded subjects under 18 years, pregnant women or prisoners, with presence of thrombosis at the prospective midline catheter insertion site, history of vascular surgery ipsilateral to the prospective midline catheter insertion site, including arteriovenous fistulas in ESKD patients, history of mastectomy with lymph node removal ipsilateral to the prospective midline catheter insertion site, arteriovenous malformations ipsilateral to the prospective midline catheter insertion site, large amputations of the upper extremity that would prevented adequate midline catheter position, with presence of skin-related problems around the midline insertion site (infection, phlebitis, scars, etc.), or known or suspected allergy to materials contained in the midline catheter.

2.4. Variables

Patients were prospectively followed to assess the development of incident UE-DVT or SVT ipsilateral to the midline catheter insertion side. All patients received a routine bilateral upper extremity venous mapping Duplex ultrasound within 35 days post midline catheter insertion, irrespective of thrombosis symptoms. These ultrasounds were performed by a registered vascular technologist certified by the American Registry for Diagnostic Medical Sonography. University of Miami Hospital is a facility accredited by the Intersocietal Accreditation Commission for vascular testing. The diagnosis of thrombosis was established merely with ultrasonography if at least one of the following key criteria were met: (a) non-compressibility, the vein could not be fully flattened with transducer pressure; (b) thrombus visualization, direct visualization of a bright (echogenic) material within the vein lumen, appearing as a filling defect on color Doppler, (c) Doppler abnormalities, such as lack of normal blood flow signals (color Doppler) or increased velocity/abnormal waveforms (spectral Doppler). The diagnosis of UE-DVT was established when thrombosis was found in the deep veins (subclavian, axillary, brachial, radial, ulnar), whereas the diagnosis of UE-SVT was established when thrombosis was found in superficial veins (basilic, cephalic). In addition, we obtained patient’s demographic variables, type of dialysis access (if applicable) and pertinent clinical data that would pose an increased risk of thrombosis, such as active malignancy, history of venous thromboembolic events and recent surgery (last 30 days). We also obtained information related to midline characteristics, including vein cannulated, catheter size (using the French scale), number of lumens, dwell time. We also identified the level of care at the time of midline catheter insertion. Patients were followed prospectively under usual follow-up regimens for this population. Patients who did not undergo follow-up ultrasound were censored at the last available assessment. All data was captured into a protected RedCap database.

2.5. Outcomes

The primary outcome focused on a composite of incident ultrasonographic UE-DVT or SVT ipsilateral to the midline catheter insertion. Any DVT in axillary, subclavian or brachial veins; or any SVT in cephalic or basilic veins could significantly jeopardize vein quality and therefore decrease the likelihood of developing a functional AV fistula. An isolated DVT in the ulnar or radial veins does not typically reduce the chance of a successful AVF. Secondary outcomes included the individual occurrences of incident UE-DVT or SVT identified by ultrasound ipsilateral to the midline catheter. We also determined the rates of combined thrombotic events (DVT or SVT) in the arm contralateral to the midline insertion.

2.6. Statistical Methods

For the present study, we provide a descriptive analysis of the data points captured. All data was de-identified prior to analysis. All statistical analyses were conducted using IBM SPSS Statistics software version 28. Continuous variables were summarized using means and standard deviations, and categorical variables using counts and percentages. Finally, to understand factors that may lead to our outcome, we calculated odd ratios (OR) in a 2 × 2 fashion as “OR” = (a·d)/(b·c), and 95% CI computed using the Wald method on log(OR). Variables were analyzed as binary variables using prespecified clinical definitions: one vs. two catheter lumens; dwell time grouped as <6 days (reference), 6–14, 15–30, and >30 days; and prior midline/PICC, malignancy, and prior surgery recorded as Yes/No (with “prior surgery” defined as surgery within the 30 days preceding catheter placement). This study used a convenience sample including all eligible patients during the study period. No formal sample size calculation was performed, given the exploratory nature of the study.

3. Results

3.1. Participants/Descriptive Data

During the study period, a total of 49 patients with advanced CKD underwent POCUS-guided midline catheter placement. The median age was 65 years, with the majority between 65–74 years (42.9%), followed by 45–64 years (32.7%), while only 6.1% were aged 18–44 years and 18.4% were aged ≥75 years. Females represented 55.1% of the cohort, and 63.3% of patients identified as Black, while 36.7% were White. In terms of ethnicity, 38.8% were Hispanic or Latino, and 61.2% were not Hispanic or Latino.

Regarding body mass index (BMI), 40.8% of patients were obese (BMI ≥ 30 kg/m²), 24.5% were overweight (BMI 25–29.9 kg/m²), 30.6% had normal weight (BMI 18.5–24.9 kg/m²), and 4.1% were underweight (BMI < 18.5 kg/m²).

Concerning kidney disease classification, 46 patients (93.9%) had ESKD, while 3 (6.1%) had CKD stage 5 but were not yet on dialysis.

Among those with ESKD, 45 (91.8%) were receiving hemodialysis (HD) and 1 (2%) was on peritoneal dialysis (PD). With respect to vascular access, 33 patients (67.3%) had a hemodialysis catheter, 15 (30.6%) had an arteriovenous fistula (AVF) or graft (AVG), and 1 (2%) used a peritoneal dialysis catheter. Most patients had a significant comorbid burden. Hypertension was the most common comorbidity (98%), followed by type 2 diabetes mellitus (67.3%), heart failure (46.9%), and hypercholesterolemia (51%). Peripheral arterial disease and active malignancy were each present in 8.2% of patients, while HIV infection and hypercoagulable disorders were less common (each 6.1% and 2%, respectively).

Regarding procedural history, 25.6% had undergone prior surgery within 30 days, 14.3% had previous midline or PICC placement, and 63.3% had a history of prior central venous catheter (CVC) use. Therapeutic anticoagulation was reported in 24.5% of cases. Most midline catheters were placed in patients admitted to general medical units (71.4%), followed by intensive care units (22.4%), and emergency department settings (6.1%).

Most catheters had a thickness < 5 French (68.8%), while 31.2% were ≥5 French. Single-lumen midlines predominated (68.8%), with double-lumen catheters used in 31.2%.

Regarding dwell time, 30.6% of catheters were in place for <6 days, 36.7% for 6–14 days, 18.4% for 15–30 days, and 14.3% for >30 days. All this data can be summarized in

Table 1. Clinical characteristics of patients included in the study *.

Patients’ Characteristics	Total (N = 49)
Age (years)
18–44	3 (6.1%)
45–64	16 (32.7%)
65–74	21 (42.9%)
≥75	9 (18.4%)
Sex
Male	22 (44.9%)
Female	27 (55.1%)
Race
Black	31 (63.3%)
White	18 (36.7%)
Ethnicity
Hispanic or Latino	19 (38.8%)
Not Hispanic or Latino	30 (61.2%)
Body Mass Index [BMI] (kg/m2)
<18.5	2 (4.1%)
18.5–24.9	15(30.6%)
25–29.9	12(24.5%)
≥30	20(40.8%)
CKD Classification
CKD stage 5	3 (6.1%)
End stage kidney disease	46 (93.9%)
Dialysis Modality
Hemodialysis	45 (91.8%)
Peritoneal dialysis	1 (2%)
Type of dialysis access
AVF or AVG	15 (30.6%)
Hemodialysis catheter	33 (67.3%)
Peritoneal dialysis catheter	1 (2%)
Other Comorbidities
Hypertension	48 (98%)
Type 2 diabetes mellitus	33 (67.3%)
Heart failure	23 (46.9%)
Peripheral arterial disease	4 (8.2%)
Hypercholesterolemia	25 (51%)
HIV	3.0 (6.1%)
Active malignancy	4 (8.2%)
Hypercoagulability	1 (2%)
Others
Prior surgery (last 30 days)	14 (25.6%)
Previous midline/PICC	7 (14.3%)
Previous CVC	31 (63.3%)
Therapeutic anticoagulation	12 (24.5%)
Level of care
General medical unit	35 (71.4%)
ICU	11 (22.4%)
Emergency Room	3 (6.1%)
Vein cannulated
Cephalic	2 (4.2%)
Basilic	23 (47.9%)
Brachial	23 (47.9%)
Device Characteristics
Catheter thickness
<5 French	33 (68.8%)
≥5 French	15 (31.2%)
Number of lumens
Single	33 (68.8%)
Double	15 (31.2%)
Dwell Time (days)
<6 days	15 (30.6%)
6–14 days	18 (36.7%)
15–30 days	9 (18.4%)
>30 days	7 (14.3%)

* Values are expressed as counts (percentages), unless otherwise specified.

.

3.2. Outcomes

The primary outcome, a composite of incident ultrasonographic UE-DVT or UE-SVT ipsilateral to the midline catheter within 35 days of midline insertion, occurred in 15 of 49 (30.6%) patients, while only 2 (4.1%) patients experienced thrombosis-related symptoms. With respect to the anatomical location of the thrombosis events, all DVT or SVT occurred in a vein relevant for AVF maturation (basilic, cephalic, brachial, axillary, or subclavian veins). Only one patient demonstrated a radial vein DVT. In this case, the thrombotic process represented a continuous DVT extending from the subclavian vein through the axillary and brachial veins to the radial vein, rather than an isolated distal event. Regarding secondary outcomes, the individual occurrences of UE-DVT or UE-SVT as distinct thrombotic events, DVT occurred in 7 (14.2%) patients, and SVT occurred in 13 (26.5%) patients, both ipsilateral to the midline insertion site. A composite of UE-DVT or SVT contralateral to the midline insertion site occurred in 10 (20.4%) patients. A composite of bilateral UE-DVT or SVT occurred in 5 of 49 (10.2%) patients. However, no patient developed bilateral UE-DVT. In Supplemental Table S1 we provide an exploratory comparison between patients that had thrombosis vs. no thrombosis.

3.3. Predictors of Midline Thrombosis

Overall, no significant associations were found when comparing outcomes based on patient or device characteristics; these included comparing one catheter lumen versus two (OR = 0.56, 95% CI 0.16–2.04), dwell time 6–14 days compared to <6 days (OR = 0.33, 95% CI 0.07–1.47), dwell time 15–30 days compared to <6 days (OR = 0.33, 95% CI 0.05–2.12), dwell time > 30 days compared to <6 days (OR = 0.46, 95% CI 0.07–3.14), previous midline/PICC (OR = 0.33, 95% CI 0.04–3.05), history of malignancy (OR = 0.74, 95% CI 0.07–7.74), and prior surgery within the last 30 days (OR = 2.17, 95% CI 0.59–7.97;

Table 2. Odds ratios and 95% confidence intervals of midline catheter-related thrombosis based on select patient or device characteristics. This was calculated to assess the plausible risk of contributors to thrombotic events.

Comparison Variable	OR	95% CI
Number of lumens: One vs. Two (reference: Two)	0.56	0.16–2.04
Dwell time: 6–14 vs. <6 (reference: <6)	0.33	0.07–1.47
Dwell time: 15–30 vs. <6 (reference: <6)	0.33	0.05–2.12
Dwell time: >30 vs. <6 (reference: <6)	0.46	0.07–3.14
Previous Midline/PICC: Yes vs. No (reference: No)	0.33	0.04–3.05
Malignancy: Yes vs. No (reference: No)	0.74	0.07–7.74
Previous surgery: Yes vs. No (reference: No)	2.17	0.59–7.97

). In Supplemental Table S2, we provide counts of events for the 2 × 2 calculations.

4. Discussion

Preserving arm vessels emerges as a critical consideration for patients with advanced CKD, particularly those anticipating AVF formation for chronic hemodialysis access [15,16,17]. The increasing prevalence of ESKD underscores the necessity of a judicious approach to selecting dialysis access options. AVFs stand out as the preferred choice due to their demonstrated superiority in long-term outcomes compared to CVC, which carry elevated risks of morbidity and mortality. While PICCs and midline catheters fulfill diverse clinical needs, their thrombotic risks, especially in the context of advanced CKD, demand closer evaluation [13,18]. A 2023 multicenter cohort study of 21,415 midline recipients found that 5272 (24.6%) had advanced CKD (estimated glomerular filtration rate < 45 mL/min/1.73 m²). Major or minor complications occurred in 15.3% of advanced CKD patients versus 14.4% without CKD (adjusted OR 1.04; 95% CI 0.94–1.14), indicating no statistically significant difference. Among advanced CKD patients, catheter-related bloodstream infection occurred in 0.2% and upper extremity DVT in 1.2% [17]. These data suggest that midline complications are relatively uncommon even in advanced CKD but underscore the importance of preventing even low rates of thrombosis to preserve veins for future AVF creation. In addition, in this study upper extremity ultrasonography was performed only in subjects who had symptoms related to thrombosis. This approach differs from the one we used in our study since we proved that the majority of thrombotic events after midlines catheter insertion were asymptomatic.

The findings of our prospective cohort study shed light on the clinical implications of midline catheter usage in patients with advanced CKD, particularly concerning the risk of upper-extremity deep vein thrombosis or superficial vein thrombosis. Our investigation aimed to fill a notable gap in the literature by examining the incidence of thrombotic events associated with midline catheters in this specific population irrespective of thrombosis symptoms. Our patients had CKD stage 5 or ESKD along with multiple comorbidities which confer a high risk for frequent hospitalizations. It is possible that they had a previous midline or PICC inserted on a different facility, and we were unable to detect it. It is important to note that 15 out of 49 (30%) of the patients with advanced CKD who received a midline had a dwell time of less than 6 days. We did not collect aggregate information on the indications for midline insertion. At the University of Miami Hospital when a midline is requested to maintain routine venous access in subjects with advanced CKD, the local policy requires documentation from at least two different floor nurses about unsuccessful attempts to place a peripheral IV line before a midline can be ordered. However, placement of peripheral IV lines at this institution is performed blindly. This can be particularly challenging in patients who present with difficult IV access. The inability to obtain peripheral IV catheters could increase the demand for midlines. Growing evidence suggests that POCUS-guided insertion of peripheral IV catheters results in significant improvements in first-attempt success rates, reduction in procedural complications, and enhanced patient comfort [19]. The implementation of a POCUS-guided peripheral IV training program for nurses at our institution could be an opportunity for quality improvement; and this could possibly decrease the number of midlines requested. Moreover, 33 (67%) of patients were ESKD on hemodialysis through a dialysis catheter. These patients would greatly benefit from transitioning to an AVF as permanent type of dialysis access and regardless of their ESKD status and need for arm vessel preservation, they received a midline at the hospital.

Our study revealed that a considerable proportion of patients with advanced CKD experienced thrombotic events after midline catheter insertion, with a composite of ultrasonographic ipsilateral UE-DVT or SVT occurring in 30.6% of participants. Notably, most of these events were asymptomatic, underscoring the importance of routine surveillance in identifying thrombotic complications early, even without overt clinical symptoms. Moreover, our study showed a rate of UE-DVT or SVT in 10 out of 49 (20.4%) patients in the contralateral arm. Since there was only a single ultrasound performed in the contralateral arm, we could not determine the timing when those thrombotic events developed. But the UE-DVT or SVT events that occurred ipsilateral to the midline were determined to be incident events, since these thrombi were absent during the initial POCUS at the time of the midline insertion. The contralateral UE-DVT or SVT events were possibly a consequence of previous venous catheterization that we could not identify or alternatively could be related to patient’s prothrombotic status and comorbidities. Lisova et al. reported upper extremity thrombosis rates associated with a midline catheter of 4.5%, but the diagnosis of catheter-related thrombosis was based solely on symptomatic presentation followed by ultrasound [16]. Additionally, in a retrospective review including over 2500 catheters, midline catheters were associated with an increased risk of thrombosis compared to PICCs (7% vs. 4.7%) [20].

4.1. Clinical Trials and the Role of Midline Catheters

Several randomized trials and observational studies provide additional insights into midline performance relative to other devices. In the outpatient parenteral antimicrobial therapy (OPAT) setting, a randomized controlled trial reported that midlines were associated with lower rates of major complications (including catheter-related thrombosis and catheter failure) than PICCs when dwell time was ≤14 days; beyond 14 days, the risk of complications was comparable [21]. This suggests that midlines may be a safe alternative to PICCs for short term OPAT, provided infusion therapy is compatible. A 2025 randomized trial comparing midlines with PICCS in bariatric surgery patients found that midlines had lower catheter failure (10.7% vs. 20.5%), longer median dwell time (7 vs. 5 days), fewer complications (13.4% vs. 27.7%), and fewer additional device requirements [22]. These results support the use of midlines over PICCs in surgical patients requiring intermediate term access and highlight the need to distinguish between device types when interpreting outcomes.

4.2. Risk Factors for Thrombosis

Despite efforts to identify potential risk factors for thrombosis associated with midline catheter use, our study did not find significant associations between thrombotic outcomes and various patient or device characteristics, including midline dwell time, catheter size, number of lumens, or prior history of catheterization. This suggests a complex interplay of factors likely contributing to thrombosis in this population, warranting further investigation to identify underlying mechanisms and inform targeted preventive strategies.

In the literature, studies have aimed to address this. The diameter of the catheter relative to the vessel size is a key determinant of venous flow dynamics. Larger-diameter catheters occupying a greater proportion of the vein lumen can impair blood flow and promote stasis. In contrast, smaller caliber catheters permit more physiological flow around the device. In one study evaluating PICCs of different sizes, 5- and 6-French catheters were significantly more likely to develop DVT compared with 4-French catheters [23]. Interestingly, our results did not demonstrate a clear advantage of smaller size midlines with a single lumen over larger size midlines with multiple lumens in terms of thrombotic risk, contrary to previous suggestions in the literature. Similarly, the duration of midline dwell time (>30 days vs. <6 days) did not significantly influence thrombotic outcomes, indicating that thrombosis may occur irrespective of catheter duration. These findings challenge conventional assumptions about midline catheter-related thrombosis and underscore the need for consideration of individual patient factors in clinical decision-making regarding vascular access.

The risk of venous thrombosis also differs by insertion sites. PICCs have been associated with a higher risk of thrombosis compared with centrally inserted central venous catheters (CVCs). A meta-analysis including approximately 1000 patients in each group compared the incidence of venous thrombosis between PICCs and CVCs in intensive care and onco-hematology units. PICCs were associated with higher thrombosis rates both in ICU patients (OR 2.58, 95% CI 1.8–3.7) and in onco-hematologic populations (OR 2.91, 95% CI 2.1–4.0) [24]. Inadequate tip positioning is another important factor linked to thrombosis. A study reported that upper extremity DVT developed in 46% of patients with catheters positioned in the brachiocephalic vein, compared with only 6% when the tip was correctly located in the right atrium or superior vena cava [16]. Furthermore, the risk of PICC-related DVT varies depending on the vein accessed. The cephalic vein is typically narrower than the basilic and brachial veins in most individuals. Cannulation of the cephalic vein has been associated with a higher incidence of thrombosis compared to basilic or brachial vein access [25].

The observed high incidence of thrombotic events ipsilateral to the midline insertion site, particularly in veins regarded as the last resort (basilic and brachial) for AVF creation, raises concerns regarding the potential impact on future AVF development in this vulnerable patient population. Given the pivotal role of preserved venous integrity in facilitating successful AVF placement, strategies to minimize thrombotic complications associated with midline use are imperative to optimize vascular access outcomes in patients with advanced CKD. In 2018, the NKF published guidelines with recommendations for vessel preservation in individuals with CKD. These included favoring venipuncture in the back of the hand with a peripheral IV; avoiding PICCs for <7 days of infusion; considering femoral venous access for central vein access; PICC placement in CKD patients require approval by nephrology. No specific recommendations were made in reference to midline catheter insertions since it remains unclear what is the effect of midlines in future AVF creation [2]. Key aspects of care in patients with kidney disease that require a midline catheter are identified below:

• Vein preservation remains paramount. The high prevalence of central vein stenosis after PICC placement and the increased risk of thrombosis in advanced CKD underscore the need to avoid central venous devices whenever possible. KDOQI clinical practice guidelines advocate avoidance of PICCs and subclavian catheters in CKD stage 3–5 [2]. The new evidence reinforces that midlines, when used appropriately, may have comparable DVT risk but must still be selected judiciously.
• Catheter selection should consider diameter, site of insertion and tip location. Use of small diameter midlines, aiming for a low catheter-to-vein diameter ratio and an ideal catheter tip position (near the right atrium) could decrease thrombotic risk [16,23].
• Placement technique. Ultrasound-guided placement by specialized vascular access teams is likely to reduce complications [20]. Standardizing protocols and ensuring proper training may further minimize thrombotic risk.
• Infusion therapies must align with catheter capabilities. Emerging infusion standards now permit midlines for intermittent infusion of irritating medications, but there is no consensus on safe administration of high osmolar or extreme solutions [26].
• Individualized thromboprophylaxis. Whether pharmacologic thromboprophylaxis should accompany midline placement in high-risk populations is uncertain. In general, prophylactic anticoagulation is not routinely recommended for catheter-related thrombosis. Current practice for symptomatic CVC-related UE-DVT involves anticoagulation for a minimum duration of three months if the CVC is removed; and continued anticoagulation if the CVC remains in place over stopping after three months [27]. Data from the RIETE registry, one of the largest cohorts of catheter-related thrombosis patients, showed an annual recurrence rate of approximately 1.5% after discontinuation of anticoagulation. Extended anticoagulation beyond 3 months or for patients with transient risk factors significantly reduced the risk of thrombotic recurrence [28,29]. In CKD/ESKD patients, careful risk–benefit assessment is required due to concomitant bleeding risk.

4.3. Study Limitations and Future Directions

Despite its contributions, our study possesses certain limitations. The absence of significant associations between thrombotic risk and patient or device characteristics may be attributed to the relatively small sample size, potentially skewing the interpretation of results. A priori sample size calculation was not performed, and the study is underpowered to detect associations between patient and device characteristics and thrombosis. Furthermore, reliance on baseline data from clinical records may have led to missing information on relevant covariates, limiting the comprehensiveness of our analysis. In addition, vintage dialysis time for the patients enrolled in this study was not considered. It is plausible that some of these patients were not candidates for an AVF creation due to lack of optimal vessels; and therefore, were in need of a tunneled dialysis catheter for long-term hemodialysis. In these subjects, the use of midline catheters and its potential thrombotic risk would be a concern of lower priority. These limitations underscore the need for larger-scale studies with comprehensive data collection to validate and contextualize our findings effectively.

Future research directions include the development of thrombotic risk prediction models that incorporate patient comorbidities (e.g., CKD stage, diabetes, obesity), device characteristics (catheter size, lumens, tip position), infusion characteristics (osmolality, pH, vesicant status) and procedural factors (operator experience, ultrasound guidance). Machine-learning approaches might stratify patients by thrombosis risk and guide personalized device selection.

5. Conclusions

Our study contributes valuable insights into the incidence and clinical implications of midline catheter-associated thrombosis in patients with advanced CKD. By highlighting the substantial burden of asymptomatic thrombotic events and the lack of discernible risk factors, our findings underscore the importance of heightened vigilance and proactive surveillance in managing vascular access in this population. Future research endeavors should focus on nursing education regarding POCUS-guided placement of peripheral IV catheters ideally in the dorsal hand veins. A specialized nursing team with advanced skills in placing such catheters should be established.