Macroscopic characteristics and handling aspects of peripheral and central venous catheters in patients treated in a German tertiary care hospital

Abstract

Background

In a cross-sectional analysis, we assessed the quality of the management of peripheral and central venous catheters (PVCs and CVCs) of inpatients at a tertiary care hospital in Germany, proportions of inflammatory signs, and catheter type-specific time periods of catheter application at the time of the detection of such inflammatory signs.

Methods

From 40 wards - excluding the department of psychiatry - data on the abundance, management and handling of PVCs and CVCs were collected and comparatively assessed.

Results

Among a total of 688 patients, 303 (44.0%) individuals had 327 PVCs (22× double applications, 1× triple application) and 80 (11.6%) had CVCs. For both types of venous catheters, signs of inflammation were in the ≤5% range. While these inflammatory signs were observed in PVCs being in place for 3.0 (±1.8) days, affected CVCs were in place for 6.8 (±3.5) days. Within both categories, however, duration of use was not an independent risk factor for inflammation, while CVCs were generally longer in use than PVCs.

Conclusions

Risk of inflammation was found to be similar for PVCs and CVCs at a German tertiary hospital, affected CVCs were longer in place.

1. Introduction

As summarized in a recent Cochrane review, critical infectious complications associated with peripheral venous catheters (PVCs) are rare events in the international scientific literature [1]. Routine replacement of such catheters has not been shown to provide a meaningful risk reduction compared with catheter exchange intervals guided by the clinical condition of the insertion site. This applies both to severe complications, such as catheter-related bloodstream infection, primary bloodstream infection, local infection and mortality, and to more common but less severe localized conditions, such as thrombophlebitis [1]. This holds true even although adherence to PVC infection prevention guidelines was shown to depend on factors like knowledge on this subject, training quality, degree of institutional support, as well as consistency of associated practices on site [2]. The localization of PVCs seems to be of minor relevance as well. When comparing PVC insertion sites at the forearm and the back of the hands, no differences regarding the observed complication rate could be observed [3].

Generally, short duration of PVC use is associated with reduced infection risk [4], but not always feasible in case of its requirement for therapeutic purposes. Although infection risk associated with PVC is lower compared with central intravenous devices, PVCs nevertheless account for about one third of catheter-related Staphylococcus aureus bacteremia cases [5]. This underscores the importance of prompt catheter removal in the absence of an ongoing clinical indication of its use, even if there are not any signs of infection [5]. Additional evidence-based factors associated with reduced complication rates include the use of catheters with smaller diameters, impregnation with antiseptic agents such as chlorhexidine [6], as well as comprehensive education and training of the nursing staff in safe catheter management [7].

Altogether, surveillance focused on PVC-associated complications is rarely conducted and prevention strategies in terms of aseptic intravenous catheter insertion and care are often inconsequently complied with [8, 9]. Available data suggest, not surprisingly, a quite frequent occurrence of PVC failure. In pediatric patients, a 38% failure rate was shown on device-level, affecting more than one third of hospitalized patients. Commonly observed kinds of PVC failure comprised infiltration, accidental removal, occlusion and leakage. While infiltration was reported in the 10% range, phlebitis as the most common infectious complication was seen in 5% of cases and accompanied by extravasation in 1% [10]. PVC-associated blood stream infection shows a prevalence between 0.1 and 0.2% [11, 12]. The most cases occur if PVCs are used over a prolongated time span, while short-term PVC application still accounts for 6.3% of nosocomial blood stream infections and 23.0% of nosocomial catheter-related blood stream infections. Consequently, a low threshold for obtaining blood cultures is critical for patients with indwelling PVCs who present local signs of inflammation and clinical features suggestive of systemic infection [11]. Preventive effects on particular types of PVC dressing or securement products, in contrast, have to be considered as uncertain due to insufficient available evidence [13]. However, early removal of non-indicated PVC is associated with a tendency towards reduced phlebitis rates [14]. Catheter infection risk is also influenced by patient factors. In burn patients, for example, insertion of the catheter close to or even within a burn wound increases the infection risk [15]. In preterm newborns, vancomycin showed some effectiveness in preventing catheter-related infections with a number-needed-to-prevent of 5 [16], but no general recommendation was proposed for this approach. Well trained staff is widely considered as a main preventive effector in terms of the prevention of PVC-associated infectious complications [17].

However, not only inappropriate handling can lead to PVC-associated complications, but vessel injuries like occlusion, infiltration or extravasation can also result from various administered drug. Among drugs used for infectious disease therapy, vessel injuries are associated with intravenous application of cefepime hydrochloride, ceftazidime pentahydrate, lincomycin hydrochloride, and flucloxacillin sodium, with the latter also causing vessel irritation [18].

Besides S. aureus [5, 19], other microorganisms commonly associated with catheter infections comprise coagulase-negative staphylococci, enterococci, Gram-negative rod-shaped bacteria and yeasts [20]. Notably, PVC-associated blood stream infections with S. aureus have infrequently been reported days after PVC removal, a medical condition known as “delayed onset S. aureus bacteremia” (d-SAB) [21]. Microbial colonization including biofilm formation is considered as a factor preceding catheter infection [22]. Consequently, catheter-related blood stream infections are usually refractory to antimicrobial therapy as long as the catheter is indwelling but respond to the treatment if the infectious focus is removed [23]. The hub, i.e., the connection between the catheter and the infusing tube, is considered as a major source of microbial entry [24]. However, while microbial entry via the catheter lumen predominates in longer indwelling catheters, microbial migration down the outer surfaces plays the dominant role in case of infections associated with short term catheterization over just a few days [25]. While strict asepsis during the insertion of the catheter will reduce the risk of the latter mode of infection, aseptic handling of the catheter's hub is the strategy of choice to prevent the intraluminal infection route [26]. Reflecting the mode of catheter contamination and catheter-associated bacteremia, specificity of bacteremia diagnosis by blood sampling through venous catheters is worse compared to traditional venipuncture, but sensitivity was found to be better [27].

The here presented cross-sectional study was designed to characterize PVC- and CVC-related inflammation and complications in a German tertiary care hospital. Specifically, the study aimed at providing real-world evidence on the quality of the management in German inpatient care, the frequency of inflammatory signs at catheter insertion sites, and catheter type-specific indwelling times at the time of their detection. We hypothesized that CVC management would be of higher quality than PVC management and that inflammatory signs associated with CVCs would occur after longer periods of catheterization.

2. Methods

2.1. Study design and observation strategy

The study was designed as a single-center, cross-sectional observational approach conducted over a two-day period across 40 wards of a German tertiary care hospital, including four intensive care units but excluding the department of psychiatry. Data collected included the number and indwelling time of PVCs and CVCs, catheter insertion sites and associated signs of inflammation, as well as the integrity of the catheters and their dressings. Assessed inflammatory signs at catheter insertion sites comprised erythema, increased local temperature, pain, swelling and functional impairment. Patient-specific data, such as age or sex, were not recorded. Incomplete datasets were not excluded from the analysis. Only patients who were present and available for assessment during the observation period were included.

2.2. Statistics

Odds ratios between categorial variables were calculated applying Fisher's exact test. Wilcoxon-Mann-Whitney testing was applied to assess numerical variables. Significance was accepted at P < 0.05, correction for multiple testing was not applied [28]. The calculations were performed using the software GraphPad InStat version 3.06 (GraphPad Software Inc., La Jolla, CA, USA).

2.3. Ethics

Ethical clearance for the study was obtained from the ethics committee of the University Medicine Rostock (reference number A 2500-0200, granted at 22 August 2025) without requirement of informed consent. All assessments were conducted in line with the Declaration of Helsinki and all its amendments as well as in line with German law.

3. Results

3.1. Characteristics of the devices of the study participants

From a total of 688 patients assessed on the chosen hospital wards, a total of 378 (54.9%) had either a PVC or a CVC or both. Stratified by the kind of device, 303 (44.0%) individuals were fitted with at least one PVC and 80 (11.6%) with a CVC. Notably, 22 patients had 2 PVCs and 1 patient even three of them, resulting in a total of 327 assessed PVCs. Predominant application sites for PVCs were elbows (n/N = 142/327 (43.4%)) and hands (n/N = 108/327 (33.0%)), next to others (n/N = 77/327 (23.5%)). CVCs, in contrast, were predominantly applied at jugular veins (71/80, 88.8%), followed by subclavian veins (5/80, 6.3%), and femoral vein (4/80, 5.0%), respectively. While devices being in place for up-to five days at the time of the cross-section assessment were mainly PVCs, CVCs showed higher proportions of devices being in place for longer durations at this time point, a phenomenon which was found to be significant for the first 15 days after application (Table 1). Insofar, the principal finding was the longer indwelling time of CVCs compared with PVCs.

Table 1.

Duration of PVC and CVC application. Significant associations are indicated in bold type

Duration of application in days	Number of affected PVCs	Number of affected CVCs	Odds ratio, significance
<5	199/267 (74.5%)	28/80 (35.0%)	5.435 (3.181, 9.287), P < 0.0001
5–10	55/267 (20.6%)	35/80 (43.8%)	0.3336 (0.1959, 0.5679), P < 0.0001
11–15	8/267 (3.0%)	11/80 (13.8%)	0.1938 (0.07502, 0.5004), P = 0.0008
16–20	2/267 (0.7%)	3/80 (3.8%)	0.1937 (0.03178, 1.181), P = 0.0826
21–30	1/267 (0.4%)	2/80 (2.5%)	0.1466 (0.01311, 1.639), P = 0.1341
31–40	1/267 (0.4%)	1/80 (1.3%)	0.2970 (0.01835, 4.806), P = 0.4085
41–50	1/267 (0.4%)	0/80 (0%)	0.9062 (0.03653, 22.479), P = 1.0000

PVC = peripheral venous catheter. CVC = central venous catheter.

3.2. Properties of insertion sites in comparison of PVCs and CVCs

The vast majority of observed insertion sites was inconspicuous, while minorities showed either traces of blood or incrustation. Incrustation was significantly more frequent at insertion sites of CVCs compared to PVCs. Other significant associations were not observed. Details are provided in Table 2. The main finding was the higher proportion of incrustation at the insertion sites of the CVCs.

Table 2.

Properties of insertion sites in comparison of PVCs and CVCs. Significant associations are indicated in bold type

Insertion site venous catheter	PVC	Used as	First CVC	Odds ratio, significance
Inconspicuous*	282/327 (86.2%)	Ref.	64/80 (80.0%)	1.567 (0.8328, 2.947), P = 0.1650
Bloody	22/327 (6.7%)	Ref.	4/80 (5.0%)	1.370 (0.4585, 4.096), P = 0.7988
Incrusted	23/327 (7.0%)	Ref.	12/80 (15.0%)	0.4287 (0.2033, 0.9040), P = 0.0420

Ref. = reference.

3.3. Properties of devices in comparison of PVCs and CVCs

Properties of PVCs and CVCs as recorded during the assessment are summarized in Table 3. Digital documentation rates ranged from 80% to more than 95%, including the documentation of catheter-related indications. Transparent dressings of standard dressing type were used in the vast majority of cases. A minority of devices showed improper dressing or even signs of inflammation. Among the latter, erythema and swelling were most frequently recorded. However, device associated complications suggestive of inflammation were too rarely seen to show statistically significant differences.

Table 3.

Comparative analysis of properties of PVCs and CVCs. Significant associations are indicated in bold type

Feature	PVC	Used as	CVC	Odds ratio, significance
Image documentation	296/319 (92.8%)	Ref.	73/80 (91.3%)	1.234 (0.5097, 2.988), P = 0.6373
Transparent dressing	286/327 (87.5%)	Ref.	66/80 (82.5%)	1.480 (0.7623, 2.872), P = 0.2733
Digital documentation “CheckPad”	230/283 (81.3%)	Ref.	78/80 (97.5%)	0.1113 (0.02649, 0.4675), P < 0.0001
Extension set used	157/327 (48.0%)	Ref.	n.a.	n.e.
Sealing cone missing	5/327 (1.0%)	Ref.	1/80 (1.3%)	1.227 (0.1412, 10.654), P = 1.0000
Signs of inflammation	14/307 (4.6%)	Ref.	4/80 (5.0%)	0.9078 (0.2904, 2.838), P = 0.7729
Redness	13/169 (7.7%)	Ref.	4/80 (5.0%)	1.583 (0.4993, 5.020), P = 0.5928
Overheating	3/160 (1.9%)	Ref.	0/80 (0%)	3.578 (0.1824, 70.162), P = 0.5528
Pain	5/163 (3.1%)	Ref.	0/80 (0%)	5.587 (0.3049, 102.37), P = 0.1746
Swelling	8/163 (4.9%)	Ref.	0/80 (0%)	8.801 (0.5012, 154.53), P = 0.0555
Functional impairment	1/160 (0.6%)	Ref.	0/80 (0%)	1.514 (0.06095, 37.616), P = 1.0000
Blood stains at three-way stopcock	26/327 (8.0%)	Ref.	9/80 (11.3%)	0.6814 (0.3059, 1.518), P = 0.3733
Intact dressing	313/327 (95.7%)	Ref.	70/80 (87.5%)	3.194 (1.362, 7.488), P = 0.0134
Bloody dressing	57/327 (17.4%)	Ref.	13/80 (16.3%)	1.088 (0.5627, 2.104), P = 0.8701
Documented indication	309/327 (94.5%)	Ref.	76/80 (95.0%)	0.9035 (0.2971, 2.748), P = 1.0000

n.a. = not applicable. n.e. = not estimable. PVC = peripheral venous catheter. CVC = central venous catheter.

Focusing on observed differences, it was noted that checkpad-based digital documentation was more compliantly conducted in case of CVCs compared to PVCs. In contrasts, CVCs were less likely to have an intact dressing. Other significant differences were not recorded. In summary, the findings suggest slightly more stringent management of CVCs, which are typically intended for longer-term use. However, despite their longer average dwell time, the quality of CVC dressings was inferior to that observed for PVCs.

3.4. Associations between the duration of device use and signs of inflammation

When comparing PVCs and CVCs with and without signs of inflammation, respectively, significant difference could not be observed within the assessed categories. However, when PVCs and CVCs with signs of inflammation were compared, CVCs had generally been in place for longer periods at the time points of assessment than PVCs. Detailed results are presented in the Tables 4 and 5. Accordingly, the principal finding was the longer catheter indwelling time of CVCs at the time points when inflammatory signs were detected.

Table 4.

Comparison of duration of venous catheter application with and without signs of inflammation for PVC and CVC applying non-parametric Mann-Whitney-U testing

Compared groups	Mean duration in days (± SD) with signs of inflammation	Mean duration in days (± SD) without signs of inflammation	Significance
PVC	3.0 (±1.8)	3.3 (±4.2)	P = 0.3916
CVC	6.8 (±3.5)	6.7 (±5.6)	P = 0.7156

Table 5.

Comparison of duration of PVC an CVC application in case of signs of inflammation applying non-parametric Mann-Whitney-U testing

Compared groups	Mean duration in days (± SD) for PVC application	Mean duration in days (± SD) for CVC application	Significance
Detected signs of inflammation	3.0 (±1.8)	6.8 (±3.5)	P = 0.0335

4. Discussion

The study was conducted to contribute to the still limited availability of data on the management, integrity, and complications of both PVCs and CVCs in German tertiary care hospitals. In a previous cross-sectional study [29], a recorded proportion of PVC use of 33% was lower than observed in our assessment. Also, the rate of recorded signs of infection in our study was more than twice as high compared with the recently detected proportion of superficial PVC infections of 2.3% [29]. The latter finding may be attributed to the fact that we counted all signs of inflammation rather than just proven infection alone. Further, the data are not directly comparable, due to the recent study's focus on peripheral wards [29] with usually less severely ill patients compared with intermediate or intensive care units, partially different professional focus of the hospitals, as well as exclusion of psychiatric wards, where patients are less frequently equipped with intravenous devices, in our study. Insofar, the round about 5% signs of inflammation at both peripheral and central devices identified by us is in a similar one-digit percentage range as compared with previous findings [29]. Summarized, the proportion of detected inflammatory signs was within the expected range.

The conducted comparison of PVCs and CVCs resulted in a number of interesting findings as well. First of all, the proportions of inflammatory signs at PVCs and CVCs was quite similar, potentially due to standardized handling of intravenous devices at our hospital. This is in some contrast to recent findings, which postulated higher infection risks associated with CVCs compared to PVCs [5]. Admittedly, signs of systemic infection were not recorded at all in our assessment, not allowing any conclusions on this disease entity. At least, the recorded inflammatory signs did not indicate substantially divergent inflammation risks between PVCs and CVCs on the assessed hospital wards.

As recent work has associated shorter duration of PVC application with reduced infection risk [4], it seems noteworthy that PVCs showing signs of inflammation had been in place for an average of 3 days. In contrast, observed CVCs with signs of inflammation had been in place for a week on average. However, when independently assessing the categories of PVCs and CVCs, devices with and without signs of inflammation were not associated with different time periods of use. This is well in line with previous observations that routine exchange intervals do not show advantages compared with exchange intervals guided by the clinical state of the catheter site [1]. The findings confirmed the hypothesis that CVCs may be in place for longer periods at the time point when inflammatory signs are detected.

It remains speculative whether or not the overall longer duration of CVC use contributed to findings like increased rates of incrustation at the insertion site and poorer states of applied dressings compared to PVCs. At least the hypothesis seems plausible, although the observed especially high adherence to digital documentation on CVC use suggests understanding of the need for CVC-associated infection prevention and control measures on the investigated wards. Overall, catheter management quality and catheter integrity were comparable for PVCs and CVCs in the assessed inpatient setting.

The study has a number of limitations. First of all, the total number of observed patients was only about one third of the number recorded in a previous similar assessment [29], reducing the likeliness of observing rare events like device-associated infections. However, we considered not only absolute numbers but more the composition of assessed patients to be of interest. Second, the distribution of this comparably small number on 40 wards made sub-stratifying at ward level or at insertion site level not promising. Accordingly, we abstained from such analyses. Third, lacking assessment of patient-related data made respective epidemiological analyses unfeasible. Thus, the study was limited to catheter status assessments, and the possibility that minor differences were missed cannot be ruled out.

5. Conclusions

Despite the above-mentioned limitations, the study provides an overview on the integrity of venous catheters and their dressings in a German tertiary hospital. Low rates of inflammatory complications were observed for both PVCs and CVCs, with inflammatory signs in CVCs occurring after longer indwelling times than in PVCs. However, the catheter indwelling time was not independently associated with inflammatory signs in either catheter type. Thus, inflammation signs appear to occur as isolated events in case of both catheter types, underscoring the need for continuous vigilance and careful clinical assessment.

Abbreviations The following abbreviations are used in this manuscript:

CVC

central venous catheter

d-SAB

delayed onset S. aureus bacteremia

n.a.

not applicable

n.e.

not estimable

PVC

peripheral venous catheter

Ref.

reference

Funding Statement

Funding: There has not been any source of funding for this investigator-initiated work.